DEPARTMENT OF MARINE BIOLOGY OF THE CARNEGIE INSTITUTION OF WASHINGTON ALFRED G. MAYOR, Director PAPERS FROM THE DEPARTMENT OF MARINE BIOLOGY OF THE CARNEGIE INSTITUTION OF WASHINGTON VOLUME XVllI Published by the Carnegie Institution of Washington Washington, November, 1922 CARNEGIE INSTITUTION OF WASHINGTON Publication No. 312 L CoplGsoftbisbooi:] DEC 4 1922 TECHNICAL PRESS WASHINGTON, D. C. CONTENTS. PAGE I. Studies on the Hybridization of Echinoids, Cidaris tribuloides. By David H. Tennent. 3 plates, 28 figures 3-20 Part I. Embryology and Hybridization of Cidaris 3 Nature and Systematic Position of the Material 3 Early Development of Cidaris trihvloides 5 Rate of Development of Cidaris compared with that of other Echinoids . . 6 Comparison with other Investigations of Cidaris 9 Formation of Mesenchyme in Echinoderms 1 1 Observations on Hybridization of Cidaris 13 Influence of Spermatozoon on Production of Paternal Characters 15 The Spermatozoon and FertiHzation 16 Cross-Activation versus Cross-Fertihzation 17 Classification of Hybrids 18 Part II. Cytology of the Eggs of Cidaris tribuloides and of Cross-Activated Cidaris Eggs 21-42 Structure of the Cidaris Egg 21 Study of Chromosomes in Species-Fertilized Cidaris Eggs 23 Chromosomes in Cidaris Eggs Cross-Activated by Lytechinus Sperm 27 Chromosomes in Cidaris Eggs Cross-Activated by Tripneustes Sperm 29 Elimination of Chromosomes 32 Alteration in Physical Characteristics of the Cytoplasm of the Egg by Action of Foreign Sperms 34 Nuclear Enzymes and Cytoplasmic Substrate 35 The Binuclearity Hypothesis 35 Discussion 37 Summary 40 Conclusions 4j Literature 42 II. The Production of Ldght by the Fishes Photoblepharon and Anomalops. By E. Newton Harvey. 1 figure 43-60 General Account of the Fishes 45 Histology of the Light-Organ /. . 48 Lack of Oxygen 5I Desiccation g2 Dilution ' | 53 Duration of Luminescence 53 Lucif erin and Luciferase 54 Temperature ' 55 Spectrum and Intensity 55 Cy tolysis 56 Sodium Fluoride 57 Potassium Cyanide 57 Culture Experiments 58 Summary en Literature ." gQ III. Hydrogen-ion Concentration and Electrical Conductivity of the Surface Water of the Atlantic and Pacific By Alfred Goldsborough Mayor. 3 charts. .61-85 Observations gc Literature cited or Explanation of the charts .........". 85 IV. Carbon-Dioxide Content of Sea-Water at Tortugas. By Roger C. Wells. 1 fig. . . 87-93 Introduction and Summary of Results gg Methods g'j General References g3 III IV Contents. PAGE V. Analytical Search for Melals in Tortugas Marine Organisms. By Alexander H. Phillips 95-99 VI. The Tracking Instinct in a Tortugas Ant. By Alfred Goldsborough Mayor 101-107 Experiments 104 VII. A Collection of Fishes from Samoa. By Henry W. Fowler and Charles F. Silvester. 2 figures 111-126 Ophichthyidse 112 Chlevastes colubrinus (Boddaert) 112 fasciatus (Ahl) 112 Muraenidae 112 Gymnothorax punctatus (Schneider) 112 pictus (Ahl) 113 Anarchias allardicei (Jordan and Seale) 114 Hemiramphidae 114 Hyporhamphus pacificus (Steindachner) 114 Mugilidae 115 Neomyxus chap tali (Eydoux and Souleyet) 115 Holocentridae 116 Holotrachys lima (Valenciennes) 116 Holocentrus punctatissimus Cuvier 116 diadema Lac6pede 116 praslin Lac^pMe 117 sammara (Forskal) 117 Cheilodipterida} 117 Amia savayensis (Giinther) 117 novemfasciata Cuvier 117 Fowleria marmorata (Alleyne and Macleay) 117 Serranidse 118 Epinephelus merra Bloch 118 Pharopteryx nigricans Riippell 118 melas (Bleeker) 118 Opisthognathidae 118 Gnathypops samoensis, new species 118 Pomacentridae 120 Pomacentrus melanopterus Bleeker 120 nigricans (Lac^pede) 120 albofusciatus Schlegel 120 Abudefduf ccelestinus (Cuvier) 120 glaucus (Cuvier) 120 zonatus (Cuvier) 120 Dascyllus aruanis (Linnaeus) 121 Chromis cseruleus (Cuvier) 121 isomelas Jordan and Seale 121 Labridae 121 Platyglossus notopis (Valenciennes) 122 CheUinus fasciatus (Bloch) 122 Scarichthyidae 122 Callyodon rubro-violaceus (Steindachner) 122 Chaetodontidae 123 Chaetodon trifasciatus Forskal 123 pelewensis I^er 123 melannotus Schneider 123 mUiaris Quoy and Gaimard 123 Holacanthus nicobariensis (Schneider) 123 Acanthuridae 124 Hepatus atrimentatus Jordan and Evermann 124 triostegus (Linnaeus) 124 Monacanthidae 124 Oxymonacanthus longirostris (Schneider) 124 Contents. y VII. A Collection of Fishes from Samoa -Continued. ^^^^ Tetrodontidae Canthigaster solandri (Richardson). . I94 Scorpaenidae '* Sebastopsis guamensis (Quoy and Gaimard) lol scabra (Ramsay and Ogilby) . . ,94 bebastopistes laotale Jordan and Seale ' ' ' ' loT Gobiesocidae * Crepidogaster samoensis Steindachner lot Gobiidae ^'*'* Eviota zonura Jordan and Seale jol afelei Jordan and Seale jo^ distigma Jordan and Seale 12^ Pseudogobiodon citrinus (Riippell) 10? Blenniidae ^"^^ Enneapterygius tusitalae Jordan and Seale Jo^ Salarius variolosus Valenciennes Jo^ gibbifrons Quoy and Gaimard ,9r Alticus biseriatus (Valenciennes) ,9^ Salarias rivulatus RUppell. . j^, Enchelyurus ater (Guntherj tZ^ Fierasferidae ^^" VTTT T J- ... . Jordanicus parvipinnis (Kaup) ..'.'.'.'.'.'.".' fo^ VIII. Leod^c^d I'^/t/ «. 23 ^' <^ 534 ^, / If'/ "•'/// // // Z//'^ /~t / Fio. 23, a and b, Cidans 9 X Tripneustes d" . Two V's and hook, 22-26. , • , , , Fig. 24, o and b, Cidaris 9 X Tripneustesd'. Early anaphase; the position and width of the chromosomes in b give clear evidence that these are merely sections of tho.se shown in a. Two V's, Elimination of Chromosomes. In 1912 I summarized the facts regarding retention or elimination of chromosomes in cross-activated Echinoid eggs as follows: (1) Elimination of no chromosomes and dominance of one species over the other with respect to the character of the skeleton. (2) Elimination of part of the chromosomes and dominance of one species over the other with respect to the character of the skeleton. (3) Elimination of no chromosomes and skeleton of intermediate character. (4) Elimination of part of the chromosomes and skeleton of intermediate character. (5) Elimination of part of both maternal and paternal chromosomes and inhibition of development. Hyhridizaiion of Ec/iwoids. 33 In group (2) I included such cases as result from the activations of the Echinoid egg by the sperm of annelids and mollusks (Godlewski, 1911; Kupelwieser, 1906) and the fertilization of eggs which had been given a certain impulse to parthenogenetic development by means of chemicals, which give, as these authors have pointed out, thely- karyotic larvse. I stated further that, practically, larvae derived from such crosses inherit from the egg parent alone just as strictly as if the eggs had been caused to develop from the first by artificial chemical fertihzation. They are false hybrids. 0^26 ^ "At J 25 ^A 5/ n ^_27 28 A » Fig. 25, Cidaris 9 X Tripneuslesd'. Median section of three, other two not shown, showing cliroinosomes becoming transformed to chromosomal vesicles before contraction of chromosome has taken place. X 2,400. Fig. 26, Parthenogenetic Cidaris egg. Prophase of monaster division. X 2,400. FiQ. 27, a and 6, Parthenogenetic Cidaris egg. Two successive sections of monaster plate. Many elements shown in b are sections of those shown in a. X 2,400. Fia. 28, Parthenogenetic Cidaris egg. Monaster division. All chromosomes have divided; 18 pairs. X 2,400. These Cidaris-LytecJiinus and Cidaris-Triprieustes crosses consid- ered in this paper constitute another group. (6) Elimination of part of the chromosomes, presumably paternal and failure in development. These may be regarded as partial hybrids. Probably several additional crosses that have been made, but which have not received adequate cytological examination, will ultimately be found to belong in this group. In these crosses, development seems to proceed well as long as the generalized type of development, common to each, continues; devel- opmental block occurs when the period of specialized development begins. The egg seems unable to accommodate itself fully to the impulse for a divergent type. 34 Hyhridization of Echinoids. Alteration in Physical Characteristics of the Cytoplasm of the Egg by Action of Foreign Sperms. In my interpretation of the phenomena exhibited in the develop- ment of Cidaris and its hybrids I have been influenced, to a consider- able extent, by the consideration of facts concerning the transformation of deutoplasmic materials to cytoplasm. The Cidaris egg, despite its transparent character, contains a relatively large amount of yolk and fat. This material gradually disappears as development con- tinues. I have also been guided by results which I have obtained from another inter-ordinal cross, Arbacia X Moira (Tennent, 1920). It is unnecessary to review in this place the facts presented in that paper, and it seems sufficient to say that certain granules and rods, present in sections of cross-activated eggs and absent in sections of straight-fertilized eggs, were explained as coarse precipitates formed as a result of the emission of a foreign enzyme from the nucleus. Even though we were to regard these bodies as artifacts produced by the fixing fluid used (acetic-sublimate) — for we know that certain metallic salts may cause the precipitation of proteids — the fact remains that physical conditions in these eggs must have been differ- ent than in the straight-fertihzed eggs which were fixed at corre- sponding stages and in the same fixing fluid, and which show no bodies of this kind. In continuation of the investigation mentioned above, one of my students, Miss Hibbard, has made a study of Tripneustes eggs fertil- ized with Lytechinus sperms, comparing these with the straight- fertihzed and parthenogenetic eggs, and has found that in the cross- activated eggs which are developing normally granules appear and later disappear as the nucleus passes into the later phases of division. Here, again, the only explanation seems to lie in the action of the enzyme introduced by the foreign spermatozoon. Miss Hibbard has also determined that some of the cross-activated eggs, which segment less normally, show other and characteristic changes in their cytoplasm. These changes are clearly pathological. The most striking result of the removal of cortical block followed by cross-activation that has come within my experience is that following the insemination of the Lytechinus egg, after preliminary treatment with NaOH or CaCl, with the sperms of Holothuria floridana (Tennent, 1911, p. 140). In this instance the sperm, upon entering, in most cases tears the egg to pieces. A deep notch or pathway made by the sperm may be seen and then the egg suddenly disintegrates. In other cases the entrance of the spermatozoon is followed by a slower, but none the less complete, cytolysis of the egg. Gray (1913) studied the eggs of Echinus acutus fertilized by the sperm of Echinus escidentus and species-fertilized eggs of Echinus Hybridization of Echinoids. 35 acutus treated by hypertonic solutions. The results were similar. There was an elimination of a certain number of chromosomes from the nucleus in both the cross-activated eggs and in the straight-fer- tilized eggs which had been treated with hypertonic solutions. The eggs of Echinus esculentus were not affected by similar treatment. Gray suggests, tentatively, after consideration of the results of McClendon (1913) and R. S. Lillie (1909, 1911), on increased perme- ability of the egg membrane after the entrance of the spermatozoon, that the degree of change in permeability after fertihzation is a func- tion of the sperm and that the cytological behavior of reciprocal crosses is explicable on this hypothesis. Gray points out that changes in permeability of the cytoplasm as induced by the sperm can not be a sufficient explanation of all abnormalities observed in cross- fertilized eggs. Doncaster and Gray (1913), in considering the probable fact that the vesicles in the eggs of the cross acutus X esculentus are derived from the acutus chromosomes, suggest that the explanation may lie in an alteration of the permeability or osmotic condition of the egg, consequent upon the development within it of a foreign spermatozoon, i. e., the physical condition of the cytoplasm is altered by the develop- ment within it of a foreign sperm nucleus. Nuclear Enzymes and Cytoplasmic Substrate. The study of a sufficient number of species, both closely related and widely separated, has now given, I believe, an adequate basis for generalization. The egg contains the mechanism necessary for development. It may develop parthenogenetically, giving a maternal larva and adult. In that development there is a harmonious interaction between the nuclear and cytoplasmic material. The results of cross-activation in which there is a complete rejection of paternal nuclear material and production of false hybrids, as in the Echinoid-Mollusk, or Echi- noid-Annelid cross, are closely related to examples of artificial par- thenogenesis. Above this level there are crosses showing maternal or paternal influence in lesser or greater degree. We might pass in review a great number of cases of^ partial hybridization and reach finally undoubted examples of true hybridization. If we restate these facts on the basis of chromatin content, we find that in false hybrids there is a complete elimination of paternal chromosomes, in partial hybrids a partial elimination of chromosomes, and in true hybrids no elimination of chromosomes. The fact that the results obtained from a given cross differ from those obtained in the reciprocal cross indicates that the differences are not due to lack of harmony between the substances of the germ 38 IJyhridizaiion of Echlnoidfi. nuclei. It is clear that cytoplasmic differences must be taken into account. The life of the cell lies in the interaction between nucleus and cytoplasm. The relative importance of each is that of enzyme to substrate. There can not be cytoplasmic control, nor can there be nuclear control, purely as such, for the processes of life lie in the interaction of both. There is nuclear control in that by the intro- duction of enzymes we obtain synthetic products; there is cytoplasmic control in that with a given substrate we provide material from which products are synthesized. From the evidence given by Arhacia-Moira material (Tennent, 1920), it seems clear that the reactions in the cj^toplasm caused by foreign sperms are unlike the reactions caused by species sperms. That fact fixes the attention on the nature of the activities of the nucleus. The Binuclearity Hypothesis. Various binuclearity hypotheses, founded in part on Richard Hertwig's chromidial hypothesis, have influenced interpretations of nuclear phenomena. The nucleus shows two phases, an active and a resting. These two phases, kinetic and interkinetic, have been made to lend themselves to an analogy with a true binucleate con- dition and to an assumption of dichromaticity. Of these two kinds of chromatin, one is supposed to be propagatory (idiochromatin), in evidence at the time of cell division; the other trophic (tropho- chromatin, somatochromatin), formed by the idiochromatin, but resident in the cytoplasm. The somatic phase of the nucleus covers a period during which it may be assumed that nuclear enzymes have passed from the nucleus into the cytoplasm, and the cytoplasm has become the seat of synthetic activities. The nucleus at this time is in a ''resting" condition; it seems comparatively empty, is acido- phile, and basophilic bodies may be found in the cytoplasm. Under the influence of the chromidial hypothesis, supposed particles of chromatin in the cytoplasm have been interpreted as chromidia, or as trophochromidia. The basophilic bodies found in cross-activated Arbacia eggs might have been interpreted similarly had there been any evidence that they were emitted from the nucleus. The possi- bility of the emission of large particles is foreign to our conception of the nature of the nuclear membrane. Hyhridization of Echidoids. 37 DISCUSSION. Masing (1910), selecting for study one of the nuclear constit- uents, nucleic acid, found that while during development in Arbacia pustulosa up to the blastula stage there is an increase of about a thousand fold in nuclear mass, there is no perceptible increase in the nucleic-acid content of the egg. He reached the conclusions that the nucleic acid of the cleavage nuclei arises from a preformed stock in the egg plasma and that nucleic acid and the chromatin of the his- tologists must be different things. (Quoted from Godlewski, 1911.) During the period studied by Masing there has been no increase in the mass of organism. Godlewski (1908) showed that the total volume of plasma in the blastula is about one-third less than in the unsegmented egg. The nuclear material has increased in volume at the expense of the cytoplasm. Godlewski's investigations show a change in ratio between cytoplasm and nucleus from 550:1 in the unfertilized egg to 6:1 in the blastula. In the light of the results of Conklin (1912) on the growth of protoplasm during cleavage, and in the light of my own results from a study of the eggs of Arbacia punctulata, as yet unpublished, and of those of Hibbard on the eggs of Echinarachnius parma, in press, these figures need some revision. This revision will not affect Godlewski's general conclusions. The unpublished results mentioned above show that in the eggs of Arbacia punctulata and Echinarachnius parma there is a considerable mass of deutoplasmic material, which is in process of transformation to cytoplasm during the cleavage stages. In the echinoderm egg, despite its frequent transparent character, there is a stock of yolk and fat. Cell volume in these eggs is not equivalent to cytoplasmic volume. The growth in nuclear mass is due, in part at least, to the assimilation of deutoplasm. Marcus (1906) and Erdmann (1908) have shown that the size of the chromosomes diminishes during cleavage. Erdmann has concluded that the chromosomes of the pluteus of Strongylocentrotus ha\'e only about one-fortieth the volume of those of the first cleavage spindle. She has also determined that there is a constant increase in total quantity of chromatin. Her figures show that at the stage of the blastula without mesenchyme, which Godlewski has estimated as a 1,256-cell stage, the chromosomal volume is 78 times greater than it was at the time of the first division. This would amount to an increase of about 6 per cent with the division of each cell. Godlewski (1908) estimated that the total volume of nuclear material is 47 times greater in the 1,256-cell stage than in the unfertilized egg, which would be 23.5 times the total volume in the fertilized egg. Inasmuch as he believed that practically all of this growth took place between the 1-cell stage and the 64-cell stage, and that but little increase of 38 Hyhridization of Echinoids. nuclear mass occurred between the 64-cell stage and the blastula stage, this would mean an increase of about 37 per cent with the division of each cell. Godlewski speaks of the increase as being nearly in a geometric ratio, this calculation being based on the volume of nuclear mass in the unfertilized egg. It should be noted that Erdmann's figures are of volume of chromosomes and that Godlew- ski's are of nuclear volume. Conklin (1912) has estimated that in Crepidula the chromosomal mass grows at the rate of 8 per cent for each division up to the 32-cell stage. Baltzer (1909) has indicated the large probable error in estimates of total chromatin volume that are based on the assumption that the chromosomes in an Echinoid egg are of equal size. Mathews believes that nucleic acid is the chromatin of the his- tologists. We have seen that JMasing found no evidence of increase of nucleic acid during development in Arhacia up to the blastula stage. The evidence of Conkhn, Erdmann, and Godlewski is con- clusive in its demonstration of increase of total volume of chromo- somes during this period. If there has been no increase in the amount of nucleic acid, there must have been increase in its basic protein portion. Goldschmidt's (1917) idea that the chromosomes act as adsorbents of enzymes which constitute the chemical basis of heredity finds much to support it in the facts demonstrated in this paper. We are able to show the presence of nucleic acid by our stains; we have no means of making the nuclear enzymes visible. Our only test for them is in the things that they do. We can not hope to demonstrate consecutive stages in processes of synthesis by specific stains. We can demonstrate, and the material discussed in this paper has demonstrated, the fact that foreign nuclear material produces reac- tions in the cytoplasm that do not occur in straight-fertilized eggs. Closeness of relationship is by no means indicative of the readiness with which the initial impulse to development may be received, nor a sure criterion of the extent to which it may proceed. In the Cidaris- Lytechinus and Cidaris-Tripneustes crosses under consideration there seems to be little cortical block to the entrance of the spermatozoon. There is little internal block during early development. Develop- ment proceeds regularly to the period immediately before gastrula- tion. To this point it has been following the general path of develop- ment taken by most Echinoderms. At the point of deviation of special from general, abnormalities appear. Development ends in the gastrula stage, as in many crosses between Echinoderms. The cells which should have gone to the completion of the archenteron and the prospective mesenchyme cells are the first to die. Sections show the nuclei to be swollen with chromatin. The nuclei finally burst, extruding irregular masses of chromatin into the cytoplasm. Hybridization of Echinoids. 39 Developmental block in this case seems to be associated with what might be described as nuclear indigestion. Fulton (1921, p. 167), in describing the result of continued stimulation of a muscle on its nucleus, says: ''With continued stimulation the nucleus must liter- ally become overwhelmed with material to be oxidized; as a result, it would become fatigued, and would no longer be capable of perform- ing its normal functions." The word ''overwhelmed" describes vividly the conditions in- volving the nucleus in these hybrids. From the Arbacia-Moira material, the evidence obtained was interpreted as showing that enzymes are emitted through the nuclear membrane and that changes occur in the protoplasm following this emission. The materials formed pass back to and are taken into the nucleus. There is no evidence that the emitted material is chromatin. There is nothing in these views that detracts from the chromosomal theory of inheritance. It is evident that what is transmitted from generation to generation is a course of development— what Brooks in "The Foundations of Zoology" calls a "capacity for nurture." The results of these cross-activations show that an orderly series of developmental reactions may be disorganized by the introduction of foreign nuclear material. In the cross-activated eggs, in the inter- action between nucleus and cytoplasm, the nucleus seems to be pro- vided with more material than it can handle. In straight-fertilization the accumulation of product would have checked the action of enzymes before that condition arose. The foreign enzymes are specifically different and their action is not checked until a greater than normal supply for this egg has accumulated. Even though the nucleus may take care of products formed, development may come to an end because of an actual exhaustion of substrate. 40 Hyhridizalion of Echinoids. SUMMARY. Part I. 1. Cidaris tribuJoides is a primitive Echinoid whose early develop- ment is less modified than that of the more modern Echinoids. 2. In its normal development it is of interest (1) because of the slowness of its development when compared with Lyiechinus and Trip- neustes, (2) in the difference in site of the formation of its mesenchyme, (3) in the place of appearance of the larval skeleton, and (4) in the form of the larva. 3. The eggs of Cidaris are activated easily by the sperms of Lyiechinus or Tripneustes. These give interordinal crosses. 4. The larvse obtained from such cross-activations die in the gastrula stage. 5. Cross-activation with these foreign sperms causes mesenchyme cells to pass into the blastocoele before the beginning of gastrulation, while in the species-fertilized eggs immigration of mesenchyme cells does not occur until after the archenteron has been formed. 6. It is suggested that the use of the term internal block be limited to block, or inhibition of the conjugation of the germ nuclei, and that the term developmental block be used in designating later inhi- bitions of development. 7. The use of a modification of the Giinther Hertwig classification of hybrids is suggested. Part II. 1. The eggs of Cidaris, although traUvsparent, are rich in fat and yolk. 2. A study of species-fertilized eggs shows that half of the eggs contain 37 chromosomes and half 38. Eggs of the group containing 37 chromosomes have one V-shaped element, while those of the group containing 38 chromosomes have two V-shaped elements. 3. Parthenogenetic eggs have 19 chromosomes. 4. Cidaris sperms are dimorphic, half containing a V-shaped chromosome, half being without this element. Half of the sperma- tozoa contain 19 chromosomes, half contain 18. 5. Some of the paternal chromosomes lag in division and are eliminated during cleavage in cross-activated eggs. 6. Investigations on Lyiechinus- Tripneustes and Arbacia-Moira material show alteration in the physical characteristics of the cyto- plasm of the egg by action of the foreign sperm. 7. The phenomena exhibited suggest that the effects are due to the action of foreign enzymes on the cytoplasmic substrate. 8. The nucleus in cross-activated Cidaris eggs seems to be supphed with more material than it can utilize. HyhruHzalion of Efhinoiih. 41 9. The facts suggest an emission of enzymes ratiier than an emis- sion of chromatin. 10. The results show that in the cross-activated Cidaris eggs an orderlj'^ series of developmental reactions is disorganized by the foreign nuclear material at the time when divergence of two systems of development occurs. 11. Through its organization and in its capacity as substrate the egg fixes the course of development. CONCLUSIONS. The consideration of the facts established during the study of the material which has formed the basis of this paper gives a deep impres- sion of the fundamental fact of organization. It gives additional proof of the existence of an underlying basis for development, which we may be able to distort, but which we are not yet able to shape at will. The normal development of the Cidaris egg is of a less-specialized type than that of the eggs of the species whose sperms were used in the cross-activations. As long as the two courses of developm_ent lie parallel, we say that development is normal. When the point of divergence between the two paths is reached, characters appear which we call aberrant. Differentiation lies in a series of reactions between nucleus and cytoplasm. In attempting to superimpose a specialized on a non-specialized type of development we fail, because of our lack of ability to harmonize two disharmonious systems of de^-elopment. The consideration of this material emphasizes again the fact that the thing inherited by offspring from parent is the capacity for development. What that development will be depends on the interactions between nucleus and cytoplasm and on adjustment to environment. The cytoplasm is the material that is shaped during the series of reactions. It is because of the fact that the cytoplasm of the egg is the material basis of the body that Conklin's statement that the egg cytoplasm ''fixes the general type of development" is true. juJ L I I 42 Hybridization of Echinoids. LITERATURE. Baltzer, F. 1009. Die Chromosonien von Strong ylocentrolus lividus und Echinus micro- iuberculnlus. Arch. f. Zellforsch., Band 2. Clark, H. L. 1907. The Cidarida. Bull. Mus. Comp. Zool., Harvard Coll., vol. 51, No. 7. . 1912. Hawaiian and other Pacific Echini. Mem. Mus. Comp. Zool., Harvard Coll., vol. 34, No. 4. CoNKLiN, E. G. 1912. Cell size and nuclear size. Jour. Exp. Zool., vol. 12. . 1915. Heredity and environment in the development of men. Princeton Univ. Press. DoNCASTER, L., and J. Gray. 1913. Cytological observations on the early stages of segmentation of Echinus hybrids. Quart. Jour. Mic. Soi., vol. 58. Erdmann, Rh. 1908. Experimentelle Untcrsuchimg der Massenverhaltnisse von Plasma, Kern und Chromosomen in dem sich entwickelnden Seeigelei. Arch. f. Zellforsch., Band 2. Fulton, John F., jr. 1921. Studies on neuromuscular transmission: I. Amer. Jour. Physiol., vol. 57. GoLDscHMiDT, R. 1917. A further contribution to the theory of sex. Jour. Exp. Zool., vol. 22. GoDLEW'SKi, E., JR. 1908. Plasma luid Kernsubstanz in der normalen und der durch aussere Faktoren veninderten Entwicklung der Echiniden. Arch. f. Entw. Mech., Band 26. . 1911. Studien tiber Entwicklungserregung. Arch. f. Entw. Mech., Band 33. Gray, J. 1913. The effects of hypertonic solutions upon the fertilised eggs of Echinus. Quart. Jour. Mic. Sci., vol. 58. Guthrie, M. J., and H. Hibbard. 1919. Cleavage and mesenchyme formation in Toxopneustes variegatiis. Biol. Bull., vol. 37. Hertwig, GiJNTHER. 1918. Kreuzungsversuche an Amphibien. Wahre und falsche Bastarde. Arch. Mikr. Anat., Band 91. Jackson, Robert Tracy. 1912. Phylogeny of the Echini, with a revision of the Palsozoic species. Mem. Boston Soc. Nat. His., vol. 7. LiLLiE, F. R. 1919. Problems of fertilization. Univ. Chicago Press. LiLLiE, R. S. 1909. The general biological significance of changes in permeability of the surface layer or plasma-membrane of living cells. Biol. Bull., vol. 17. . 1911. The physiology- of cell division: IV. Jour. Morph., vol. 22. Marcus, H. 1906. iJber die Wirkung der Temperatur auf die Furchung bei Seeigeleiern. Arch. f. Entw. Mech., Band 22. McClendon, J. F. 1912. The osmotic and surface tension phenomena of living elements and their physiological significance. Biol. Bull., vol. 22. Mortensen, Th. 1920. Studies in the development of Crinoids. Papers from Depart- ment of Marine Biology, Carnegie Inst. Wash. Pub. No. 294. . 1921. Studies of the development and larval forms of Echinoderms. Published at expense of Carlsberg fund, G. E. C. Gad., Copenhagen. OsHiMA, H. 1921. On the development of Chtcumaria echinata. Quart. Jour. Mic. Sci., vol. 65. Pinney, Edith. 1911. A study of the chromosomes of Hipponoe esculenta and Moira atropos. Biol. Bull., vol. 21. . 1918. A study of the relation of the liehavior of the chromatin in development and heredity in teleost hvbrids. Jour. Morph., vol. 31. Poll, H. 1920. Mischlingsstudien VIII. Arch. f. Mikr. Anat., Band 94. Prouho, H. 1887. Recherohes sur le Dnmcidaris papillata et quelques autres fichinidea de la Mediterranee. Arch. Zool., espcr. gener. 2., Ser. V. Tennent, D. H. 1911. Echinoderm hybridization. Papers from Department of Marine Biology, Carnegie Inst. Wash. Pub. No. 132. . 1 91 la. A heterochromosomc of male origin in Echinoids. Biol. Bull., vol. 21. . 1912. Studies in cytology. Jour. Exp. Zool., vol. 12. . 1914. The early influence of the spermatozoon upon the characters of Echinoid larvae. Papers from Departm.ent of Marine Biology, Carnegie Inst. Wash. Pub. No. 182. . 1920. Evidence on the nature of nuclear activity. Proc. Nat. Acad. Sci., vol. 6. ILLUSTRATIONS. Plate 1, All of the figures on this plate were drawn with a Zeiss compensating ocular 8 and 2-mm. immersion objective, giving a magnification, when drawn with a camera lucida on an inclined drawing-board slightly above table level, of about 1,800 diameters. These figures have been reduced one-half in reproduction. Fig. a, Cidaris X Cidaris. Anaphase first cleavage. Fig. b, Cidaris X Cidaris. Later anaphase first cleavage. Fig. c, Cidaris 9 X Tripneustesd'. Telophase first cleavage. Zwischenkorper at line of contact between cells. Fig. d, Cidaris 9 X Tripneustes d' . Late telophase second cleavage. Fig. e, Cidaris 9 X TriprmtMes (f . Earlier telophase first cleavage. Plate 2. Magnification and reduction as in plate 1. Fig. a, Cidaris 9 X TripneustesiS'. Anaphase second cleavage; lagging chromosomes at center. Fig. b, Cidaris 9 X Tripneuslesd'. Second cleavage completed; Zwischenkorper. Fig. c, Cidaris X Cidaris. Prophase of mitosis; fat droplets small and have been moved out from region of nucleus. Fig. d, Cidaris X Cidaris. Late prophase of mitosis; region of asters freed from fat droplets. Fig. e, Cidaris 9 X Lytechinm<^. Large fat droplets surrounding nucleus. Fig. f, CidarisX Cidaris. Late telophase; chromosomal vesicles; small fat droplets moving in. Fig. g, Cidaris X Cidaris. Chromosomal vesicle; large fat droplets. Figs, h to l, Cidaris X Cidaris. Small area of cytoplasm from one side of the amphi- aster during anaphase of mitosis, showing variation in size and relative abun- dance of droplets. Fig. h, Clear area surrounding each droplet. Fig. I, Cytoplasm with a few large droplets. Fig. j, Cytoplasm with both larger and smaller droplets. Fig. k, Cytoplasm with numerous small droplets only. Fig. l, Cytoplasm with numerous large droplets. Plate 3. In all of the detailed figures on this plate only the archenteron and a small portion of the posterior end of the gastrula are shown. The relation of the part shown to the entire gastrula may be seen by reference to figure j. All figures except figure j were drawn at a magnification of 1,900 diameters. They have been reduced one-half in reproduction. Fig. A, Cidaris 9 X Lytechijimd'. Blastula, 18 hours. One cell in blastocoele; cells in wall seem to be in process of withdrawal. Fig. b, Cidaris 9 X Lytechinus d' ■ Gastrula, 23 hours. Seven cells in blastocoele. Fig. c, Cidaris 9 X Lytechinus d . Gastrula, 24 hours. Fig. d, Cidaris 9 X Lytechinusd. Gastrula, 24 hours. Fig. e, Cidaris 9 X Lytechinmd'. Gastrula, 40 hours. Fig. f, Cidaris X Cidaris. Blastula, 18 hours. Fig. g, Cidaris X Cidaris. Gastrula, 18 hours. Fig. h, Cidaris X Cidaris. Gastrula, 23 hours; 2 mesenchyme cells are in process of withdrawal from wall of archenteron. Fig. I, Cidaris X Cidaris. Gastrula, 24 hours; 5 mesenchyme cells in blastocoele at plane of section. Fig. j, Cidaris gastrula, 23 hours; X 180. Camera sketch of living gastrula; 3 mesen- chyme cells in the blastocoele. ♦ * . •• W''. ' • ♦ Jj' "■'■•:: c •• • *v 't . * ^ * '; ">■ t.i J) IJ TENNENT J'. D 3> ., o. • • • • • « • .1 ^ ir- » . » • • . • • • . .% ' 'r{^X •:..••". ^^'^\^{- •• ; K '•••• . ^ .• B Jj %• TENNENT II. THE PRODUCTION OF LIGHT BY THE FISHES PHOTOBLEPHARON AND ANOMALOPS. By E. NEWTON HARVEY, Princeton Universiiy. 43 THE PRODUCTION OF LIGHT BY THE FISHES PHOTOBLEPHARON AND ANOMALOPS. By E. Newton Harvey. GENERAL ACCOUNT OF THE FISHES. In the Banda Islands, southeast of Amboina, the Moluccas, Dutch East Indies, occur two luminous fish with relatively very large luminous organs. One, Photoblepharon palpehratus, called ikan ( = fish) leweri ( = ?) batu ( = stone) by the natives, is known only from this general region, and was first described by Boddaert (1781) from a specimen obtained at Amboina. It is caught with hand-nets in shallow water, swimming singly or few together, among the stones or corals. The other, Anomalops katoptron, called ikan leweri ajer (= water) or laut (= sea) by the natives, occurs also, although rarely, at Amboina, Menado (North Celebes), Fiji, New Hebrides, and the Paumotus. It was described by Bleeker in 1858. Unhke Photoblepharon, it swims in schools of 100 or more near the surface, in somewhat deeper water. In no other place in the world can these fish be caught so easily or in such numbers as at Banda. During a trip to this region under the auspices of the Department of Marine Biology of the Carnegie Institution of Washington, in the autumn of 1920, I was able to obtain all the material needed for my investigation without encountering any more difficulty than the general indisposition of the fishermen to work beyond their usual amount, or the advent of rough weather. In dealing with the fish- ermen, my friend Sech Ahmed bin Said Baadilla acted as interpreter, and it gives me great pleasure to acknowledge his kindness during my stay at Banda Neira. I am also very greatly indebted to Dr. A. L. J. Sunier, director of the Laboratorium voor het Onderzoek der Zee, at Batavia, whose kind hospitaUty, letters of introduction, and interest in my undertaking made my stay in Netherlands India a pleasure and a success. Through his efforts also I was able to study luminous forms of the Java Sea on the Dutch government's fisheries steamer Brak. I take this opportunity of expressing my thanks to the Dutch officials for their courtesy. These luminous fish of Banda possess more or less commercial value to the fishermen. Both species are caught at night-by the natives and used as bait for other larger fish. The luminous organ is cut out and placed on a hook, the light, which is said to last a night, serving as a lure for other fish. From my own experience in the 45 46 The Production of Light by the laboratory the light from both species lasts between 7 and 8 1 hours. While the fish occur the year around, they are most easily obtained in October-November and April-May, during the change of monsoon, when the weather is calm. They differ in this respect from the lumi- nous squid {Watasenia scintillans) of Japan, which come in toward shore to spawn at only certain seasons of the year and then return to the deeper water. Although the Banda Sea about these islands is very deep (12,000 feet), these luminous fish keep near the surface, living in the relatively shallow water immediately around or between the islands of this volcanic archipelago. They are not so readily caught on moonlight nights, but this is because their hght can not be so easily seen by the fishermen and not because they avoid the light of the moon or show any such periodicity in appearance as that of the palolo or other worms. Nothing is known of the life history of these fish. In October eggs can be squeezed from the females of both species. These are surrounded by jelly, are transparent, and contain large oil droplets, but sink in sea-water. Eggs of Photoblepharon are about 1.7 mm. in diameter, and those of Ariomalops 1 mm. in diameter, without the jelly. Anonialops eggs plus sperm did not develop under laboratory conditions, but these were not very favorable. They showed no luminescence on stimulation mechanically, or upon the addition of ammonia to the sea-water containing them. It is quite natural that the earlier observers, with preserved material at their disposal, should be ignorant of the purpose of the luminous organ. Boddaert thought the function of the organs of Photoblepharon was to shield the eyes of the fish from injury by the branches of the coral among which it lived. Lacepede thought it a protection of some sensitive tissue against the rays of the tropical sun. Giinther first considered that the structures were light-organs, while Vordermann (1900) actually observed the living fish in 1897 and saw its light. The short paper of Vordermann and the descrip- tion recorded in the narrative of the Siboga expedition by Weber (1902) contained all our knowledge of these fishes until the appear- ance of an extensive paper by Steche (1909). This excellent mono- graph deals largely with the histology of the luminous organ and also contains a few physiological observations. Thus, Steche determined that the light shone day and night continuously with an intensity (for one organ of Photoblepharon) of 0.0024 meterkertze. Mechanical pressure or chemical stimulation is quite unable to increase the intensity of the hght, which can be "turned off" completely or not at all only by mechanical contrivances connected with the organ. Discussion of Steche's histological observations will be given in describing the histology of the organ. The Anomalops obtained by me varied in length from 4 cm. to 11 cm. Photoblepharon is about the same size, but the proportions Fishes Photoblepharon and Anomalous. 47 of the two fish are different, Photoblepharon being stouter in pro- portion to the length. Sometimes the former grows to large size and is caught by hook and hne. Mr. Baadilla, of Banda Neira, told me of one about 250 mm. long. The usual size for Anomalops is 102 mm. from head to base of tail, 32 mm. high, and 28 mm. thick at the gills. Its luminous organ will measure 12.3 by 5.7 by 1.4 mm. and weigh about 0.08 gram. In Anomalops the two luminous organs together make up about 0.58 per cent of the body-weight. This is not as large a proportion as we might expect when we consider that the organ looks very large and is actually one-eighth to one-tenth the length of the fish itself. No doubt the eggs which filled the abdomen of the Anomalops examined greatly reduced the weight proportion of luminous organ to whole fish, that would be observed in a fish without eggs. In both fishes the luminous organ is a compact mass of white to cream-colored tissue, flattened oval in shape, lying in a depression just under the eye and in front of the gills. The organ looks as if made for experimentation, as it is attached only at the dorso-anterior end and can be cut out with the greatest ease, giving a piece of prac- tically pure luminous tissue. The back of the organ is covered with a layer of black pigment which serves to keep the light from shining into the tissues of the fish. In both fish there is a mechanism for obscuring the light, but curiously enough the mechanism developed is totally dift'erent in the two species, notwithstanding the fact that in structure the organ is identical in the two and in every detail except proportion the fish are very similar. In Anomalops, the organ is hinged at the antero-dorsal edge and can be turned downward until the light surface comes in contact with a fold of black-pigmented tissue forming a sort of pocket. The light is thus cut off. In Photoblepharon, a fold of black tissue has been developed on the ventral edge of the organ socket, which can be drawn up over the light surface like an eyehd, thus extinguishing the light. Why these two fish, so similar in most respects, and especially in the general structure of the luminous organ, should have developed such totally different means of extinguishing the light is a mystery. As observed by Steche, I find that the organ itself emits light continuously and steadily, in the day as well as at night. No amount of mechanical or electrical (strong interrupted induced shocks) stimulation causes any effect whatever on the intensity. In this respect it differs completely from all other animals and resembles the bacteria and fungi, which also produce a steady light independent of stimulation. One other fish, Monocentris japonica, with a paired light-organ at the tip of the lower jaw, behaves as do Anomalops and Photoblepharon. This peculiarity of steady light, independent of stimulation, should be borne in mind, as it has a very important bearing on the nature of the light of these fishes. 48 The Production oj Light by the In life, the light of Anomalops is constantly being turned on and off, according to Steche, 10 seconds light and 5 seconds dark. Photo- hlepharon in its natural environment shows its light continuously, but in glass jars, either as a result of partial asphyxiation or excite- ment, its light is also intermittent. The animals appear to be swim- ming about with bright flash-lights which are turned on periodically. According to the natives, the use of the organ is as a search-light. Its use, according to Steche, is as a " Scheinwerfer" or search-hght to attract prey, and the flashing is to mislead its prey. When brought into the laboratory, in small glass jars, where the supply of air is limited, the fish lose control over the closing mechan- ism and the flashes become irregular. Just before death from asphyx- iation, when the fish swim upside down and slowly, the light is usually not visible, but on actual death, when movement ceases, the organ is exposed and gives forth light, both in Anomalops and Photoblepharon. This simply means that the muscles involving the closing mechanism pass into the relaxed condition on death. The relaxed condition of these muscles, then, corresponds with the exposed position of the organ. HISTOLOGY OF THE LIGHT-ORGAN. Although this paper deals with the chemistry of light-production in Anomalops and Photoblepharon, an accurate description of the structure of the organ is of considerable interest in view of the rather startling conclusion to which I have come from chemical investiga- tion, namely, that the Hght-organ is a mass of tissue designed for the nourishment of luminous bacteria, which produce the Hght. The gross anatomy and histology of the organ is the same in both fishes, with the exception of the movable screen to obscure the light, pos- sessed only by Photoblepharon. Steche describes the organ as an acinose gland made up of a large number of gland-tubes parallel to each other and extending completely across the organ, from the back pigmented surface to front trans- parent surface, where the light emerges. The ascini have been elongated and arranged parallel to each other. In a parallel section of the fresh organ it is very easy to see these tubes and also the blood- vessels which run between them. A cross-section of the tubes shows that they are pressed into a polygonal shape, separated from one another by sparse connective tissue and arranged in a ring about a blood-vessel, whence they get their oxygen and food-supply. The back ends of these tubes meet a layer of cells containing small gran- ules of guanin, which are believed to act as a reflector layer, in analogy with such a layer found in the luminous organs of other fish. Steche observed that these granules dissolved in formalin and were doubly Fishes Photohlepharon and Anomalops. 49 refractive, but made no chemical tests for guanin. Back of this reflector layer is a layer of cells containing small black pigment granules which effectively screen the tissues of the animal from its own light. In addition, the socket in which the hght-organ lies, the movable screen in Photoblepharon, and the whole of the light-organ except the front, almost flat, surface, are covered with this black pigment layer. Just below the front surface of the organ a number of the tubes unite to form a common reservoir which connects with the exterior by a pore (20 to 30 microns wide) passing through cutis and epidermis. The pores are scattered over the surface of the organ and were over- looked in Steche's first description of the tissue. There is no doubt of these pores, however, as they are very clearly visible in sections of the gland made by Professor Dahlgren, of Princeton University. At the base, the light-tubes are protoplasmic, but the rest of the tube appears to be one large vacuole of secretion made up in fixed material of small droplets and granules uniformly distributed from base to front of tube. The outUnes of the tubes are so diflScult to see at the front that Steche gained the impression that the cell itself is directly converted into the secretion. Although no mitoses were observed, the appearance of cells with large nuclei at the base of the tube was such as to suggest a growing layer there which supplies new gland-tube cells to take the place of those wasting away by formation of the Hght-giving secretion. No luminous material is passed out of the gland, a point which I can confirm from my obser- vations of the living fish, so that Steche beUeves the light to be burned in the reservoir at the front surface of the gland. The light would therefore be extracellular but intraglandular. At the front end of the gland-tubes there is a 1- to 2-layered epi- thelium that passes at the pores into the epithelium of the outer surface of the organ. Here, also, is a tough connective tissue in which are embedded blood-vessels and nerves. The red blood- vessels are very clearly visible in the living organ outhned against the white of the gland. They arise as 9 to 13 vessels passing from both the lower and upper edges of the front surface of the gland and branch to smaller vessels meeting near the middle. A peculiar valve occurs where branch vessels leave the main artery. The blood-supply is exceedingly rich. Steche could not determine the point of ending for the nerves which follow the blood-vessels, but thinks they pass to the gland- tubes and control the secretion. I should judge it more likely that they are vasomotor in function, but I have no evidence for this point. The nerve is a large branch of the trigeminal-facial complex. There was no indication of a marked center in the brain, such as that possessed by some electric fishes to control their electric organs. 50 The Production of Light by the Steche describes the living organ as clear and transparent and bright yellow in color, but becoming cloudy and coagulated in fixing fluids or on death of the fish. I am quite unable to agree with this description. The luminous organ of living fish is white to cream- colored and not transparent. It is fairly firm in consistency, and I have never noted any marked change on death of the fish or removal of the gland from the body of the fish. The fresh gland is firm enough to be cut with a razor into fairly thin sections, in which appear very clearly the parallel gland-tubes with blood-vessels running between them and filled with a row of oval, nucleated red blood- corpuscles. On application of pressure to the cover-glass the gland- tube contents flow downward and mingle with the fluid (sea-water) bathing them, forming a white, milky emulsion. In this emulsion can be seen a great number of small granules and rods, often arranged in spirillum-like rows. The rods are unquestionably bacteria, as they can be seen to move of their own accord, often with a corkscrew- hke motion. Some of the granules are probably end views of bacteria, but others may be cell granules of one kind or another. The granules have a weak Brownian movement. The spirilla-like rows may be almost as long as a red blood-corpuscle is wide. The contents of the luminous cells, then, give an emulsion of bac- teria in the sea-water. In the dark this emulsion glows brightly, but whether the light comes fiom the bacteria or not can not be directly proved. The light is perfectly homogeneous under the microscope, showing no trace of light-giving granules so character- istic of the medusae and pennatulids. Needless to say, the light of luminous bacteria would also appear homogeneous, as the light of a single bacterium is too weak to be seen through the high power of the microscope and the bacteria are too small to be observed as individuals with the low power of the microscope. If the material of the gland-tubes is pressed out into fresh water a coagulation appeai-s to occur and the material assumes a finely granular net structure in which it is difficult to make out individual particles. Brownian and bacterial movements cease and the emul- sion also ceases to luminesce. Bacteriolysis has taken place. Fresh smears of the luminous gland, kindly stained for me by Pro- fessor Dahlgren, show the bacteria very nicely and in great abun- dance. While Steche's description of the structure of the light-organ is essentially correct, failure to examine the fresh organ material has led him, I believe, to a misinterpretation of its nature. The bacteria are so numerous and the chemical behavior of the emulsion of the organ (as we shall see) so similar to an emulsion of luminous bacteria that I feel that they are the real source of the light of these fishes. The organ becomes, then, not a glantl for the pioduction of a secretion Fishes Photoblepharon and Anomalops. 51 whose combustion is intraglandular but extracellular, but an organ for the nourishment of symbiotic luminous bacteria. On this view the extraordinary richness of the blood-supply is at once apparent — to supply the oxygen for respiration of the organ cell, for respiration of the bacteria, and for the production of light by the bacteria. Luminous bacteria require an unusually abundant supply of oxygen, as those who are familiar with the growth of them will readily under- stand. On this view the existence of pores in an organ which does not produce an external secretion becomes explicable — namely, a means of exit of dead bacteria. It is very likely that other fish may also be found to possess luminous organs for the growth of luminous bacteria. To decide this it will be necessary to study the living animals, as no certain remains of bacteria are to be made out in fixed material. A fish which I suspect may prove to be similar to Photo- blepharon and Anomalops is Monocentris japonica of Japan. My suspicion is based on the fact that the light of this fish, two specimens of which I observed at Misaki in 1917, is produced continuously day and night and without change in intensity, just as the Banda fishes. The view that animal light is due to bacteria is not a new one. It has been advocated by Pierantoni (1918) for some years. He had grown cultures of luminous bacteria from the luminous material of squid and has even gone so far as to suggest that in forms where no bacteria can be demonstrated we are dealing with ultramicro- scopic organisms similar to ultramicroscopic pathogenic forms sup- posed to be present in filterable viruses. While it is certainly not true that the light of all forms is due to luminous bacteria, I think the chemical evidence which follows, together with the ocular evi- dence already described, is very strongly in favor of the view that the fight of these fishes is due to symbiotic luminous bacteria. LACK OF OXYGEN. One of the conspicuous anatomical peculiarities of the light-organ is its rich supply of blood-vessels. These can be seen in life, running from the lower (principally) and upper edges of the organ to branch over its surface into capillaries which run between and parallel to the columns of luminous material. The organ is also in a position just anterior to the gills to receive fully oxygenated blood, and in cutting the organ out blood flows freely, showing the presence of large arteries in this region. It is not surprising, then, to find that the organ is very sensitive to lack of oxygen, the light disappearing promptly in its absence. This can be shown for both Anomalops and Photoblepharon by three methods. First, it may be readily observed that the light-organs of fish dying in sea-water from lack of oxygen become dimmed. If the fish 52 The Production of Light by the are now lifted out of the water to the air, the light immediately becomes bright again. This is not due to stimulation of the fish in handhng, because the same amount of stimulation while the organ is under sea-water containing less than the proper amount of oxygen does not increase the intensity of the light. It is interesting to note, in this connection, that strong, interrupted, induced electrical shocks will not cause a luminous organ, dim from insufficiency of oxygen, to glow more brightly. This shows that, for this material at least, the cell is able to absorb all the oxygen available and that electrical stim- ulation can not cause additional oxygen to enter as a result of increased permeability from stimulation. Second, if a luminous organ is removed from a fish and laid face down on a piece of glass, the light, as observed through the glass, disappears almost instantly, except about the edges of the organ in contact with air. On lifting from the glass the organ glows brightly over its whole surface. The glass keeps the air away and the light then disappears with extraordinary rapidity. Finally, if one makes an extract of the organ by grinding in sea- water with quartz sand and places the extract in a test-tube, the light quickly disappears, except at the surface of the extract in contact with air. Shaking the tube causes its contents to glow throughout. In these experiments it is the rapidity with which the oxygen is used up that is astonishing. Similar experiments can be performed with an extract of Cypridma or pennatulids, but oxygen is not con- sumed nearly so rapidly. The situation is exactly as if one had, in sea-water, an emulsion of luminous bacteria, which also use up oxygen with extraordinary activity. The difference between lumi- nous bacteria and extracts of Cypridina lies in this, that the bacteria use oxygen not only for light-production but also for respiration. Large amounts of oxygen are necessary for the latter. The extract of these fish behaves like an emulsion of luminous bacteria and fur- nishes an additional fact in favor of the view that the light is really of symbiotic bacterial origin. DESICCATION. If the light-organs of Anomalops or Photoblepharon are cut out, carefully freed of adherent sea-water, and placed in a desiccator, they dry to a hard, leathery mass which is powdered only with diffi- culty in a mortar. On adding water, there is no light, or at most a faint light, if the tissue has been dried very rapidly by placing it very near lumps of CaCl2. No doubt a vacuum desiccator would give better results, but certainly there would be no fight after drying to compare with the great intensity before drying. In this respect these fish behave as luminous bacteria, which give light on moisten- Fishes Pholohlcpharon and Anomalops. 53 ing only if dried very rapidly and if tested with water within a short time after they have been desiccated. DILUTION. Despite the fact that there is so little light-material in an organ of these fish which can be dried, there is an amount in the fresh glands which will give a visible light when distributed through an immense amount of sea-water. In one experiment with Anomalops, the light-organ was carefully cut out, ground with sand in a mortar, and 10 c. c. sea-water added; 5 c. c. of this briUiant emulsion was then diluted with equal volumes of sea-water successively until the light could no longer be seen. The organ distributed in 1,280 c. c. still gave a good light and in 5,120 c. c. to 10,240 c. c. the light was just visible. The organ measured 12.3 by 5.7 by 1.4 mm. and the organ of the opposite side weighed (fresh) about 0.08 gram. This tissue, therefore, gave light visible in, let us say, about 8 liters of sea-water, or 1 part in 100,000 sea-water. A strong emulsion of the organ is milky in appearance, due to the small, suspended particles. It resembles a suspension of luminous bacteria in sea-water, and like them the luminous material passes ordinary filter-paper readily. As no porcelain filter was obtainable, I can not say whether the particles will pass one of these or not. DURATION OF LUMINESCENCE. A concentrated sea-water extract of the luminous organ of either Anomalops or Photoblepharon allowed to stand will give only a very faint light after 7 hours and no light after 8j hours. More dilute extracts in sea-water give light for a shorter time, as do also extracts in more dilute sea-water (1 part fresh water to 2 parts sea- water and 1 part fresh water to 1 part sea-water). These experiments with diluted sea-water were tried in the hope of getting a medium more like the blood of teleosts, which usually have a salt-content considerably lower than sea-water. If the light of these fishes is of bacterial origin, the bacteria are so dependent on cultural conditions within the Uving organ that they will not live for more than 7 or 8 hours when removed. The organ of Anomalops or Photoblepharon kept intact in sea-water or in the dead fish will also give no Hght if ground in a mortar after a period of 8 hours. These experiments were performed at 29° C. temperature. Once the light has disappeared from an extract of luminous organs on standing, it can not be regenerated in any way. In this respect the extract of these fish differs from those of pennatulids and jelly- fish, which again give light upon addition of fresh water, and from 54 The Production of Light hy the Cypridina and fireflies, which again give Ught on mixing with luci- ferin. Neither Anomalops nor Photoblepharon gives the hiciferin- luciferase reaction. If fresh water is added to a sea-w^ater extract of the organ of these fish, the Ught grows rapidly dimmer and dis- appears. There is no sudden increase in intensity, such as one gets on adding fresh water to extracts of pennatuhds and jelly-fish. All these peculiarities are identical with those of an emulsion of luminous bacteria, and again serve to strengthen the evidence that light is due to symbiotic bacterial organisms. LUCIFERIN AND LUCIFERASE. In Cypridina, Odontosyllis, Photos, and fireflies, the presence of luciferin and luciferase can be demonstrated. In Noctiluca, penna- tulids, jelly-fish, luminous bacteria, and ChcEfopterus it is not possible to demonstrate them; it is also impossible in these fish. Luciferin, if present in the organ, should be prepared by some one of the following methods: (1) Adding boiling water to the excised luminous organs and extracting them in a mortar; (2) heating the concentrated luminous sea-water extract of the organ to (a) boiling or (6) to a temperature (55° C.) which permanently extinguishes the light. Luciferase, if present in the organ, should be prepared by some one of the following methods: (1) Extracting with sea-water and allowing the extract to stand for 8 hours till the light disappears; (2) extracting with fresh w-ater; (3) extracting with sea-water and adding such substances as saponin or sodium glycocholate, which causes the light to disappear quickly. Dark solutions which should contain luciferin and luciferase have been obtained by these various methods, but on mixing no light whatever has appeared. The luci- ferin^ of these fish also gave no light with Cypridina luciferase, nor the luciferase' of these fish with Cypridina luciferin. In the impos- sibility of demonstrating a luciferin-luciferase reaction, despite an apparent abundance and persistence of luminous material, Anomalops and Photohtepharon again agree with luminous bacteria. The luci- ferin-luciferase experiments may be summed up as follows: Photoblepharon luciferase X Photoblepharon luciferin — negative " X Anomalops « — « (( X Cypridina (< — 11 Anomalops X Anomalops " — " " X Photoblepharon u (< (( X Cypridina " — " Cypridina X " — brilhant 30"' p.m June 29, 1918, noon 1 17 N. Equator. 4°23' S. 1 25 S. Equator. 4°36' N. 6 35 N. 1 18 N. 4 06 S. 28.5 27.7 28.3 Easterly. Easterly Easterly. Sept. 8, 1918, noon Sept. 8. 1918. 4''54" p.m. . Sept. 9, 1918, noon 28.9 29.1 29.5 Westerly slight. Westerly slight. Westerly slight. July 17, 1919, noon July 18, 1919, noon July 19, 1919, noon Sept. 18, 1919, noon Sept. 19, 1919. noon Sept. 20, 1919, noon 28.3 28.6 28.8 No current. No current. No current. 4 57 S. 31 N. 6 05 N. 7 09 N. 1 31 N. 4 32 S. 28.8 28.6 29.4 No current. W. slight. W. slight.* Mar. 29, 1920, noon Mar. 30, 1920, noon Mar. 31, 1920. noon July 30, 1920. noon July 31, 1920, noon Aug. 1, 1920, noon 27.2 26.9 29.1 Westerly. Westerly. Southerly. 4 08 S. 1 18 N. 6 44 N 28.4 27.5 28.5 No current. No current. Strong easterly. * At 5 p. m. September 20 current was strong easterly until 9 p. m. It will be seen that on six of these eight voyages we crossed a region of low temperature near the equator, with warmer water both to the north and to the south of it. Moreover, on five of the eight voyages we encountered a surface current setting toward the east in this region, and thus contrary to the prevailing westerly drift of the surface water over the tropical belt of the Pacific. Now, it is well known that a surface current moving toward the west will set up a current toward the east in the deeper waters. This subject has been admirably treated by W. J. Sandstrom in his Hydrodynamics of Canadian o/ the Surface Water of the Atlantic and Pacific. 69 Atlantic Waters, Canadian Fisheries Expedition, under the Direction of Dr. Johan Hjort, 1914-1015. Sandstrom applies the Bjerknes theory of circulation to the waters of the Newfoundland region, which shows that Archimedean forces have more to do with the formation of ocean currents than has the wind. Thus, the Gulf Stream flows from the tropics toward Spitzbergen because the light surface layer of the ocean, which is heated by the sun, is 600 meters deep in the tropics and only 200 meters deep at Spitzbergen; hence the thicker, deeper part of this wedge of light water must constantly flow toward its thin edge at Spitzbergen in a futile attempt to overcome this difference in thickness and reduce the whole wedge to a horizontal surface layer of uniform thickness. Of course, if this were ever accomplished, the current would stop after a few oscillations, but the unequal heating ability of the sun in the tropics and in the Arctic regions keeps the current moving. We can not dwell upon this interesting subject, nor is it necessary so to do, for it could hardly be more clearly and simply explained than by Dr. Sandstrom. Suffice it to say that the prevaihng westerly drift of the hot, light surface waters in the tropics must produce an easterly current in a heavier and colder layer of water which lies beneath the surface. At or near the heat equator, however, we would expect that this deep layer would at times be brought to the surface by the upwelHng known to exist in this region. This would, however, not interfere with its horizontal movement until it reached the surface and moved against the prevaihng wind. As is well known, the rotation of the earth about its axis gives to ocean currents in the northern hemisphere a tendency to swerve toward the right. On the equator, however, there is no such tendency, but it increases as the sine of the latitude, although in a latitude so low as 5° N. it would be practically negUgible. Hence, in this region, if the surface current moves toward the west the deep layer would move toward the east, and there would be but little tendency for it to diverge toward the southeast and south. This tendency to swerve toward the right is of course not wholly lacking even in low latitudes in the northern hemisphere, and thus the great westerly drift of the surface water of the tropical Pacific eventually turns northwestward and finally northward, and flows along the coasts of China and Japan as the "Japan Current." If the easterly surface current which is sometimes met with in about 5° north latitude in the mid-Pacific consists of deep water which has been brought to the surface by upwelling, it might, for a time at least, retain some of the characteristics of the deep water of the ocean, even after it reaches the surface. Thus we may readily account for the sHghtly lower temperature of the water of these easterly currents as compared with that of the 70 Hydrogen-ion Concentration and Electrical Conductivity westerly surface drift to the northward or to the southward. More- over, the Ph of this easterly-moving water ought generally to be lower than that of the westerly drift, for the cold waters of the depths are richer in CO2 than is the warm surface layer. Our observations indicate that this is probably the case; the results upon the eight voyages across the region in the mid-Pacific from 6° N. to 2° S. are shown in table 4. Table 4. Direction of surface current. Observed range in Ph of surface water. Average Ph for all observations. Westerly 8.17 to 8.25 8 . 08 to 8 . 22 8.1 to 8.23 8.21 8.19 8.18 No current Easterly A low Ph when the current is easterly is, however, not always observed. Thus, on August 1, 1920, the ship ran out of the prevaihng westerly drift at about 8 a. m. and was in a strong easterly current until about 4 a. m. August 2, and yet the Ph of this easterly current was exactly the same as that of the westerly drift to the north or the south of it. Naturally, if deep water rich in CO2 comes to the surface, it soon becomes warmed, and thus its excess of carbon dioxide is discharged into the air until the CO2 tension of the water is about in balance with that of the air; and after this process is completed it might, due to momentum, still retain some of its easterly movement, but its Ph would be the same as that of the prevailing westerly surface drift. It seems probable, therefore, that in cases where we find an easterly drift in the mid-Pacific in about 5° N. latitude, and its Ph is about 8.23, it has been for so long a time on the surface that it has attained the temperature and Ph characteristic of the westerly surface drift. If these easterly drifts be derived from deep water which has up- welled to the surface, their average CO2 tension ought to be higher than when the current is moving in a westerly direction, and averaging the calculated CO2 tensions of the water between about 6° N. lati- tude and 2° S. latitude, observed on our eight voyages across this region in the mid-Pacific, we find that this seems to be the case, as appears in table 5. It seems, then, that the CO2 tension of the surf ace water in these easterly counter currents is appreciably higher than in that of the prevailing westerly surface drift. The atmospheric CO2 has a tension of about 0.0003 of an atmosphere, and thus it appears that these easterly currents probably discharge CO2 into the air. In this con- nection it may be of interest to observe that L. R. Cary (1917) found that the oxygen content of these easterly currents was higher than that of the westerly drift. of the Surface Water of the Atlantic and Pacific. 71 Thus the facts seem to warrant the inference that easterly counter currents, such as are found in about 5° N. latitude in the mid-Pacific and possibly also the Guinea Stream of the tropical Atlantic, are composed of water which has upwelled to the surface. Henderson and Cohn (1916, p. 621) conclude from laboratory experiments that upon the whole, in most places and at most seasons, carbon dioxide must be escaping from the sea into the air, although they also state that the balance is doubtless restored by CO2 entering the water from the air in the polar regions. These authors did not consider the effect of photosynthesis by marine plants, which McClendon afterward showed to be such an important factor. Were it not for photosynthesis, it is probable that large quantities of carbon dioxide would escape from the sea in the tropics, but, instead of this, McClendon, Wells, and I find that the warm waters are nearly in balance with the atmosphere. Table 5. Direction toward which surface current was moving. Range in CO2 tension of water in terms of 0.0001 atmosphere. Average CO.; tension in terms of 0.0001 atmosphere. Easterly 3 to 4 3 to 4 . 4 2 . 7 to 3 . 2 3.5 3.3 3.1 Westerly My observations along the Atlantic coast between Nova Scotia and Florida, from December to March, show also that the coastal current during these cold months has a CO2 tension somewhat below that of the atmosphere, and this may be due to the great concentration of plant life in these cold waters. Thus, according to my observations, the averages for the shore current, the sahnity of which ranges from 30 to 33, between Nova Scotia and northern Florida in winter, are: Temperature 6.7° C, salinity 31.7, Ph 8.05, and CO2 tension 2.5 ten-thousandths of an atmosphere, while similar data for the Gulf Stream at the same season between the Straits of Florida and Cape Hatteras are: Temperature 22.3° C, salinity 36.35, Ph 8. 21, and CO2 tension 2.8. Thus the cold shore-water seems to be in a condition to absorb CO2 from the air, w^hile the warm Gulf Stream waters are more nearly in balance with the atmosphere. In summer, when the shore current is warmed to about 22° C, its CO2 tension rises to be quite in balance with the atmosphere. It is well known, from the extensive work of Blackman and Smith, that photosynthesis about doubles in effect for 10° C. rise in temper- ature, but the tropical waters are deficient in nitrogen and can thus support only a meager plant life in comparison with that of colder 72 Hydrogen-ion Concenlralion and Electrical Conductivity regions. Thus, McClendon found less than 0.01 mg. of nitrogen per liter as nitrates and nitrites at Tortugas, while Raben (1910) found more than ten times these amounts in the North Sea; and, as shown by McClendon, the tropical ocean, despite its high temperature, can eliminate only a small part of its free CO2 by photosynthesis, due to the scarcity of pelagic plant hfe. Krogh (1904) calculated that if the average CO2 tension of the ocean is the same as that of the air (about 0.0003 atmosphere), it must contain 27 times as much C'02 as the air. Thus, if the ocean gave off one-tenth of its CO2 to the air, the carbon-dioxide tension of the sea would sink to 0.0002 atmosphere. He found that the CO2 in the air of Disko Island, Greenland, ranged from 0.00025 to 0.007, being high with winds from the north and west and low when the wind blew from the south and east. The turbid sea-water at Disko Island had a CO2 tension of 0.0001 to 0.00035, while the clear water in the same region had a tension of 0.00035 to 0.0006 atmosphere, thus apparently lower than that of the surrounding air. Also, according to Krogh, the CO2 tension of the surface water between Cape Farewell, southern Greenland, and the Shetland Islands was distinctly lower than that of the air. The average CO2 tension of the air over the ocean seems to be about 0.000295, this being the mean of 51 observations made by Thorpe between Brazil and England. In 1917, however, using apparatus given to me by Professor McClendon, I tested the CO2 tension of the air over the Pacific between Samoa and San Francisco, but there was apparently no relation between the local CO2 tension of the air and that of the water under the air, this lack of coordination being due in all probabiUty to the great mobiUty and rapid fluctuation in CO2 tension in the air as compared with that of the water. It would apparently be necessary to obtain several thousand determinations of the CO2 tension of air over the ocean, taken at all seasons and in all weathers, to determine its mean CO2 tension with accuracy, but the determinations that have been made indicate that it is not far different from that of the air over the land. My average for 83 observations made during eight voyages over the tropical Pacific from 23° north to about 10° south latitude is: Temperature 27°, Ph 8.22, CO2 tension 2.99 ten-thousandths of an atmosphere, and salinity 34.87 (range 33.96 to 35.71). Thus the tropical waters appear to have a CO2 tension about in balance with that of the atmosphere. If we consider the subtropical and temperate surface water 400 miles and more ofT the CaHfornia coast and extending from 35° to 23° north latitude near Honolulu, we find for the average of 46 obser- vations that its temperature appears to be 19.8° C, Pii 8.18, the CO2 tension 2.51 ten-thousandths of an atmosphere, and salinity of the Surface Water of the Atlantic and Pacific. 73 34.53 (range 32.94 to 35.64). Thus it seems to be absorbing CO2 from the air. The surface water within 400 miles of the coast of CaUfornia, due to upwelHng of deep water in this region, seems as a whole to be about in balance with the atmosphere, although closer to shore it is doubtless gi\ing out CO2 into the air. Thus, my average for the eight voyages is: Temperature 13.45° C, Ph 8.09, CO2 tension 3 ten-thousandths of an atmosphere, and salinity 33.15. We lack sufficient data for a definite statement as to whether CO2 is on the whole passing from air to sea, or vice versa, but the surmise may seem reasonable that a balance is maintained; the absorption of CO2 from the air by the Polar Seas being offset by its passing slowly out of the ocean over the wide area of the tropics, at least during the warmer months of the year. The temperate regions, on the other hand, stand in an intermediate relation, the water absorbing CO2 during the winter and giving it out to the air during the summer. Table (i. Locality. Ph of sea-water. T^, ,0 (surface 8.34 7.45 8.35 8.27 8.22 8.25 8.08-8.22 8.00-8.05 Black bea< , , 1 deep water Sea of Marmora and Bosporus Eastern Mediterranean Western Mediterranean Coast of Portugal Off Scotland and Faroe Island Southeastern part of North Sea and Skagerak . . It may be of interest to compare our observations with those of Palitzsch (1911), who was the first to apply Sorensen's colorimetric methods to the study of the hyrogen-ion concentration of sea-water. Palitzsch used naphtholphthalein and phenolphthalein as indicators and tested the Ph of the Black Sea, Sea of Marmora, Mediterranean, Atlantic, and North Sea in summer, with the results shown in table 6. 74 Hydrogen-ion Concenlration and Electrical Conductivity CO 05 3 •*-> CO g5 c^ s . 13 < •njnionog in o J J 93 1! W •oo8ton«Jj[ ■Qiaqd -Boun« lOOOO }0 -Bas JO U0ISU3i -B88 JO not? -BJjaaouoD uoi-H -ja)BAi-Bas JO Ej IN t- 00 ■* 0 00 CO rH o CO CD Oi CO t- r-l >-"'*' (N t1< CO «5 CO IN N M N C^ X Oi »>. CO 0> CD to 00 CD CD U) U3 to C3 O O O O O O -H I CO 06 00 00 X 00 •cas JO 'duiax •dvas% iiy ; ^ . ) bC 1 ■»* £3 CJ .^ *^ ^ , CO •* CD ■-< to CO ' (N >0 1^ O '-' CO .-I i-i rt IM C-J (N . CD •fl to 00 i-i CO ' ■-( ,-( i-i .-1 (N IN > ; ^ in ; o o ,r? ea o oqO Q ^ o 00 "O O to IN t>. 01 CO ■* •* 10 40 CD CO O t~ •<*' •-' CO CO CO IN IN e>j a a a a a ^ 00000" o o o o o o c a la a o r-1 -<' M CO Tjl' U3 to (N M CS M IN IN Qi Qi Q) Q) Q> ^ fit A ^IL, & (i, f£i ' lO CO 00 o >o ^ . ^ ^ rt N N CO t>. 10 o to to NNINCOCOiNCOCOiM?> A X 10 10 to >o 10 tDCDO5C0r-tOC0O5CDCO miO»OCOCDCDCOiOiOW5 odddddddod WJlOeO h-OO COIOU5 C>l(NNNi-lr-iMIN(M(N 00060606060606060000 .(NtJIt-iosMOOtIIOO 0> y^J'iOtD'O'^'J'tDCDOOt^ ° C^CSININININC^C^lNIN .r^u3coc^t^u5eo->*i^T)< ~'^lOCDtO'*'<. I3 .•0 13 a t . Mo fc. • P • o . .0000 •iP feP !3 fePPPP 00 C3 to 0000 <-> r r r M o - _ - fe:^^^"PPPQP 1^ 02 02 M r • ■ • • ■ ^^^^^ ; : : : : cor^—ioooMiio-^N ^rHe^'>1ON.t>'0 lOcOCDtOCOCOCOCDCOt^ 2? lOrHco'-io aco-^-iM ?-NC0l0t-lr"T''*"5O—i o I, fl 1^ 13 S a fl B'oa'^a'oflio IrHi-TMINCOCO-*'^ J3 J2 oj the Surface Water of the Atlantic and Pacific. 75 n °9J3l{d -soiniB lOOO'O }0 -899 JO uoisna} »00 p8>«ino[BO -was JO 8tnBj3 000' I nt 8i[B8 JO SUIBJS 'jijIUIJTIg 'J3^«Ai-6aU JO 8tnBj8 000' I ui eiuBiS 'atliJOiqQ -Baa JO uoti -BJ^uaouoa uoi-{j ■it^iniiBS joj pajoajjoo Hj anjx ■duia:) Ja'jTjjVi •duiaj Jiy I " t: a 5 & o a « ■?-^ Q & a> ^ o> 00 t~ r>-t^X«DOcOCOiOCD oddddoddoo C-JM OO-HiNNOlM oococ<5^ o o o o o "S" o " as caaafliot-a>o^ o> o" "-^ c^" ro fo ■* ># 1? ■* t» !0 O "-■ OS c to m o> I-. o> r- c« o 00 -< ■♦ •* lO lO ffO CO CO CO CO CO CO CO 1*1 to CO N I I lO •* t>. CO O 0> 0> 0> 00 00 X to ■<»'■<«< to CO rH to 05 C5 to to to to t- r^ d o d o ^ o rt (M CVI rt 1-t -< 00 00 00 00 00 CO (^ 0> to to 00 to CO ° CO CO ■-( o to •* e-) C^ (N fl .-1 r-1 . ^ to o> -^ to ^ -* CI IN fH Ol to to (M (N W .-I r-< -H ^^ fe fc S b I R 5 g 5 i rl • § ^ I -^ H . CO Z Z Z M CC O M CQ O ■ • - 0. : • a « O Oj ll> ki ^ ^" »_• ^" I* "> ■ w 3 -ti feXJ- d ■* I O 5" ' »„ CO I O "5 ■ H — O o CO I «3 CS I If o IS ^ A a a a d a CO o o o o o i o o o o o to a a a a a >o to h^od o> o c^ N (N e^ N CO 76 Hydrogen-ion Concentration and Electrical Conductivity e 82° .2C35 §00 « S ■^ i « eg TOOOOPSuijat u I u o I B « a ^ =00 p3;«inoi«0 000' I ni eii«B JO BtUt^jS 'XitUIJBg •j3}0M-eas }o etuBja 000' I "? euiBiS 'auuoiqQ sec e • &QCQ Si.* ^ s -CdB JO UOIl -■BJlUBDUOa UOI-H aoj pa^oaiioo hj •diaai ja^B^ ■diua-} Jiy ■o -a M *3 t3 Its £ c C 3 ^ <» Q a ^ S.2 o g * .h ja 3 K « •<* 1^ 00 OS 00 (M O) IM (N IN CO « O t» N CO >0 •* CO 0» CO CO CO Tji U) r)< CO CO CO CO CO •o •* 00 CO lO O "5 CO CO CO 00 00 05 05 O <3> in >o lO 't> ■^ t- t~ fC -* to to <£• ii o V- O t, 4) t, ."" I) r- (U " - m ? "m S fe " C « gQQQQ ^ i s S 2 > m > ■« CO to O CO t- a a ^ O CJ (^ . S ° -"S B, "3 a - .. ■3 ,2 (S o"' rt rH ,H W O 00 •<1< t' ?» 2 CO ic >-i o Tj" a ■ 00 ■* '-' t~ IN .2 ' CJ CO •* Tt" lO p ■^ o o ■* CO i-l 00 •* CO CO CO PJ M r-( M IN C< N St of the Surface Water of the Atlantic and Pacific. 77 e o s 8§ §'^ S5 1^ &&5 S b c . •ojaqdsounB lO lO >o TOOO'O P sniw> OS O ■* "5 i-H M to t- ^ ai uotsna:^ M COMCOCO."OCOe<3C<3MCO "OO pa^Binoi^O •jajB.tt !uin;g •ja-jB.tt-Bas JO » CO CO CO ^ CO N IN IN rf N • CO h- to to sniBja 000' I ni OJ o> e» o> o> euiujS 'anuopi3 | i-l rH "^ '^ "^ i i ■ja^BA X -sas JO not^ to X -■Bijuaauoo uoi-jj o to ooocDcoO'-icoaios tOtOCDtOtOCOtOtOtOtO to to to o oooooooooo o o o •X:jtnijBS 25 NNINOOOOC^i-loOtOtO NININrt>->INMrt>HpH N CO (N CO IN joj pa-joajioo hj 00 00000000X0000000000 00 00 00 . N tOC0tOC0-o ^ a «3 3 XI 3 2:^ 2 '^ fe ^ ja ^ i> 73 m _2 c ^ c C c C c c (N >, 1 o lU " 0, 3 & 1 IV c "Z, ^^^« M is O is >. d H |g "3 3 a a '13 -<' n S fe 6 a r 'it m oj c i r a a o O a c 11 Sap s -cQp §q:3 !- j; 1- 11 %%\\% % < 3 3 m Om -OO ■ -O ■« at oward d was and nd by scale I CO CO CO (N N (N IN C^ 0) 0) Qji-Ht-H^Hir ■ > g o c c c 3 3 H" z ;t^!5coioooZ'>j<'o2 ; iz ;2; k OOtOtOOJC^Ot^OOl^ c ) OStOCOOOt^iNt^TtitOC 1 c a ^ lOOOCTiOSOINlNCOCOtt 5 C ) ^" ^ "S " 2 <£ 2. > (N00-*-Hr-lO»U5Or- < ■< < t 3 CiO g MiOINCOOO^INCOr- rHCS|COCO■*lOK5lO'0^ o 1 > 2 2 I-) CO to to to to to to to to t£ 3 t£ > > -5 M ^ ** 1 to "S ^ •^iNomo'-ii^M'taa 3 3 ^ i< 3 o - 1' hi hJ oco'<<<'. INt»BAV-T!38 JO eoinjS OOO'T ni anuoiqo JO btugjo -B39 JO noi') -Bj^uaouoD noi-jj •A^inij'BS joj pa^DDjjoo Hj anjx •duia} J3)BAi rf — 1 U5 (N O i^ r^ to CO ■-1 O O Ci C5 05 00 00 00 00 r-. 05 CO W CO Tfl -1 CO to (O O IN OS --I —I —I O lO CO CO CO CD o o o o o TO w ^ ^ cq IM CI IN M IN 00 00 00 00 00 CS IN IN IN 00 00 00 00 00 CO 05 .-< lo t>. 00 00 05 05 C^l M IN IN IN t>->OCOCO "^ S-?^ 6 s •^ (£ ^ a .t: s .a ^ 'BAiig luojj sajij^ ■opnjia -no| Djwuixojddy 05 t>- CO b- 05 I t^ t^ 00 05 00 IN IN IN IN N ooooor^oocot^t^co COININININC^ININC-) t3-a . Mo .S.2 7 wwwO'T'^mdO'T-'OOOOOO - - -■=Ov^--m ^ iiQ-SPQ 2; -s (3 m C3 J O -3 C3 -5 o O O O jp-S g-SPPPP a • a » CI • • • ■ r • 3 > 3 ; P S 5 S QP:gpp ■ hj PPQ^ § . . . un o '. '. I Tjl Z ass d d oi lOOUtl o • •'O oo Tfico-^ o aa^ o>o CO>Ot-ZcOcO00'^t- CO lO t^ o •* ■^ CO CO o) CO TS< 00 IN CO CD CO'OCOCOOCOCO'IIO 0)iNr^M- •-1 .-(1-1 c^ IN IN C< CI IN :4^ ■^ -S t^ 1^ o 05 1^ \3 g t^ t^ t^ CD CD ^ _^ ^ ^ ^ ^ ^ CDMiOiOcOOOi/S o ococoifj-^coiNcohM a CDCD1. 00 00 0> 'CC-i'^ocoLooco 3 ■<}-ic-)oo a, "5C0050.-ICOCOCOO. OSOiOOOO-H-HM 0> -H >-l t^ CO N C) CO CO CJ IN CO 00 00 ■* 00 00 o CS cq CO CO CO CO >0 'O -"^ CO CO IN CO CO CO CO CO CO CO CO 00 ■<)<■* 00 U3 U5 rH CO CO oo 05 05 Ol 00 00 t^ •-I CD o> h- r» o> _ _J t^ 1-. I_ CD CO CO CO CD oo o<6 O <6 ^ o OP(i4 jtt to -^ 'i -_ y M » S'i : z 2; so Ifi J2 '-I — , ^ _ o o o o o gppppp C) CO O ■* 00 cs .2 CO CO lO »0 O r> O IN O >-< C-l ^ >0 o 'O C5 ' oi CO ■>»' "^ o e^ N N ei IN N CO 1^ M 00 CO 00 00 00 lO (N t-- ro ■^ O) X 1(5 "5 "O to O to t>. t^ r- to ■-< rt T-H CD to to to to to to o d d d o o o t^ r^ t>. 00 — 1 —I ^ ,-( i-H -H .-I (N (N (N 00 00 00 00 00 00 00 , t^ lO lO rH 0> •-< ' '^ to o cq CO K5 to il r-l rt (N IN (N 0 O) 0> O •-I -H 00 Nc^icococ-icococococ^ b-iococo'-ir^coco'-ito rJ(tO>-l'-lCOtOOOOOOlO cocococococococococo O050ooooooi00soa> tototototototototoo ddooddtDOod CqCJ(NOJ"00->l< 'irjtoododoooooooooooo C^CqOllNC^lNiNlMC^M .C000NIO-* i-(T)<-ooo5o6o>oo ' (NCSCvl(NlMINCqC^ . :3 cd O CO O ■ M :-5;3' o o ;pp, B 6 d Q a d a d .„". Q.". ooo„ iqc30D iz; to 00 !2; ft !2; . s o t^ o ^ a> j^ ■«*l •* r» CO ^ "5 o iotor-t^h-ooooooo50 80 Hydrogen-ion Concentration and Electrical Conductivity 'fi J* « k. e « CQ •w s o e 3^ 5 ~i ^ 3> w <~i en ^ C^ o e^ ^' fJ- a. *^ o OjJS (^ HI K n. a) o^ "* ^ t s 1 ? S ;^> •J^ e s o ^ ^ > e t:i ■i&: e e: »-% »• •^ o c o ^ 1 s c li- » :: S 02 •c* 'ec<5 cc "o" ?> *> 3 e f? s 1^ la -sounB lOOOO P euija^ ui uoisua^ '00 pa^Binoi^O 000' I ni r(i«8 JO SUISJ3 'X^imiBg •j8}BM-Bae JO suiBia 000' I n? autJOiqo JO suibjo •jajBAi -Bas JO noi:} -Bj^naonoo uoi-jj -ES loj pajoajjoo ja^BM-BSS JO HJ •duia% jajBj^ ■d\ua% Jty •^ t^ o -*-> c p- O -O to S it COCOCOCOCOCONCO ooO'-'iom'^o OiOOiOiOiOOOlCi X to (D to to C^ C^ aio>oit~-r»co I ;:3 O O O 4J O ; <» Q Q Q "= Q ^Z^ fe-' J* £ O sJ -a § ° S g o- d fi g oQ OS oQP a o •OO-^NlN-^iOCO Q « '^'Z O <0 rH (N t^ h.OOOSOO'^'-iC^ 05 Oi C5 t* '^ (N (N . to -H (^ n lO nt ■* CO (N »- -g ,1 rH r-I rH i-( O g OJ t^ T»< -Of CO -2 O T)< o (N — ■<»< a ^ CO 1^ O CO >o O (N IN CO CO CO lO I u) to r> 00 oi I N N O O t^ h- CD 10 CO CD U5 »0 000000 CO M IN ■!)< •<)« N Cvi N CI O) -l ■* OS N CO ^ i^ 05 '^ ^ rt rt ,-c « CV| C^ . >0 IN 00 »» t>. « j-' CO CO "5 l> d CO 'H « rt .^ Ji oOQQQQ ^ : : : : : 1 2; 1 •OOOOO^.OSCD 5 C C coco— ICOUJIN ge a ^iNOOiiNcoo ^ Tjl 11 M U3 rt CO CO CO e^ IN M CO CO t~. (N M N W ci CO CO CO ^ ^ ^ ^ 10 ^ CO CO CO CO CO CO 00 r» 00 05 00 I-- N 05 "-I 01 »0 OJ 05 00 OJ OJ 05 OC CO CO CO M CO t^ 10 »0 O CO CO U5 000000 IN M W OJ IN IN 06 00 00 00 06 06 Tjl 01 OJ >-l "5 "5 U) t^ CO 05 OJ OJ O) iN OJ IN (N . f >>. OJ ■<)< IN 00 r "5 CO r» t^ OJ o Ol OJ O) IN OJ CO ^'=3Q^^' ■2 aj Qj S< 'i^ C3 W -S ?> 4^ 03 (n ^ ^ n 03 0/ ^ 0) d) 0) OJ ooGoCo I « ^ ^ ^ fe (3 Z M 2; s S g ^ 9 o' o' o" § d.§QQQ :z:ojZ : : : 5 ,1, •* 10 lO •* ^ OJ a t^ CO lO OJ ^ *-* OJ OJ CO k aj -^ - 01 t^ 00 CO "I t". •* UO -H rH ,^ OJ -^ 00 — I CO UO t^ 05 •O CO CO CD CO ^ 00 OJ t- i-H -.J" o r» 00 05 o •-" 1-1 OJ N IN CO CO CO ■^ IN M 00 00 ^^"'. CO IN Oi CO Oi Oi a-. •-I t-- U5 U3 to "O dodo -■88 aoj pa'jDOj -joo aa:>'BM jo h J3')0AV •dui3^ jyV Is . r^ to 6^ M *^ •- .H o ^ ^ ■ S S ^ ■^ T3 3 ^ I ^ •noi) -BAJOsqo JO auitx te £ S Q g .S « « " S o ^ t. a. -■ 1-H (N 00 t^ "5 CO ^ 00 00 t ■^ O '-' ■^ O! ^ rl CD 0> O —1 1- N CO r^ t- t» «o to c< ci ei N (N ci e^ lO Tji CVI 00 00 CO CD r» "-I lo to to T)> Tjl lO Ttl CO M CO CO CO CO CO CO 00 CO ■* CO •* to rH C^ TJH O to O d 05 <35 05 00 00 . 05 —1 O CO -< ID M _ _, rt o I lO kO to to to to o o o d o o Tt< CO •-< M —I N M W (N M (N od 00 00 00 00 00 ' lO T(< CO IN O to IN o to tt CO to OS f-i ^ to M M C^ IN CO CO I 1 CO (N ri M C< « « I of the Surface Water of the Atlantic and Pacific. 83 •ajaqdsoui'jB lO lO m •4 lOOO'O JO euii8» m OS CO lO e<5 a> CO 00 00 CO CO lO -• Ol 1< 1^ ^^ -«3S JO 8Uiej8 Tfi CO ■* " CO T}< CO -H O ■* OS 00 o> •O t^ t» t^ 000' I «! s?IBs JO « C^ CO CO CO CO to CO to CO to CO CO CO CO CO CO CO CO -1 o CO CO CO CO O Ui BoibjS 'Aiini[Bg •jajB.u-Bas JO O) CO FH CO CO lO C3) ifs ^ Ttl ^ siu«j3 000' T UI CO OJ lO o '-' rH rt 0> r^ ■<« o> CD r-< •* >-< o t^ t^ o. 00 o . 05 t~ t^ t- 00 1-t N. o> •ja:)BAi 00 -B39 JO non X «5 W) •* IN CO -BJ^USDUOO UOI-JJ o; oj CO lO CO I> co' o» •* ■^ So c 5 M . £ fe S a — 60 o o 1 03 6 ?:i=^ 3 9 3 0) M 3 S3 ° ?3 (0 1 §• £ ' ^^ 3=3 (- si p Q C p « p c o P a P P « Q P i « d ;; ^ ; ji c o O o O o £c o o o OS a § c p P p C c P b oC p p p Si oP p ^ > HI a ■ o o o o Direction toward which wind was blowing. 5 c d p 5 J3 c c i 0] 03 a a2 o ■ Q :3 £ £ oj ■a R 6 P c p .a bU "3 " ^ « tux, 1^ <-i 1-; ;3 u O u ^ I? m z O ^d a a a a a 6 a a : a a a a VH.° a a 03 a 0. ca 0. 08 OS CS a Id Time o servat s ^ 1 CO ■«1< o o "A o p D S O CO OO O o o o o S a o 2 00 o o CO o S o <« -a u of 03 s w a 0, a o M O - S a O > a a a O a> a 6 i ;^ 2 1 i 03 O o a =a il g Ji o "o O w c 2 S o; a '3 a o « 1 §•1 a » 3 B S o 03 t. M ^6< II 1 & C9 1 1 1 U "o 03 o •n a 3 O 03 _o "S CO to a ■§ B o >a "5 o Is 6 8 ^ •§ ■s^-3 !0 .(a a o •« to o a' a^ in 1 o I a 1 o w o CO o t>. '"' rH IN ■* C^ r-l (N CN| -c ^ ^ to t^ (N ^_j CO CD CO O lO 00 00 IN M ■* CO rH IN "O ''J' CO N e CO ic >o 05 o T^H M o f.^ (Tl 00 _. h-) 00 00 00 t^ K t^ t^ t» t* t^ » 2: ?- N l-H C'l d CO 00 (N >o Tt< Ui t» o CJ CO IN M r^ rH M 1 « 2 S. >o * •* U5 o N CO rH 1 •-! CO (N o o ^ ^ lO ,__ ■<»< lO IC CO CO I^ ■<« IN M IN C^ IN Q i-l to p 2 a OS C35 1 84 Hydrogen-ion Concentration and Electrical Conductivity. TOOO'O JO sviid% Ut U O I S U 3 ) 500 ps^BinoiBO o to IN IN 00 -H CO ■jaye.'A. -■C3S JO SIUBJS 000' I ais^lBsjo f-l N IN 1-1 •J3^B.tt-BaS JO suibjS 000' I uj o o> "5 ■«»< 00 t>. t^ r- •J3^BM -B3B JO a O I ^ -Bj'}n3ouoo not-^ O O -I '^ « r^ d d JOJ p3)03JJ0a J3^BM-B3B JO HJ rH W r- O •diuat JS-jB^ o o 00 00 IN U5 ■«1< (N iH i-> 00 00 •duiaj 'jiv U5 1> CO d 3 IP Q Q P 3 M '% P Q O -0 T •.- ■~ u. f ■fi u ^ O p is . oj o ~ -S S .H P S M S P fe T! 3 3 o -ri tt ^ t a 0) 3 S "3 ? ~ ■3 3 M P P P p pp p p o o pp O P P w w S5 B S s e d B. P P E S S o ffi o O E'aiP ^E Sii HE « C3 J3 n H H '^ o » o (u * E.=2 ^.2 ^ .5 'q "C o o (^ Wtetn (a o ea o 03 o = ca 'e j= 45 ■ 5 fcrl ez ''z i;io 03 ^ ^ a ^ a i - .2 & 3 3 ■a; n P_d ^ -3 1 Z P d^- 13 OS •3« Sf **- j3 0! -u a ba i. H J ,,--^ .-c »-• ri^ fi- E 1-1 ^3" J w Si ,• . _ in .^ 10 o •-1 t^ to « It! US (N IN ^ *" to 00 - ^ lO U5 « ,-, rt IN TO n ^ ^ ■<>< ■"J" U5 IN IN C) -H IN 01 IN e./3 - 2S.3°- 3*. 5e-June 27 O- a.lS- 2a.^s"- 3*.79-june 27 . a.23-2« 29/-3S*.*-Sfpt.t O e.2-2B.i „ 5 - June ze B.9-3S.5-5ept.a** Phoenix 7 -3S.2e-june2e ^Oa 18-283 -3i.7l-Junt.23 8.23.-28.3 Ellice , It.l3nds Mari^i esa Islands •• '"-'3.23-26.3-35.77- 8.25-27-3S.6 -S\ept.6»-r '. Samoa Island -;<53s. *-a8 2i-27.'s-3S.1l-Aug.l/ >-S e'25-l6.e°-3S.-H-Aug.l2 8.25- 26.^ -Aug. I'Z. S.'i5-ZS.*°-3SJ7-5ept.6 ■ •'Tonea O VOYAGE OF S.S. SONOMA. JUNE 18-30,1918. « S.S.TE.LUNE .AUGUST 11-12. • S.S.NIAGARA , SEPTEMBER 6-18,1513. Chart 3. — Voyages in Pacific Ocean in 1918. IV. CARBON-DIOXIDE CONTENT OF SEA-WATER AT TORTUGAS. By ROGER C. WELLS. One figure. 87 CARBON-DIOXIDE CONTENT OF SEA-WATER AT TORTUGAS." By Roger C. Wells. INTRODUCTION AND SUMMARY OF RESULTS. It is generally considered that the carbon-dioxide content of sea- water may be increased by accessions from the air, by animal life, by the decay of organic matter in the sediments on the bottom or elsewhere, by the solution of carbonate rocks, by the contributions of rivers, and by gas vents beneath the sea. Although all rivers carry carbon dioxide into the sea, most of them actually dilute the sea-water, and should, therefore, as a matter of fact, not be classed as increasing the concentration of the carbon dioxide in sea-water. On the other hand, sea-water may lose carbon dioxide to the air, to plants, and in the formation of carbonate rocks and the carbonaceous parts of organisms. Mere evaporation and precipitation also alter the carbon- dioxide concentration somewhat if other conditions remain un- changed. It is obvious that the actual condition of the water at any given time and place depends on a complex set of factors whose evaluation requires many observations as well as a knowledge of the previous history of the water, including information concerning the currents, flora, fauna, and other agencies that may affect it. With the hope of gaining information on some of these points, the writer has made determinations on sea- water from various localities, which supplement to some extent the work of Schlosing (1880), Dittmar (1884), Murray (1889), Fox (1909), McClendon (1918), and Mayor (1919). Determinations were made at Tortugas, Florida, in June 1919,' on water taken directly from the sea at various points about Loggerhead Key, which reveal unmistakable diurnal variations, first noted by McClendon in 1916.'^ The water has sufficient contact with plants and sea-weeds to show the effect of photosynthesis on its CO2 content. There is a loss of CO2 by day and a gain by night. The respiration of the animals is not sufficient to keep the equilibrium steady during the day. It is impossible to say to what extent the balance is maintained by the atmosphere, but in this locality, at least, where the water is ' Published by permission of the Director, United States Geological Survey. » Carnegie Inst. Wash. Year Book for 1919, p. 195. • Carnegie Inst. Wash. Publication No. 252, p. 216 (1918). 90 Carbon-Dioxide Content of Sea~\Vater at Tortugas. relatively shallow, plant life appears to be the chief agency in causing a daily variation in the CO2 content, and also a slight deficit in the average content from the theoretical. Even here, however, the departure from the equilibrium conditions required by the equation of Fox, which is based entirely on the physical factors of temperature and salinity, is not great. It is to be inferred that the slight deficit in CO2 is accompanied by an excess of oxygen in the w^ater, increased activity in the animal-vegetable cycle, and increased precipitation of calcium carbonate. The diurnal changes caused by photosynthesis are in the same direction as those caused by the diurnal temperature changes, assum- ing complete adjustment of the equilibrium between the carbon diox- ide in the water and that in normal air. This equilibrium can scarcely reach a full adjustment each day, however, even in the relatively shallow water of the lagoon, but if it did it would not account for all of the observed variation in carbon-dioxide content. Theoretically, all surface sea-water should show slight diurnal changes in its gas content, but the mixing of the water and experimental errors have heretofore prevented the demonstration of such changes. I have attempted to discover whether Dole's figures^ on total carbon dioxide at Tortugas reveal diurnal changes, but unfortunately the data were gathered solely with reference to the tides and do not give the time of day exactly, so that it is impossible to use them for this purpose. As far as can be determined, there is no general cor- relation with the tidal flow, but if the effects of photosynthesis were considered, it is possible that some relationship to the tides could be determined. However, as photosynthetic action extends to considerable depths, it appears doubtful whether unaltered water gains access to Tortugas by tidal action, the keys being situated on the western part of the Florida platform. There is thus no opportunity for an up-weUing of deep vv^ater rich in CO2, as noted by Mayor in the open water of the Pacific. The source of the water in the flowing tides should be determined, if possible, at the time further samples are taken to test these points. Wood-Jones has shown how markedly the whole developmient of the Cocos-Keeling Islands is affected by the prevaihng winds and currents of the open Indian Ocean. The average diurnal variation in CO2 found around Loggerhead Key was about 4.3 per cent of the total CO2. There was greater variation than this on some days, as much as 5.6 per cent. Moreover, the absolute quantities of COa found varied from day to day, the mean being 0.0895 gram per liter at 27.2° C. On one rainy, cloudy day the CO2 found averaged 0.091 gram per hter, whereas on a fine day it averaged 0.0885. This is as would be expected with regard to the photosynthetic effect of sunlight, but the data are still too few » Carnegie Inst. Wash. Publication No. 182, p. 73 (1914). Carbon-Dioxide Content of Sea-Water at Tortwjas. 91 to establiyh the relationship definitely. Even wider differences are shown in Dole's figures on different days. No diurnal variations were found in the excess base, and it is doubtful whether any will be, as solu- tion and precipitation of carbonates from sea-water are not thought to be an action that reaches equilibrium rapidly. Even the final CO2 distribution attained between sea-water and air appears to be reached slowly, according to the experimental evidence at present available. Determinations of CO2 should probably be made soon after the time the samples are collected, on account of the possibility of the decay of organic matter, such as algse, in preserved samples. More- over, the writer' has shown that sea-water preserved for a long time in glass loses part of its CO2, owing to the precipitation of aragonite brought about by the slow solution of alkaU from the glass. This was ordinary glass. It is well known that '' nonsol" glass is attacked more slowly. The other change found accompanying the decrease in CO2 was a decrease in the titration alkalinity of the water, corresponding to the decreased quantities of carbonate and bicarbonate, but the Ph value was not greatly changed, as sodium had merely taken the place of calcium. The average ''excess base" found at Tortugas corresponds to a normahty of 0.00239. This titration includes everything that con- sumes acid; it represents chiefly bicarbonate, about 0.00183, some carbonate, about 0.00041, and other substances that contribute to the alkahnity, about 0.00015. The last figure, however, was not deter- mined at Tortugas, but with Gulf Stream water that had been shipped to Washington. This water was collected November 30, 1919, tem- perature 17.94° C, lat. 24° 24' 20" north, long. 81° 31' 15" west, about 7.5 miles from American Shoal Light, depth 105 meters. Its titration alkalinity, or excess base, was 0.00258, total CO2 0.0968 gram per liter, Ph 8.08 at 25° C. These figures give on calculation bicarbonate alkalinity 0.00197, carbonate alkalinity 0.00046, other alkahnity 0.00015. The observations here reported are too few, however, to warrant discussion. The methods used in arriving at these figures are given below. METHODS. Carbon dioxide. — The .total CO2 was determined by adding an excess of hydrochloric acid to 500 c.c. portions of the water and boiling about 15 minutes, while a current of CO2 free air was passed through the water, then over calcium chloride, and finally through two soda- lime tubes, each having its last third part filled with calcium chloride. The soda-lime tubes were weighed with a counterpoise to minimize errors likely to be caused by the high humidity. The second tube served as a check on the absorbing power of the first. » Jour. Wash. Acad. Sci., 10, 249 (1920). 92 Carbon-Dioxide Content of Sea-Water at Tortugas. Chloride. — The chloride was titrated with silver nitrate, using potassium chromate as indicator, and the density of the water was calculated with the aid of Knudsen's tables. u -, 0.094 .092 .090 .088 .086 Cloudy Ram N.wind Cloudy Cloudy Rough Fair Fair Ram Fig. 1. — Total carbon dioxide and time of day. H-ion concentration. — The Ph values were estimated colorimetri- cally by comparison with a set of standard tubes very kindly loaned by Professor J. F. McClendon. Excess base. — The excess base was obtained by titrating 100 c. c. portions with 0.02 normal sulphuric acid, using methyl red as indi- cator and blowing out the carbon dioxide by means of pure air for Table 1. — Determinations on Sea-Water at Loggerhead Key, Tortugas, June 1919. [t, temperature, °C.; CI, chlorine, grams per kilogram; D, density; Ph, hydrogen-ion concentra- tion expressed as — log [H + ]; Alk., excess base, in terms of a normal solution, or alkalinity titrated with acid, using methyl red and blowing out CO2; CO2, total, gram per liter.] Date, time, and condition of water. June 10, 11, 11. 12. 13, 13, 13, 13, 14, 14, 14, 15, 15, 16, 16, 16, 16, 16, 17, 17, 17, 17, 18, 18 18, 3^3CP p.m.; cloudy. . 12 noon; fair 5''30"'p. m.; cloudy. . 12 noon; after rain. . . GMS" a.m. ; after rain 9 50 a.m.; cloudy. . , 12 50 p.m.; cloudy. . 4 10 p.m 9 45 a.m.; after rain and wind 3M5™ p.m.; wind. . . . 5 15 p.m.; wind. . . . 6 30 a.m.; wind. . . . 11 a.m 7''10™ a.m.; fair 8 15 a.m.; fair 11 a.m.; fair 4''15™ p.m.; fair 5 45 p.m.; fair 6 30 a.m 10 15 a.m.; slightly cloudy 4*^45™ p.m.; rainy. . . 7 p.m.; after rain. . . . 7'' 15™ a.m.; after shower 11 a.m.; rain 1 p.m. ; rain 26.6 27.6 28.2 27.4 27.1 27.1 27.5 27.6 27.0 27.4 27.8 26.5 26.8 26.8 27.0 27.0 28.1 27.8 27.1 27.3 26.8 27.6 26 8 26. S 26.6 CI 20.01 20.00 20.05 19.90 19.28 19.90 20.07 19.96 19.89 19.87 20.00 19.94 19.31 19.99 20.00 19.99 20.05 19.98 19.99 20.09 19.16 19.77 20.04 19.81 19.73 D 02373 02339 02326 02332 1 . 02258 1 . 02342 1.02352 1.02334 1.02344 1 . 02328 1.02333 1.02366 1.02371 1 . 02368 1.02359 1.02357 1.02330 1 . 02330 1 . 02354 1.02361 1.02239 1 . 02308 1.02370 1.02339 1 . 02334 Ph 8.23 8.19 8.19 8.21 8.20 8.19 8.22 8.21 8.21 8.21 8.07 8.19 8.20 8.23 8.21 8.20 8.18 8.20 8.19 8.18 Alk. 0.002370 . 002368 .002398 . 002394 .002394 .002410 . 002402 .002406 .002402 .002412 . 002435 .002424 .022368 .002414 . 002440 . 002374 .002410 .002414 .002396 .002306 .002326 .002418 . 002400 .002374 CO2 . 0890 .0925 .0902 .0924 .0913 .0900 .0897 .0881 .0863 .0873 .0910 .0869 .0904 .0896 .0888 .0873 .0873 .0928 .0910 .08.57 .0893 .0918 .0928 .0879 Carhon-Dioxide Content of Sen-Water at Tortugai^. 93 15 to 20 minutes. This method gives sharp results, but the alka- linity thus found probably includes a small amount due to substances other than carbonates. For the alkalinity due to carbonates alone the method at present used by the writer is as follows: A slight excess of 0.02 normal sul- phuric acid is added from a burette to 100 c. c. of sea-water in a non- sol flask. The mixture is then boiled 15 to 20 minutes, while a gentle stream of pure air is passed into it to assist in blowing out the carbon dioxide. (This step might be done under reduced pressure.) The solution is then titrated back with 0.02 normal sodium hydroxide, using cresol red (o-cresolsulphonephthalein) as indicator. This indi- cator is not entirely satisfactory, but is used in order to arrive at an end point of approximately the same Ph value as sea-water. If this end point is not used the titration will include in part the acidifi- cation of certain substances besides carbonates that are present in sea-water. The water is first more than acidified in order to assure the removal of all carbon dioxide. The record of determinations made at Tortugas is given in table 1. Figure 1 shows the relation between the carbon-dioxide content of the water and time of day. GENERAL REFERENCES. DiTTiiAR W. 1884. Challenger Rept., Physics and Chemistry, vol. 1, p. 215. Fox, C. J. J. 1909. Faraday Soc. Trans., vol. 5, p. 82 McClendon, J. F. 1918. Carnegie Inst. Wash. Pub. No. 252, pp. 213-258 Mayor, A. G. 1919. Proc. Amer. Philos. Soc, vol. 58, p. 150 Murray and Hjort. 1912. The depths of the ocean. ScHLOsiNG, T. 1880. Compt. rend. 90, p. 1410 Wells, R. C. 1918. U. S. Geol. Surv. Prof. Paper 120-A. WooD-JoNEs, F. 1910. Coral and atolls. V. ANALYTICAL SEARCH FOR METALS IN TORTUGAS MARINE ORGANISMS. By ALEXANDER H. PHILLIPS, Princeton University. 95 ANALYTICAL SEARCH FOR METALS IN TORTUGAS MARINE ORGANISMS. By Alexander H. Phillips. This is the second instalment of analyses of marine organisms collected at the Tortugas. The first was printed in the report of the Carnegie Institution of Washington for 1917, where also the general method of analyses was described. Owing to the smaller amounts of both copper and zinc in these specimens, the method was modified and the copper was determined by the ferrocyanide colorimetric method, and zinc was determined by the terbidimetric method described by Victor Berckner/ n .^ / Specimen. No. 614-0.... 614-S 614-T 614-U.... G14-A-B.. 614-A-R.. 614-Q... . bl4-B-H.. 614-A-W . 614-W. . . 614-C-A.. 614-A-C.. 614-A-E.. 614-A-D. . 614-B-W\ 614-B-A 614-C.. .. 614-B-F. . 614-A-lt. 614-A-2i: . 614-AA-2 . 620-a-4... 615-A-l . . Name. Plexaura homomala Gorgonia flabellum Pseudoplexaura crassa . . . Plexaura flexuosa Eunicea crassa Eunicea rissose Biiarium Xiphigorgonia ameeps. . . Gorgonia acerosa Bryozoa Toxopneustes Mellita Clypeaster Ciypeaster Diadema Astrospecten Cassiopea xamachana . . . Pentaceras Holothurla bermudiana. . Holothuria bermudiana. . Loggerhead sponge Unio % . . . Mud from Marquesas.Fla Weight fresh. yms. 1055 550 1205 400 110 560 1235 445 480 1205 1345 1955 * 255 10370 1175 1945 725 3623 Weight dried at 110. gms. 398 288 418 172 98 260 461 190 141 717 698 956 97 127 472 402 404 74 280 Copper. 0.0006 .00025 .00022 .00024 .00014 .00018 . 00009 .00008 .00012 .00014 .00016 .00032 . 00026 . 00022 .00021 .00029 .00021 . 00036 .00012 .00011 .00013 .00026 .0032 Zinc. .0016 .0025 .00023 . 00024 . 00052 . 00066 .00045 .00148 . 00208 . 00084 .00046 . 00042 . 00027 .00034 . 00006 . 00092 .00062 . 00056 .00034 .00064 .00094 .0032 .0018 Iron. .0021 .0032 .0043 .0043 .0026 .0034 .0036 .0032 .0060 .00196 .0026 .00015 .0032 .0036 .0028 .0032 .00016 .0029 .00071 .0031 .0015 .0261 Manga- nese as Mno. 0.00021 .00024 .00029 .00015 .00012 .00008 .00014 .00011 .00019 .00021 .00009 .00065 .00072 .00046 .0002 .00041 .00013 .00037 .00021 Trace. .0030 .0058 * Not weighed. t Composed of the body walls, muscles, etc. j Composed of organs and intestine only. The weight of the sample taken for the determination of copper and zinc was in each case 20 grams of the material, dried at 110° C. to » Jour. Bio. Chem. Soc, June 1919. 97 98 Analytical Search for Metals in constant weight, and the amounts of all the metals, reported in the table above, are in each case expressed in grams per 20-gram sample. The four metals sought occur in all the samples analyzed, but not in as large amounts as in those forms where it is possible to separate the soft tissues from the calcareous skeleton. In some of the present material the calcareous skeleton was by far the larger part, by weight, of the sample taken for analysis, thus reducing the amount of each metal reported in the 20-gram sample, as there is no doubt that the metals here reported are associated to a much greater extent with the soft tissues than with the calcareous skeleton. From a consideration of the amounts of each metal reported in the various organisms, there seems to be no ratio of occurrence or quanti- tative relation of one to the other, but zinc is present in larger amounts than copper. Sample No. 620-a-4 was composed of the soft parts of a number of fresh-water mussels taken from the Millstone River east of Prince- ton, at a point where all the water of the river is collected from a low, level, sandy area far removed from any possible metallic veins. It is interesting to note that this fresh-water form collects both copper and zinc from such a natural water and in about the same amounts as do the salt-water organisms. This may be true of other fresh- water forms, but as yet this sample of Unio is the only one analyzed. Iron and m.anganese in this case were not determined, as it was diffi- cult to completely eliminate the silt contained in the intestines, and as this silt would contain both iron and manganese, their determina- tion under the circumstances would be of no interest. In the note'' A possible source of vanadium in sedimentary rocks''^ it was stated that the conditions for the fixation of these small quanti- ties of metal in the mud and slime at the bottom of the shallows and lagoons are ideal, about the Tortugas at least, as the constant libera- tion of hydrogen sulphide by the mud in the slightly alkaline sea- water would precipitate all four of the above metals as sulphides, even though they were present in very small quantities. To test this, the sample No. 615-A-l was analyzed for these metals. The sample consists of a mud collected in the shallow lagoon of the Marquesas, Florida, and taken by means of a special sampling apparatus at a depth of 2 feet below the surface of the mud. The results of this analysis are shown in the table. It is not surprising to find iron and manganese, but zinc and copper are both present and in larger quantities than the analyses show them to be present in the organisms; therefore these metals are concentrated in this bottom mud. Unfortunately, no other samples of muds were taken and the extent of concentration of copper and zinc can not be determined at present. > Amer. Jour. Sci., vol. xlvi, p. 471. Tortugas Marine Organisms. 99 Copper has been reported as a component of the surface mud of the ocean bottom at other localities, and both zinc and copper have been reported as being present in small amounts in many limestones and dolomites. The calcareous mud of the Marquesas is limestone in the making, and when the conditions under which it accumulates are considered there can be very little doubt that the copper and zinc content is derived from organisms which have concentrated these metals from sea- water, and at death the metal content of their decaying tissues is fixed as sulphides and becomes a part of the limestone or sedimentary rock thus formed. Even though the percentage (0.016 per cent for copper and 0.009 per cent for zinc) is seemingly small, the actual weight of metal present when calculated for large areas of hmestone is considerable, as each cubic meter of rock would contain 432 grams of copper and 243 grams of zinc — an amount quite sufficient to produce metallic deposits of commercial value after secondary concentration by natural agents. VI. THE TRACKING INSTINCT IN A TORTUGAS ANT By ALFRED GOLDSBOROUGH MAYOR 101 THE TRACKING INSTINCT IN A TORTUGAS ANT. By Alfred Goldsborough Mayor. In the preparation of this paper it is a pleasure to acknowledge my indebtedness to Professor William M. Wheeler for his identifica- tion of the ant in question and for references to the literature. All papers previous to 1910 are referred to in Wheeler's masterly work ''Ants," published by the Columbia University Press, while the studies of later authors, such as Pieron, Turner, Santschi, etc., are referred to by R. Brun (1914, Rie Raumorientierung der Ameisen und das Orientierungs problem in allgemeinen, 234 pp., 51 fig. Jena). Another important paper is by V. Cornetz (Les exploration et les voyages des fourmis, 192 pp., 83 fig. Paris, 1914). In this brief paper we will not attempt to review the already voluminous literature, but refer to it onlj^ as it relates to our observations. Monomoriuin destructor Jerdon, a tropicopolitan ant of East Indian origin, was identified in Florida by Wheeler (1906, Entomo- logical News, vol. 17, p. 265). It is a small, reddish-brown ant, and is a great pest in- the wooden buildings of the Tortugas laboratory, making its nests in crevices of the woodwork. So voracious are these insects that we are obliged to swing our beds from the rafters and to paint the ropes with a solution of corrosive sublimate, while all tables must have tape soaked in corrosive sublimate wrapped around their legs if ants are to be excluded from them. These pests have the habit of biting out small pieces of skin, and I have seen them kill within 24 hours rats which were confined in cages. The experim-ents herein described were made on the flat wooden floor of the laboratory, this flatness having possibly prevented the ants from orienting themselves with respect to conspicuous objects or unevenness in the ground, although I have no evidence that they do this under any conditions. In order to attract the ants, a number of recently killed house- flies were impaled upon a pin; and then, upon looking over the floor, one soon found an ant wandering in a tortuous course over the flat surface. The pin with its lure of flies was then thrust into the floor in front of this foraging ant, which would often pass v/ithin 0.25 inch of the lure without perceiving the flies; but if its course were such that it came appreciably nearer than 0.25 inch, the ant suddenly turned toward the flies, and without apparent excitement appeared to "inspect" them, spending a half minute or more crawl- ing over them and stroking them with its antenna?. This applies to 103 104 TJw Tracking Instinct in a Tortugas Ant. dead flies, for if a fl}^ be wounded and moving the ant usually be- comes much excited and proceeds to bite it, and it is remarkable how efficacious the bites are in quieting the fly. In any event this ''finder ant" soon leaves the flies without carrying off any piece of them, but instead of moving off in the erratic and tortuous path it was pursuing before it found the flies, it now goes in a fairly straight path toward some crevice in the floor, out of which there soon pours an excited swarm of its nest-mates, who proceed toward the flies in a fairly straight path, but which is not necessarily identical with that taken by the ''finder ant" in returning from the flies to the nest. Often the path of the return swarm is not quite right in direction, and thus the ants would pass to one side or the other of the flies; but, curiously, when the right distance has been made and the ants are about to pass the flies, the swarm suddenly breaks up into individuals coursing in random fashion in all directions (the "Turner's curves" of authors). It is very dramatic to see the straight path of the once orderly file of ants suddenly break into this random wandering, but so accurately do they gage the distance that I have never seen them miss it by more than 2 inches in a journey of 8 feet, while often the direction may be so much in error that in going 8 feet the ants may tend to pass as much as 4 inches to one side or the other of the flies. Of course, in cases when the path of the main swarm does not happen to "strike" the flies, but passes to one side and then breaks up, many of the ants will not succeed in finding the flies, but must wander in erratic curves over the floor. A considerable number, however, do find the flies, and within a few minutes a fairly straight swarm-path is established between the nest and the flies. Cornetz, observing ants in Algiers, finds that when an ant returns to the nest it pursues a fairly straight path which is more or less right in direction, but in any event, when the ant has gone the correct distance, it begins to wander in more or less tortuous courses until it finds the nest. EXPERIMENTS. In studying the behavior of the Tortugas ants I w^ould usually place a few recently killed house-flies upon the floor in order to draw out the ants, so that many of them would be constantly moving over the floor in all directions and the entrance to the nest would be well defined by crowds of moving ants in its neighborhood. Often paper was pinned down to the floor in order that the paths of the ants might be drawn accurately in pencil. I. When an ant has discovered a dead fly and is engaged in "inspecting" it, we may draw a circular line of a solution of cor- rosive sublimate in 35 per cent alcohol about a foot in diameter The Tracking Instinct in a Tortugas Ant. 105 around the fly. Then, when the ant leaves the fly and proceeds more or less in the direction of the nest, it soon meets with the barrier of corrosive sublimate. This arrests it. It then crawls around close to the inner edge of the barrier, then suddenly goes straight back to the dead fly, and again starts out for the nest, to be again arrested by the barrier of corrosive sublimate. Finally, after re- peating these movements a number of times, it crosses the barrier and goes in a more or less direct path toward the nest; but when it meets other ants and rubs antennae with them they are not excited and no swarm comes back to the dead fly. The fact that the "finder ant" has crossed the corrosive sublimate seems to have destroyed its power to excite other ants or to draw them back to the lure it has found. II. Conversely, if after a "finder ant" has gone back to the nest, and the swarm is well estabhshed and the fly is being torn to pieces, a circle of corrosive sublimate be drawn around the fly, the ants coming from the nest are at once arrested when they reach the outside of the ring of poison. A block occurs and in about a minute nearly every ant between the outside of the ring and the nest is seen to be returning straight to the nest, so that the swarm vanishes in a short time. The ants caught within the ring usually show some hesitation in crossing the poisoned area, and once having crossed, they rarely return to the fly, so that the numbers attacking the fly constantly decrease, due to lack of new recruits. III. On one occasion an ant which was carrying a grain of sand in its mandibles found a dead fly and "inspected" it in the usual fashion, but did not drop the grain of sand. It then crawled off toward the nest, carrying the sand, but no return swarm came. No inferences can be drawn from this observation, however, for it is based on only a single case, although it seems possible that under certain conditions a "finder ant" does not produce a return swarm. IV. If after having found the fly the "finder ant" is allowed to go toward the nest and to rub antennae with several of its nest- mates, and is then gently brushed up from the floor with a camel's- hair brush and removed, the ants it has encountered show normal excitement; and this excitement spreads by contact to others in their neighborhood, but no return swarm occurs. The excited ants rushing to and fro often cross the path the "finder ant" traversed in going from the fly toward the nest, but none of them attempts to follow the trail back to the fly. V. If the "finder ant," after having "inspected" the fly and started toward the nest, is brushed up and carried through the air to the nest-crevice, the ants it falls among may at times display some excitement, but of this I am uncertain. Certainly, however, no return swarm comes back to the fly. Apparently, having been 106 The Tracking Instinct in a Tortugas Ant. carried through the air, the "finder ant" becomes incapable of con- ducting a return swarm to the fly. Indeed, it appears probable that the ''finder ant" has lost its sense of orientation, for if we brush up an ant from the midst of the crowd moving in the neighborhood of the nest-crevice and carry it through the air to a dead fly, say 8 or 10 feet away, the ant "inspects" the fly in a normal manner, but instead of starting on a fairly straight path back to the nest- crevice, it courses widely over the floor in all directions, every now and then turning and going straight back to the fly, and then starting out again. If in the course of these wanderings it meets with several of its fellows and rubs antennae with them, it then returns to the fly, while its mates follow it in much excitement and a swarm starts; but apparently the ''finder ant" has lost its sense of the direction of the nest after having been carried through the air. But while this may apply to Monomorium destructor, it seems not to be true for certain other ants. Thus, Brun found that Formica rufa and several other species of ants have a remarkable sense of direction which it is diflScult to confuse. This sense, according to Brun, is complex and composed of perceptions which are chemical, topographical, tactile, gravitational, and (as Santschi showed) a perception of the direction of hght. They seem also to be able to remember the "lay of the land," even after an interval of 3 weeks between visits to a given region. It is possible that the flat floor of the laboratory at Tortugas presented no topographical features of sufficient definiteness to serve as guides to the ants, who became lost, much as a good woodsman might be lost at sea. It is also possible that different species of ants differ widely in their sense of orientation, and that in Monomorium destructor this sense is some- what poorly developed. VI. If the abdomen of an ant be shghtly notched or split with a pair of fine dissecting scissors, so as to mark the ant, she does not seem to be rendered abnormal in behavior by the operation, although such ants are apt to die after a few hours, apparently through injury to the tracheal system. If such an ant finds a dead fly it "inspects" it in the normal fashion, and then starts oft" normally, but the other ants pay no attention to it and are not excited if it rubs antennae with them. At times, indeed, the normal ants may interfere with the maimed ant, seizing it by the legs, so as to arrest its progress, but generally they wholly ignore its presence. In any event, when the marked "finder ant" has met several of her nest-mates and rubbed antennae with them, she starts back in a straight course to the fly, but none of the nest-mates follow it, and thus no swarm occurs. In one experiment such a maimed "finder ant" repeated this return journey eight times, each time after having rubbed antennae with its nest-mates, but none of them followed her back The Tracking Instinct in a Tortugas A7it. 107 to the fly. It seems that her instinctive behavior as a "finder ant" is unimpaired by her injury, but due to this injury her nest-mates no longer recognize her. VII. It is difficult to prove that the "finder ant" actually con- ducts a swarm of her nest-mates back to the lure she has found, and I tried many experiments to demonstrate or refute this hypothesis. Most of these were unsuccessful, due to the "finder ant" becoming indistinguishable from the numerous ants crowding around her. Several instances, however, seem to give a positive result, leading me to infer that the "finder ant" actually takes the lead and con- ducts the swarm back to the dead fly. These successful experiments were made with unusually small ants, so small that they can be distinguished even among a crowd of their nest-mates, unless, indeed, another equally small individual enters the swarm. These small ants are only about half the size of the average worker. Moreover, workers of normal size are much more numerous, outnumbering these small ants perhaps 50 to 1. These small ants appear to be normal in behavior, for if one of them finds a dead fly it goes through with the usual "inspection," and then starts off deliberately more or less in the direction of the nest. When it meets a cluster of its nest-mates it rubs antennae with them, and this causes intense excitement, which spreads rapidly by contact from ant to ant. As soon as this is accomphshed, the small ant starts back toward the fly in a fairly straight course, but which is rarely or never identical with the path she took from the fly to her mates. The nest-mates crowd around the "finder ant," and others follow these, so that a moving file of ants is seen rushing toward the fly as if an army were moving along its length with the small "finder ant" in the lead. Of course, if other dwarfed workers were seen in this army the observation was thrown out as non-convincing; but in several instances I clearly saw one of these dwarfed ants lead a small swarm, composed entirely of its larger nest-mates, back to the dead fly, and am thus inclined to think that it is the normal function of a "finder ant" to "personally conduct" her nest-mates back to the food she has discovered. VII. A COLLECTION OF FISHES FROM SAMOA. BY HENRY W. FOWLER, Academy of Natural Sciences of Philadelphia, AND CHARLES F. SILVESTER, United States Army. Two figures. 109 A COLLECTION OF FISHES FROM SAMOA. By Henry W. Fowler and Charles F. Silvester. The specimens forming the basis of the present paper were col- lected at Pago Pago in the spring of 1917 by the Carnegie expedition to Samoa. Efforts were made chiefly to secure small or incon- spicuous forms, and, though the collection embraces only 53 species, several are rare and one species is described as new. The collection consists of five lots of small fishes taken from the following locahties. First lot, April 5, 1917, from the cove just south of Aua village and 100 feet northwest of Dr. Mayor's "Aua line." These were taken by Ufting bunches of coral from the bottom and then breaking the coral. The second lot has the same data, except a few specimens screened at the bottom with wire and mosquito screening. The third lot, taken April 6, consists of specimens shaken from coral in the reef in front of the hospital, Pago Pago Harbor, Tutuila. The fourth lot was obtained March 20, 1917, from tide-pools near Double Point, just west of the entrance to Pago Pago Harbor. The fifth lot is simply labeled Pago Pago. The collection is now contained in the Museum of the Academy of Natural Sciences, Philadelphia. The ichthyology of Samoa has claimed the attention of several investigators. The most important general account is the "Fische der Sudsee" by Giinther.^ This was founded largely on the colored drawings of fishes from various Polynesian Islands made by Andrew Garrett. After the first few parts were published the work was discontinued for a number of years, though in 1910 it was finally completed. Previously some of the fishes collected by the Godeffroy firm, which also financed Gunther's ''Fische der Sudsee," were sent to the Vienna Museum and described by Kner and Steindachner.^ Later a collection from Savaii and Upolu was made by Rev. S. J. Whitmee and sent to the British Museum. The percoids from this collection are published in Boulenger's Catalogue.^ Streets made a small collection about 1876, which he later described.* In 1900 Fowler^ reported a small collection made at Apia, Upolu, by Dr. 1 Jour. Mus. Godeffroy, I (Heft i), 1873, pp. 1-24, pis. 1-20; II-III (Heft v-vi), 1874, pp. 25-96, pis. 21-60; IV. 1875, pp. 96-128, pis. 61-83; V (Heft xi), 1876, pp. 129-169, pis. 84-100; VI (Heftxi), 1877, pp. 169-216, pis. 101-120; VII (Heft xv), 1881, pp. 217-256, pis. 121-140; VIII (Heftxvi), 1909, pp. 261-388, pis. 141-160; IX (Heft xvii), 1910, pp. 389-519, pis. 161-180. > Sitz. Ak. Wiss. Wien, 54, 1866, pp. 356-395, 5 pis.; /. c, 58, 1868, pp. 26-31, 293-356, 9 pis. » Cat. Fish. Brit. Mus., I, ed. 2, 1895, pp. 1-394, pis. i-xv. * Bull. U. S. Nat. Mu3.. VII, 1878, pp. 43-102. 6 Proc. Acad. Nat. Sci. Phila., 1900, pp. 524-528. Ill 112 A Collection of Fishes fro??} Sa7noa. H. C. Caldwell, of the U. S. Navy, and received at the Academy of Natural Sciences of Philadelphia in 1857. The most complete work appeared as "The Fishes of Samoa," by Jordan and Scale/ though its scope is widened to include a list of all the species then known from Oceania. Finally Steindachner/ under the title *'Zur Fischfauna der Samoa-Inseln, " reports a collection made by Dr. Rechinger in 1905. OPHICHTHYID^. Chlevastes colubrinus (Boddaert). One example, 683 mm. Aua Reef, Pago Pago Harbor, June 14, 1920. Head 8.4 to vent. When fresh in alcohol grayish white generally, lower surface of tail slightly tinted with pale cream-color. Blackish-brown cross-bands broad, nearly or quite half width of pale interspaces, most all complete, little narrower below, and about edges of each narrow whitish border. Beginning at vent, 10 interspaces with rounded, black blotch within each along fin edge. Also along side large, round, black blotch in each of interspaces, of which several may be dorsal, or some absent and extend for some extent as small blotches. Chlevastes fasciatus (Ahl). One example, 549 mm. Same locality as preceding. Head 8.87 to vent. Differs in dark cross-bands, much narrower, at least much less than one-third Avidth of pale interspaces. Also, all along dorsal surface of trunk pale interspaces, each with small, round blackish blotch but little larger than eye. These extend only on first four interspaces of tail. All dark cross-bands interrupted below, except last three on tail and no dark blotches on anal, which uniform whitish, except last three dark cross-bands. MUR^NIDiE. Gymnothorax punctatus (Schneider). Head about 8; depth at vent about 19; head width about 3.66 in its length; snout 5.5; mouth 3.5; interorbital 5.5; eye 2 in snout. Body moderately long, well compressed, rather slender with convexly flattened sides; tail long, slender, and tapers largely from vent. Combined head and trunk about 1.75 in rest of body. Head rather small, compressed, with slightly swollen pharynx, apparently rather blunt in front. Snout (damaged above) apparently conic and about as broad as long. Eye rounded, little backward in mouth length, without eyelid. Mouth rather small, horizontal. Teeth uniserial in jaws, entire, compressed, attenuate. First 7 teeth each side in front above little larger than others. Vomer in front with 2 similar teeth, front one smaller. No tongue. Row of very small and rather wide-set cutaneous points, minute, along lower lip. Upper lip (damaged) not examined. Jaws apparently equal, ^ lower jaw with low rami, convex and strong. Front nostril in short, fleshy tube near snout tip. Interorbital convex. Occipital region well swollen or convex. Gill-opening, little below median body axis, little inclined from horizontal, length about equals snout. Pharynx smooth. Skin smooth, tough, rather thick. Along each side of mandible 5 pores. Lateral line obscure, with row of indistinct rather wide-set pores along side medially. Dorsal origin apparently about last third in space between posterior edge of eye and front of gill-opening, fin high, especially on last half of tail, and narrowly contin- >BuU. Bur. Fisheries. U. S., XXV, 1905 (Dec. 15. 1906), pp. 173-455, pis. 33-53. » Sitz. Ak. Wisa. Wien, CXV (1), 1906, pp. 1369-1426. • This specimen has the head slightly damaged and due allowance should be made in these proportions. A Collection of Fishes fi'om Samoa. 113 uous with very small obsolete caudal to anal. Caudal length less than eye. Anal less than half high as dorsal. Vent directly in front of anal origin. Color in alcohol very pale grayish white generally, everywhere marked with small, pale brownish, irregularly crowded dots or specks of variable size and density. Spots pale or obsolete along fin borders, but distinct on basal portions of fins. Under surface of head and belly pale, nearly immaculate. Gill-openings very inconspicuous, pale, same as general color. Iris slaty. Teeth whitish. Length about 173 mm. Only the above example from Pago Pago. Probably pale yellowish generally in life, with dark specks. Gymnothorax goldsboroughi Jordan and Evermann, a synonym of the above species, differs from our example in coloration, as it is marked with very many minute whitish or pale spots and has a distinct white fin edge, not seen in our specimen. Gymnothorax pictus (Ahl). Head about 7.75; depth at vent about 15.4; head width 2.4 in its length; head depth 2; snout 6; eye 8; mouth 2.8; interorbital 6. Body moderately long, well compressed, moderately deep, and with convexly flattened sides, long tail tapering largely behind. Combined head and trunk length equals rest of body. Head moderate, compressed, pharynx scarcely swollen, flattened sides but slightly approximate below, front rather robust and upper profile little concave over eye. Snout conic, tip and surface convex, length seven-eighths its width. Eye rounded, little nearer upper profile than mouth, about midway in gape of latter, without eyelid. Mouth moderate, horizontal, closing completely. Lips rather tough, fleshj'', and row of minute papillte or filaments around edge of each. Teeth conic, entire, uniserial along jaw edges. Front teeth in each jaw enlarged as patch of several (6 to 8), strong, erect. Single row of small, erect conic teeth down vomer. No tongue. Upper jaw sHghtly protrudes. Mandible rather low, strong, surface convex. Front nostril in short, fleshy tube about half of eye. Posterior nostril simple pore nearly over middle of eye within interorbital space. Interorbital convex. Occipital region well swollen convexly. Gill-opening near median body axis, slightly inclined from horizontal, about 0.66 of eye. Pharynx smooth. Skin smooth, tough, of about uniform texture. Along each upper lip at least 6 distinct pores well above edge, first slightly in front of nasal tube. Pore directly above upper anterior eye edge in front of posterior nostril. Pair of pores little above bases of front nasal tubes, and another pair well up about midway in snout length. Four distinct pores along each mandibular ramus, well below edge of lip. Lateral line not developed. Dorsal origin about midway between posterior eye edge and front of gill-opening, fin moderately high, though more elevated posteriorly, where confluent with small caudal. Caudal rounded, about long as eye. Anal similar to dorsal, though much lower. Vent about an eye diameter in front of anal origin. Color in alcohol, olive-brownish generally, washed with pale lilac-gray, producing a more or less uniform tint. Though visible to the naked eye as very fine reticulations or specks, under a lens bodj^ seen to be everywhere marked with dusky to blackish- brown vermiculations, extremely minute, though well defined. End of tail and muzzle tinged slightly more brownish. No dusky blotch at gill-opening, or at mouth corner ; latter pale inside. Iris dull slaty. Teeth pale. One 115 mm. long, from Pago Pago. It differs from any example of the species we have seen in its very minute, dark vermiculations. Among the many figures of Bleeker is none of the small size of our own example or with its color pattern. 114 A Collection of Fishes from Samoa. Anarchias allardicei (Jordan and Seale). Head 6.87; depth at vent about 15.25; head width 3.16 in its length; head depth 2.87 ; snout 5.75; eye about 9; mouth 3.12; interorbital 5.75. Body moderately short, well compressed, rather deep, with convexly flattened sides and taU tapering rather abruptly behind. Combined head and trunk 1.2 in rest of body. Head moderate, compressed, depth slender and pharynx not swollen, about even in width, convex above and below. Muzzle rather obtuse, upper profile slightly con- cave over eye. Snout obtuse, convex at tip and on dorsal surface, length three-fourths its width. Eye rounded, about median in depth over mouth, little backward in gape length, without eyelids. Mouth rather small, horizontal, completely closes. Lips rather tough, fleshy, entire along edges. Teeth conic, entire, subequal, strong. Upper teeth with one series small, mostly uniform and erect all around outer edge of jaw and inner series of enlarged, depressible, wide-set sharp-pointed teeth on both sides. Lower jaw with similar dentition. Front of vomer with two large fangs and row of few small teeth down its shaft behind. No tongue. Upper jaw tip slightly protrudes, and mandible with strong rami. Anterior nostril short, fleshy tube about as long as pupil, near snout tip. Posterior nostril simple pore over eye center within interorbital space, which is convex. Occipital region not especially swollen, convex. Gill-opening close to ventral profile, as simple pore, size of posterior nostril. Pharynx smooth. Skin smooth, tough, 5 pores along each side of upper hp, first in front of anterior nostril, second close behind base of anterior nostril, third nearer eye than second, fourth below front pupil edge, and fifth close behind eye. Pair of pores above, close to and within front internasal space, second pair midway on snout above, third pair adjoin hind nostrils over eyes. Lower edge of mandible with 5 pores each side^ grad- ually more distant from one another backward. No lateral fine. Dorsal begins as very slight ridge over gill-opening; extends back also as very slight fold to caudal, where it is a little broader. Caudal rounded, about as long as eye. Anal developed only as low fold on under surface of tail about last two-elevenths of its length, continuous also with caudal. Color in alcohol uniform dusky brown above. Under surface of head, belly, and end of tail tinted brownish; hind edge of latter whitish. Iris pale slaty. One example, 116 mm. long, from Pago Pago. Varies from the original account in the presence of two large anterior vomerine teeth, and no smaller posterior vomerine teeth are mentioned by its describers. The figure of A. allardicei shows the dorsal origin beginning apparently nearer the mouth corner than the gill-opening, while in our example it begins over the gill-opening. A . allardicei has been united^ with A.knighti Jordan and Seale, but its mottled coloration and more elevated dorsal doubtless renders it distinct. HEMIRAMPHIDiE. Hyporhamphus pacificus (Steindachner). Head from upper jaw tip 4.6; depth 9.33; D. iii, 14; A. iii, 16; P. i, 10; V. i, 5; scales 66 in lateral series from shoulder to caudal base and 6 more on latter; about 7 scales above lateral line to dorsal origin; 2 scales above anal origin to lateral fine; snout about 2.66 in head without beak; eye 3.33; maxillary 4.5; interorbital 4.33; pectoral 2.25; first branched dorsal ray 2.87; first branched anal ray 3.5; least depth of caudal peduncle 5.25; lower caudal lobe 1.25; ventral 3.2. Body elongate, rather robust, slightly compressed, though sides are convex and not flattened, deepest medially. Caudal peduncle compressed, least depth half its length. » Glinther, in Jour. Mua. GodefiFroy, XVII, 1910, p. 421. A Collection of Fishes from Samoa. 115 Head compressed, flattened sides approximated below where width is half that of cranium, well attenuated forward. Snout long, depressed, width 1.75 in its length. Eye elongately ellipsoid, close to upper profile, slightly advanced in head (without beak). Free portion of upper jaw nearly an equilateral triangle as seen from above, its length 2.4 in snout. Maxillary 1.5 to eye, broadly vertical, width equals pupil. Lower jaw long, slender, so rest of head from upper jaw tip only 0.75 of remainder of beak. Teeth fine, simple, in narrow bands in jaws. Upper buccal fold narrow, lower broader. Tongue elongate, depressed, smooth, free. Nostrils rather large, together, their depression as long as pupil along upper snout edge close before eye. Interorbital depressed to very slightly concave. Opercle broad, smooth, width 1.25 eye diameters. Preorbital slightly less than eye. Gill-opening forward to front eye edge. Rakers 10+23, lanceolate, 1.5 in gill- filaments and latter 1.75 in eye. Isthmus narrowed, trenchant frenum in front. Scales deciduous, all well imbricated and above computations, largely according to pockets. Dorsal and anal mostly covered with small scales, at least basally, also caudal base. Scales with basal circuli 19. Lateral line complete, apparently low along side, touches at ventral origin, tubes simple and each well exposed. Dorsal origin well posterior, much nearer caudal base than ventral origin or little behind last third in space between caudal base and pectoral origin, anterior rays longest, though their tips extend only to middle of fin when depressed, and entire depressed fin three-fourths to caudal base. Anal inserted opposite dorsal, similar. Caudal well forked, lower lobe much longer than upper (damaged). Pectoral base high, fin 4 to ventral origin, latter midway between pectoral origin and caudal base, fin short or but 2.25 to anal. Vent close before anal. Color in alcohol brownish on back, paler on under side, apparently whitish in life. Down middle of back well-marked dusky hne with narrow one each side and parallel, also scale edges same tint. From shoulder, narrow silvery-white band to caudal base, widest below dorsal, where about two-thirds vertical eye diameter, and its upper border tinted slaty narrowly or with deeper line. Fins all pale brownish, vertical ones and pectoral above tinted little with grayish. Iris silvery white, also side of head. Inside gill-opening marked with dusky dots. One 205 mm. long, from Pago Pago. Agreement was found with Hawaiian examples, which have rakers 9+22. Hyporhamphus samoejisis Steindachner,i as suggested by Giinther, is probably the same. This species is doubtless identical with Hemiramphus dussumieri Valencieimes. MUGILID^. Neomyxus chaptali (Eydoux and Souleyet). Head 3.4; depth 3.4 to 3.5; D. IV-I, 9; A. Ill, 9; scales 37 to 39 in median lateral row to caudal base and 5 more on latter; 12 or 13 scales transversely between dorsal and anal origins; about 21 or 22 predorsal scales; snout 3.5 in head; eye 3.25 to 3.33 in head; mouth width 2.87; interorbital 2 to 2.4. Body compressed. Head broad above, constricted below, upper profile nearly straight. Snout broadly obtuse as seen dorsally, length two-fifths its width. Eye large, posterior edge midway in length of head, rim free. Premaxillaries concealed. Upper front lip thick, width slightly over half of eye. Edges of Ups with single row of rather large fleshy papillae. Mandible included in upper jaw. Nostrils small, close, near upper edge of snout. Interorbital broadly convex, with slight depression in front. Rakers 22+36, slender, lanceolate, little less than filaments, the latter 1.66 in eye. Scales large, firm, in even longitudinal rows; basal radiating striae 4 to 6, with 3 to 5 incomplete accessory ones; circuli rather coarse. Dorsal, anal, and caudal largely 1 Sitz. Ak. Wiss. Wien, CXV (1), 1906, p. 1418, Upolu. 116 A Collection of Fishes from Samoa. covered with small scales. Spinous dorsal origin about opposite pectoral tip. Soft dorsal inserted little behind anal origin. Second anal spine but little shorter than third. Pectoral reaches half-way to anal. Ventral inserted about opposite middle of depressed pectoral. Vent close before anal. Color in alcohol brownish on back, sides and below silvery white. Dorsals and caudal tinted with dusky, also pectoral, the latter with small dark spot at origin. Iris whitish. Length 73 mm. (caudal damaged). Two small examples from Pago Pago. HOLOCENTRID^. Holotrachys lima (Valenciennes). One example, 68 mm. long. Holocentrus punctatissimus Cuvier. Head 2.6 to 2.75; depth 2.66 to 2.8; D. XI, 14 or 15; A. IV, 9 or 10; scales in latera; line 35 or 36 to caudal base and 2 more on latter; 4 scales above lateral line to soft dorsal origin; 7 scales below lateral line to spinous anal origin; 7 predorsal scales snout 3.8 to 4.2 in head; eye 2.66 to 2.75; maxillary 2.75 to 2.66; interorbital 4.12 to 4.2. Head about half as long as wide. Snout length two-thirds its width. Pos- terior edge of pupil about midway in head length. Jaws about even. Maxillary two-fifths in eye, expansion 2.75. Bands of villiform teeth in jaws, on vomer and palatines. Interorbital level. Cranial bones striate. Preopercle spine long, strong, reaches back slightly beyond gill-opening, length two-fifths of eye. Preorbital narrow, with several strong marginal spines. Suborbital chain about equally wide as pre- orbital, its serrated edge finer. Edges of suprascapula, preopercle, opercle, and subopercle serrate. Preopercle ridge entire. Rakers ii 2-f 8 iii, lanceolate, nearly long as filaments, which one-third of eye. Scales largest on flanks, smaller on predorsal and vertical fin bases. Cheek with 4 rows of scales. Scales with basal parallel vertical striffi 18 to 25 (more numerous in larger examples); 7 to 13 strong, broad apical spines. Color when fresh in alcohol pale orange, generally as ground-color. Back with two and a half longitudinal rows of dusky brown, narrow, and not sharply defined, parallel with lateral line above. Below lateral line six and a half broad longitudinal rows of similar color, lower much narrower. Head brownish above, tinged with pale orange below, and each scale on cheek and oparcles with pale dusky spot. Spinous dorsal grayish, with first three and last membranes jet black. Other fins all pale orange to whitish. Iris silvery, fading slaty. Four examples, all young, shaken from coral from reef in front of hospital, Pago Pago Harbor, 42 to 45 mm. long. Also 3 examples 64 to 71 mm. long, from tide-pools near Double Point, just west of the entrance to Pago Pago Harbor. These largely in agreement with Giinther's figure.^ Two small examples from cove just south of Aua village. These have a conspicuous dark blotch in the front of the spinous dorsal. Holocentrus diadema Lac6pfede. Head 2.87; depth 3.33; D. XI, I, 12, i; A. IV, 9; scales 50 in lateral line to caudal base and 3 more on latter; 5 scales above lateral line to origin of soft dorsal; 9 scales below lateral line to origin of spinous anal; 9 predorsal scales; snout 4.25 in head measured from upper jaw tip; eye 3; maxillary 2.87; interorbital 4.25. Width of head half its length. Length of snout four-fifths its width. Eye with posterior edge of pupil midway in head-length. Closed lower jaw slightly projects. Maxillary reaches beyond anterior edges of eye, not quite to pupil, expansion about one-third in eye. Bands of villiform teeth in jaws, on vomer and palatines. Interorbital level. Cranial > Proc. Zool. Soc. London, 1871, p. 660, pi. 60, 2 figs. A Collection of Fishes from Samoa. 117 bones striate. Preopercle spine reaches back only to bony edge of infraopercle. Preorbital very narrow, and suborbital chain but little wider, its edge more weakly serrate than finely serrated preorbital edge. Rakers ii 2+9 iii, lanceolate, about three-fourths of filaments, which one-third of eye. Scales largest on flanks, smaller on predorsal and breast. Cheek with 5 rows of scales. Scales with parallel vertical striae 60 to 70 basally; 8 blunt, short basal denticles; 16 to 18 broad, strong apical spines. Color when fresh in alcohol bright rosy-red generally, with 3 rows of narrow, dark longitudinal bands above the lateral line and 5 broad ones below. Later these faded out below and made up of brownish dots, as seen under a lens. Upper surface of head washed with pale browTiish. Iris silvery, fading slaty. Spinous dorsal pale rosy generally, fading whitish, except large median black blotch on first two membranes, then black blotch submarginallj'^ after each dorsal spine, and from fourth spine basally black band back to last spine. Other fins all uniform pale rosy, fading whitish. One example, 78 mm. long, from coral in reef in front of the hospital, Pago Pago Harbor. It differs from examples in the Academy of Natural Sciences of Philadelphia in the much shorter preopercular spine and the coloration. It has more whitish on the spinous dorsal, lacks entirely the blackish on the front part of the ventrals and anal, besides having a pale or whitish pectoral axil. Holocentrus praslin Lacdp^de. Small examples from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Holocentrus sammara (Forskal). One example from Pago Pago, 100 mm. long. It agrees with Bleeker's figure in the anterior dark blotch on spinous dorsal median and lengthwise. Jordan and Scale' describe four examples from Samoa. First has "spinous dorsal broadly edged with blood red." Second has "dorsal maroon, whitish spots at base, tips white, and front of fin with large, black, red-washed blotch." Third with "large black blotch on front of spinous dorsal." Fourth with "front of soft dorsal with very large blotch of maroon- black, fin otherwise flesh-color, tips white." In a Hawaiian example Jordan and Evermann show lengthwise lines made up of dark spots. Also 3 small examples from cove just south of Aua village, April 5, 1917. In two of these the front of the spinous dorsal has a large black blotch and succeeding mem- brane with less distinct dark blotches. Remaining example with spinous dorsal uniformly pale or whitish. CHEILODIPTERID^. Amia savayensis (Giinther). Five from reef in front of hospital, Pago Pago Harbor. Fourteen from cove just south of Aua village. These agree in every way with the large series of Philippine examples in the Philadelphia Academy. Amia novemfasciata Cuvier. Adult and young example from tide pools near Double Point, just west of entrance to Pago Pago Harbor, March 20, 1917. Two also from Pago Pago. Fowleria marmorata (Alleyne and Macleay). Head 2.5; depth 3, D. VIII-I, 9; A. II, 9; scales 23 in lateral hne to caudal base and 2 more on latter; 2 scales above lateral line, 6 below; 8 predorsal scales; snout 4,16 in head; eye 3.87; maxillary 2; interorbital 5.5. Body elongate, compressed, deepest at spinous dorsal origin, profiles alike. Caudal peduncle compressed, least depth 1.12 its length. Head large, compressed, profiles > Bull. Bureau of Fisheries, XXV, 1905, p. 227. 1 J 8 A Collection of Fishes from Samoa. alike. Snout convex on dorsal surface, slightly so in profile, length about three-fifths its width. Eye large, impinging on upper profile, posterior edge about midway in head length. Mouth large, well inclined, lower jaw slightly included. Maxillary extends beyond posterior edge of pupil, not quite to posterior edge of eye, expansion little less than pupil. Villiform bands of teeth in jaws and on vomer, none on palatines. Tongue elongated, rounded tip free. Nostrils together, directly in front of eye. Interorbital level. Preopercle edge entire. Rakers 1+6 short points, about equal to filaments, which one-third of eye. Scales large, Httle smaller along body edges, and 2 rows on cheek. Basal radiating strise 13 or 14; apical denticles 80 to 90; circuli fine. Lateral line of 5 well-exposed, simple tubes first, then only as row of pores to caudal base, one in center of each scale exposure, and all concurrent largely with dorsal profile. Third dorsal spine longest, reaches back to soft dorsal origin. Soft dorsal and anal alike, opposite, rather elongate, height of former slightly less than half of head. Caudal rounded. Pectoral reaches anal origin, ventral little shorter. Color in alcohol deep brick-brown, with 9 vertical cross-bars, twice width of pale interspaces. Head mottled brownish, \\ith pale lilac tint on mandible and branchi- ostegal region. I^arge jet-black, round blotch, little less than eye, though larger than pupil and margined narrowly with golden-brown. Small, black crescent above opercular spot. Pale bar from eye to preopercle angle, lower edge dusky. Each scale on caudal peduncle with median dusky blotch, rather small, though distinct. Two rows of scales between pectoral and ventral bases on side of abdomen, with slightly oblique, narrow dusky line. Fins all dusky-red. Iris brownish. Length 47 mm. Also smaller examples same locahty. Head 2.5; depth 2.87; D. VII-I, 9; A. II, S; scales about 21 in lateral hue to caudal base and 2 more on latter; snout 3.75 in head; eye 3.33; maxillary 1.8; interorbital 4; length 33 mm. This approaches Apogonichthys isostigma Jordan and Seale, in the more spotted appearance, which possibly may not be distinct from A. marmoratus Alleyne and Macleay, as the black spots on the trunk seem to be the chief character of distinction. Both from cove just south of Aua village, April 5, 1917. SERRANIDiE. Epinephelus merra Bloch. Two young examples from cove just south of Aua village. It differs from the adult stage in the much larger dark blotches. Pharopteryx nigricans Riippell. Two small examples, from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. One example from Pago Pago. All show D. XII. Length 48 to 57 mm. Pharopteryx melas (Bleeker). Two from cove just south of Aua village. In alcohol, Iwdy dusky-bro\\Ti generally, clouded with blackish. Head same, Httle paler below. Iris slaty. Bases of vertical fins pale or largely whitish, all broadly blackish about outer or terminal portions. Spinous dorsal edge, together with upper soft dorsal edge, especially in front, orange. Pectoral and ventral brownish. Length 50 to 55 mm. One 35 mm. long, same locaUty. All have D. XL OPISTHOGNATHID^. Gnathypops samoensis new species. Fig. 1. Head 2.75; depth 3.33; D. VII, 20; A. Ill, 17; P. 15; V. I, 5; scales from shoulder to median caudal base about 50, and 10 more on latter; 31 tubes in lateral line; 4 scales above lateral line to soft dorsal origin; 22 scales in vertical series below lateral line to A Collection of Fishes jrom Samoa. 119 spinous anal origin; about 40 predorsal scales; head width 1.8 its length; head depth at occiput 1.4; mandible 2.1; sixth dorsal spine 3.8; sixth dorsal ray 2.75; second anal spine 6; sixth anal ray 2.75; least depth of caudal peduncle 3; caudal 1.87; pectoral 1.6; ventral 3.12; snout 5 in head, measured from upper jaw tip; eye 5; maxillary 2; interorbital 7.5. Body oblong, compressed, deepest at front of spinous dorsal, edges rounded con- vexly. Caudal peduncle well compressed, length about seven-eighths its least depth. Head compressed, upper profile little more inclined than lower, flattened sides plightly approximated above. Snout convex over surface and in profile, short, length half its width. Eye small, advanced, posterior pupil edge near first third in head length. Mouth large, oblique, lower jaw prominent and slightly protrudes, rami robust and moderately high inside mouth. Maxillary extends back slightly beyond posterior edge of eye, though not quite halfway in head length, expansion equals eye. Lips fleshy, moderately broad. Teeth fine, pointed, in bands in jaws and on vomer and palatines. Tongue rather slender fleshy point, free and smooth. Anterior nostril in short tube near front end of snout, posterior one simple pore close to anterior edge of eye medially. Interorbital narrow, nearly level. Preorbital narrow, less than half of eye. Preopercle edge uneven, largely convex. Fia. 1. — Gnathypops samoensis Fowler and Silvester. Type. Gill-Opening in front of posterior edge of eye. Rakers iv 1+6 in, broad, asperous knobs, longest three-fourths of filaments and latter one-third of eye. Pseudobranchife large as gill-filaments. Isthmus moderately broad, slightly constricted forward. Scales moderate, smooth on front part of body, as on head, predorsal, breast, and trunk toward end of depressed pectoral, after which finely ciliated entirely. Scales on trunk in even longitudinal rows, small and crowded along edges of body, though less on breast than at most areas. Snout, front preorbital, maxillary, lips, chin, and branchiostegal region naked, head otherwise scaly. Several pores on suborbital chain close to eye, others on mandible. Fourteen scales across widest extent of cheek. Scales on opercle largest on head. Fins all with small scales basally, extending well out on dorsals and anals. Scales with basal parallel marginal stria? 11 to 13; circuli parallel laterally and same end as 4 or 5 apical denticles of small size. Lateral line only as superior branch from shoulder along back near upper profile, and not extending beyond soft dorsal. Tubes simple, large, well exposed. Spinous dorsal inserted over origin of pectoral, spines all slender, low, pungent, graduated up to third and then about uniform. Soft dorsal inserted about midway between posterior edge of preopercle and base of caudal, fin uniformly high, rounded behind, and like soft anal posterior rays extend back to caudal base. Second anal spine longest, first little longer than third, origin much nearer origin of pectoral than base of caudal. Soft anal like soft dorsal. Caudal rounded, with median rays longest. Pectoral elongate, pointed median rays longest, reaching anal. Ventral origin slightly 120 A Collection of Fishes from Samoa. in front of pectoral origin, not quite reaching halfway to anal, spine half of length of fin. Vent directly in front of anal. Color when fresh in alcohol rich blackish-chocolate, largely uniform, with shght lilac tinge on branchiostegal region. Fins all largely blackish; also blackish blotch on opercle. Iris dark brown. Length 61 mm. Type No. 50,563 A. N. S. P. Cove south of Aua village, 100 feet and northwest of Dr. Mayor's "Aua hne." Taken by lifting bunches of coral from the bottom, April 5, 191 r. Also No. 50,564 paratype, A. N. S. P., same data: Head 2.5; depth 3; D. VII, 20; A. Ill, 17; scales from shoulder to median caudal base 50 and 11 more on latter; snout 5.2 in head from upper jaw tip; eye 4; maxillary 2; interorbital 7; length 55 mm. This interesting species has no close aUies and is the first occurrence of the family in Samoan waters. (Named for Samoa.) POMACENTRIDJE. Pomacentrus melanopterus Bleeker. One from Pago Pago, 86 mm., differs from Bleeker 's figure,^ as he shows the pre- orbital with a spine, the suborbital rim serrate, and the posterior edge of the preopercle almost entirely serrate. The black pectoral basal blotch is shown as a dark bar within a crescent. Bleeker says,^ however, "ossibus suborbitalibus alepidotus — non vel vix denticulatis; osse prseorbitali . . . incisura plus minusve profunde ab ossibus suborbitalibus ceteris distincto, postice rotundo vel in spinuUura desienta." Pomacentrus nigricans (Lac^p^de). Two young and one adult from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Apparently the same as Jordan and Scale's material, as the squamation extends much further forward than shown by Giinther.^ All our examples show a black blotch at pectoral axil and another at last dorsal ray bases. Pomacentrus albofasciatus Schlegel. Our material includes 4 examples from coral in the reef in front of the hospital at Pago Pago Harbor; 10 from cove just south of Aua village; 2 from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Abudefduf ccelestinus (Cuvier). Young example with 6 transverse dark bars, largely reflected on fins, though dorsals and anals largely and caudal completely whitish. Length 18 mm. Pago Pago. Abudefduf glaucus (Cuvier). Four small examples from cove just south of Aua village, and 24, all dull and uniform in color, from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Abudefduf zonatus (Cuvier). Seventeen from Pago Pago. No trace of the white lateral bar, though head and back are thickly spotted with pale blue. Bleeker's figure^ does not show the blue spots as distinct and variegated as in our examples. Glyphidodon brownriggii Gtinther^ has been referred to the present species, but none of his figures show spots, and though his figure A is perhaps closer, it has the dorsals and anals broadly dark. " Atlas Ich., IX, 1877, pi. 42, fig. 6. 2 Nat. Verb. Hollands. Maatsch. Wetensch. (Mem. Pomacent.) (3), Deel. 2, No. 6, 1877, p. 65. •Jour. Mus. Godeffroy, VII (Heft, xv), 1881, pi. 124 f.y. • Atlas Ich. IX, 1877, pi. 407, fig. 3. • Jour. Mus. Godeffroy, VII (Heft, xv), 1881, pi. 127, figs, a, c, c. A Collection of Fishes from Samoa. 121 Dascyllus aruanus (LirniiEUs). Three small examples from cove just south of Aua village. Chromis caeruleus (Cuvier). Two adults from Pago Pago. Chromis isomelas Jordan and Seale. Head 3.16; depth 1.87; D. XII, 13; A. II, 14; P. I, 15; V. I, 5; tubes 14 in upper arch of lateral line and 7 porous scales in horizontal section before caudal base; 3 scales above lateral hne to origin of spinous dorsal; 9 scales below lateral line in vertical row to origin of spinous anal; 19 predorsal scales; width of head 1.5 its length; head depth at occiput 1 ; snout 4; eye 2.75; maxillary 3.2; interorbital 2.4; fourth dorsal spine 2.12; ninth dorsal ray about 1.75; second anal spine 1.9; ninth anal ray 1.4; least depth of caudal peduncle 2. Body strongly compressed, deeply ellipsoid, deeper midway in combined head and trunk. Edges all convex. Caudal peduncle well compressed, long as deep. Head deep, profiles about evenly inclined, with upper very slightly concave over eye, compressed and flattened sides slope evenly above and below. Snout surface convex, also profile, length half its width. Eye, also pupil, slightly ellipsoid, Uttle advanced or with posterior edge about midway in head length. Mouth small, oblique, terminal. Lips narrow, rather thin. Teeth conic, in rather broad bands in jaws, outer row slightly enlarged, also extend all along premaxillary edge. Mandible even with upper jaw tip when closed, rather shallow and rami well elevated behind inside mouth. Buccal membranes (breathing valves) present inside mouth, upper broader. Tongue pointed, free, smooth, rather elongate. Nostril small, simple pore, about midway on side of snout. Interorbital evenly convex. Preorbital narrow, about two-fifths of eye. Suborbital and preopercle edges entire. Gill-opening forward to anterior edge of eye. Rakers 7+20, lanceolate, but httle shorter than filaments and latter shghtly less than half of eye. Pseudobranchiae large as gill-filaments. Isthmus narrowly constricted in front. Scales large, minutely ctenoid, rather narrowly imbricated, smaller along edges of body. Cheek with 4 rows of scales. Single row of scales on preorbital and infra- orbital. Small scales crowded densely over bases of vertical fins, though on spinous portions forming a sheath basally, row of scales up behind each spine on membrane. Small scales at pectoral base. Ventral with pointed axillary scale about one-third of fin-length, median flap between fins about three-fourths length of axillary flap. Scales with basal radiating striae 7 to 10, sometimes 2 or 3 auxiliaries; small apical denticles 98 to 110; circuli fine. Upper arch of lateral Une extends back opposite eleventh dorsal spine base. Tubes large, simple, each well exposed or over first three-fifths of scale; also continued irregularly as 4 pores, then drops a scale and 2 more pores below soft dorsal. Horizontal section of lateral line of simple pores, begins below soft dorsal opposite third pore of upper section, skips 1 or 2 scales, then continues to caudal base. Spinous dorsal inserted inmiediately after pectoral base or much nearer snout tip than origin of soft dorsal, spines graduated up to third and fourth, the longest, others posteriorly but slightly shorter, edge of fin notched and little cutaneous flap behind each spine tip. Soft dorsal inserted about last third in space between origin of spinous dorsal and base of caudal, fin pointed, median rays longest. Spinous anal inserted opposite ninth dorsal spine base or httle nearer pectoral origin than caudal base, second spine longer, first two-fifths its length. Soft anal like soft dorsal, only larger. Caudal deeply forked and outermost rays of each lobe produced in long slender points, length 1.87 in combined head and trunk. Pectoral reaches anal, upper rays longest, and fin about 3 in combined head and trunk. Ventral inserted sHghtly behind origin of pectoral, first ray ends in filament reaching soft anal origin. Ventral spine three- fifths of fin. Vent diiectly in front of anal. 122 ^4 Collection of Fishes from S>amoa. Color when fresh in alcohol, the front half of entire body is deep blackish brown, hind portion white, line of demarcation very striking or exactly midway between eye center and caudal base. Pectoral base and dorsal jet-black. Iris blackish-brown, golden circle around black pupil. Length 75 mm. Pago Pago. Another with same locality shows: Head 3.16; depth 1.87; D. XII, 14; A. II, 13; 14 tubes in upper arch of lateral line; snout 4 in head; eye 2.5; maxillarj^ 3.25; inter- orbital 2.25; length 60 mm. Concerning C. dimidialus, Jordan and Scale state; "It is very close to our Chromis isonielas, but according to the figure by Dr. Giinther, and the description of Dr. Klun- zinger, the posterior boundary of the black area is at the front of the anal fin." Turn- ing to Giinther 's figure of Heliastes dimidiatus,^ one finds such is not the case, as Giinther shows the dark anterior area extending almost to the caudal peduncle, at least over a good portion of the soft dorsal and certainly over more than half of the anals. No dark pectoral blotch is indicated, though Giinther says that the pectoral base is black. It thus appears that his figure represents a variation of C, dimidiatm, and it is quite likely C. isomelas is simply another variation. Klunzinger says" that the dark anterior color extends to origin of anal, bases of pectorals and ventrals black, pectoral hyaline. He mentions only one example, 60 mm. long, and states that its caudal has elongate points. LABRIDJE. Platyglossus notopsis (Valenciennes). Tliree young examples from Pago Pago, quite unlike the adult in coloration. In alcohol our specimens are generally dull brown. Five longitudinal bands, expanded medially, each bordered broadly with blackish-brown so as to form 10 bands all together. Bands on head much narrower and pale areas thus wider. On trunk inter- vening dark areas of each pair of dark bands mottled or blotched obscurely darker. Dorsals and anals black. Middle of spinous dorsal with large black ocellus, edge narrowly whitish, equal to 1.5 eye diameters. Caudal base with blackish bands extending short space, then end abruptly, rest of fin white. Pectoral with dusky base, fin otherwise gray-white. Ventral dusky. Iris slaty-brown. Largest example 44 mm. Smaller examples show pale brown, general color paler or more whitish, also pale bands much broader. In very young each dorsal white, with 2 black blotches, side wth 2 broad longitudinal blackish bands and parallel short band on back and breast. Also bands end on caudal base as black blotch to each caudal lobe. Median blackish band on head above and another below upper forks at interorbital to form band each side of back, and lower extends back to join abdominal band of each side behind. Otherwise coloration whitish or pale. Also, small example from cove just south of Aua village. Cheilinus fasciatus (Bloch). Young example from cove just south of Aua village. SCARICHTHYIDiE. Callyodon rubro-violaceus (Steindaclmcr). Head 2.33; depth 2.8; D. IX, 10; A. Ill, 9; scales 19 in upper arch of lateral line, 5 in lower section to caudal base and 2 more on latter; 1 scale above lateral Une and 6 below; snout 3 in head; eye 4.25; mouth 6.5; interorbital 3. Three rows of scales on cheek, of which lower row on preopercle limb. Body elongately ellipsoid, com- » Jour. Mu8. Godeffroy, VII (Heft, xv), 1881, p. 237, pi. 125, fig. E. ' Verb. Zool. Bot. Gee. Wien, XXI, 1871, p. 29. A Collection of Fishes from Samoa. 123 pressed. Head ratlicr pointed. Eye large, slightly advancetl. No posterior canines. Lips wide, upper covering greater part of teetli. Caudal slightly convex behind. Color in alcohol dull olive-brownish generally, scarcely paler below, and without conspicuous markings, though center of each scale little paler. Dark line on upper lip. Dorsals and anals mottled brownish medianly. Caudal pale brownish, with about 4 -^'ery obsolete or faint browTiish cross-bars. Iris slaty. Length 46 mm. From coral reef in front of hospital, Pago Pago Harbor. Head 2.33; depth 2.33 to 2.87, D. IX, 10; A. Ill, 9; scales 19 in upper arch of lateral line, 4 in lower section to base of caudal and 1 more on latter; 1 scale above lateral line and 6 below; snout 3.25 to 3.33 in head; eye 3.33 to 3.5; mouth 4.5 to 4.66; inter- orbital 2.75 to 3. Three rows of scales on cheek. Body elongately ellipsoid, compressed. Head pointed. Eye large, slightly advanced. No posterior canines. Lips moderate, teeth broadly exposed. Caudal slightly convex behind. Color in alcohol pale brownish-olive with 4 broad dark-brownish longitudinal bands, expanded medially and broader than pale interspaces, but not so on head. Pale-brown bar down from lower anterior edge of eye, another from posterior edge. Lips dusted brownish. Head mottled brownish above. At caudal base each median dark body-bar ends as blackish blotch at base of each lobe, fin otherwise whitish. Dorsals and anals largely deep dusky, at least basally, rest with other fins pale. Length 27 mm. Two from cove just south of Aua village. Giinther unites this species with C. rvberrimus Jordan and Seale, and questionably includes Paeudoscarus rubro-violaceus Steindachner and Scarus paliica Jenkins. CHiETODONTID^. Chsetodon trifascialis Quoy and Gaimard. Two examples, 20 to 31 mm. long, from coral in reef in front of hospital, Pago Pago Harbor. Chaetodon pelewensis Kner. One from Pago Pago. A comparison with Giinther's figure shows how crude his representation really is. The scales in our specimen are all much finer on the fins, pale ocular bar has dark border above the eye, ends of upper jaw dusky, 7 oblique dark bars but little curved and lowest nearly straight, with one on anal sub-basally, the other close to the body edge, though not extending on caudal peduncle. Row of dark spots between each defined dark bar and parallel. Chaetodon melannotus Schneider. Young example, 26 mm. long, from coral in reef in front of hospital, Pago Pago Harbor. It agrees largely with Day's figure,' except that the ocular bar is broader, no black submarginal dorsal, anal, and caudal line, and the black on caudal peduncle encompasses most all of the fin, leaving only narrow white crescent across caudal base. Two other examples, 24 mm. long, from cove just south of Aua village. Chaetodon miliaris Quoy and Gaimard. Four from cove just south of Aua village, largest 23 mm. Close to the young of C. nielannoUiS, but differ in slightly less inclined lines on sides of body and presence of blackish ventral and anal edge, last more broad anteriorly. Holacanthus nicobariensis (Schneider). One small example from same locality as last. > Fishes of India, I, 1875, pi. 28, fig. 1. 124 A Collection of Fishes from Samoa. ACANTHURID^. Hepatus atrimentatus Jordan and Evermann. Small example from coral in reef in front of hospital, Pago Pago. Traces of longi- tudinal blue lines, better separated and fewer than in the original figure.' Also only- trace of black blotch at bases of last dorsal rays. Hepatus triostegus (LinnaBUs). Four small examples from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Jordan and Seale say: "This seems lilce Hepatus sandivichensis, but lacks one cross-bar and is very pale, only four bands on sides." Such is not the rase viith our material, as all are like Jordan and Evermann's Hawaiian figure,'' except that thej^ all have the dark cross-bar on caudal peduncle above and below, but broken m.edianly on each side of caudal peduncle. Jordan and Evermann do not show it on the lower surface of the caudal peduncle. Black bar at pectoral base in our examples extending below only slightly, if at all. MONACANTHIDiE. Oxymonacanthus longirostris (Schneider). Small examples from cove just south of Aua village. TETRODONTID^. Canthigaster solandri (Richardson). Three examples, largest 32 mm., from cove just south of Aua village, tide-pools near Double Point, just west of entrance to Pago Pago Harbor and Pago Pago. SCORPiENIDJE. Sebastopsis guamensis (Quoy and Gaimard). Two from Pago Pago, larger 84 nrni. Six from cove just south of Aua village, largest 75 mm. Four from coral reef in front of hospital, Pago Pago. These agree with Giinther's figure,' except that he does not show the supraorbital cirrhus. Sebastopsis scabra (Ramsay and Ogilby) is said to differ in its longer anal spine, though this is no longer than in our examples if Jordan and Scale's* figure is correctly identified. S. parvipinnis (Garrett) is alleged to differ in its minute dermal flaps and rather low, uniform dorsal, both characters possibly due to variation. Sebastapistes laotale Jordan and Seale. One 53 mm. from coral in reef in front of hospital, Pago Pago. Two, 64 and 46 mm., from cove just south of Aua village. Also another, same locality, screened at the bottom. GOBIESOCID^. Crepidogaster samoensis Steindarhner. Two small examples from cove just south of Aua village. GOBIIDJE. Eviota zonura Jordan and Seale. Two small specimens, same locality as last. » BuU. U. S. Fish. Comm., XXIII. 1903 (1905), p. 393, fig. 171. > L. c, p. 395, fig. 172. » Jour. Mu8. Godeffroy, IV, 1875, pi. 76, fig. c. « Bull. Bur. Fisher., XXV, 1905, p. 375. fig. 71. A Collection of Fishes from Samoa. 125 Eviota afelei Jordan and Seale. Six from coral in reef in front of hospital, Pago Pago; cove just south of Aua village; tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Probably E. smaragdus Jordan and Seale is identical. Eviota distigma Jordan and Seale. Two from cove just south of Aua village. Small black spots, two in number, on each pectoral base distinctive. Two from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Pseudogobiodon citrinus (Riippell). Twenty examples from cove just south of Aua village. Variably pale or dark. Some with first dorsal olive, border bright orange, edge narrowly black. Second dorsal olive with yellow border, narrowly edged blackish in some examples; others show pectorals with yellowish tints. The darker examples mostly uniform slaty and without the brilliant borders to the fins. Eleven examples, same locality, screened at the bottom, are variably light and dark, some yellowish, others with orange-bordered dorsal. Three from coral in reef in front of hospital, Pago Pago Harbor, and one from Pago Pago. BLENKIID^. Enneapterygius tusitalas Jordan and Seale. One from cove just south of Aua village and 5 from tide-pools near Double Point, just west of entrance to Pago Pago Harbor. Salarias variolosus Valenciennes. One example from cove just south of Aua village. Jordan and Evermann^ say : "The fish figured and described by Giinther in Fische der Siidsee as Salarias variolosus from Tahiti^ is a different species." It is also further inferred that the specimen in the Academy of Natural Sciences of Philadelphia from the "Sandwich Islands," collected by Thomas Nuttall, is not identical with Giinther's fish, A comparison of Nuttall's specimen with our Samoan leaves no doubt as to their identity. Salarias gibbifrons Quoy and Gaimard. One from cove just south of Aua village. Alticus biseriatus (Valenciennes). (Fig. 2). Head 3.33 to 4.12; depth 3.66 to 5; D. XIII, 18 to 20; A. 22 to 24; P. 15; V. 2; head width 1.5 its length; head depth (without crest) 1.4; eye 2.5 to 3.5; mouth width 2; first dorsal spine 2; fifth dorsal ray 1.33; fourth anal ray 2.33; least depth of caudal peduncle 2.33; caudal 1; pectoral 1.25; ventral 2. Body elongate, slender, tapers back gradually and evenly from head to caudal peduncle. Latter not free, compressed. Head small, robust, cheeks and lower sides little swollen. Snout very obtuse, front profile vertical and slightly convex, breadth opposite front of eyes about equals twice its length to upper jaw en"d medially. Eye large, antero-lateral, moderately elevated, posterior edge near first third in head. Mouth broad, with short gape, inferior, so front of lower jaw about opposite eye center. Each side of lower jaw with large posterior canine. Teeth minute otherwise, very close-set, pointed, in single narrow flexible row. Interorbital narrow, not one-third of eye, level. Anterior nostril level with and close before lower eye edge and with short fleshy tentacle about one-third of eye. Posterior nostril close above anterior or lower nostril, also nearer eye, simple pore. » Bull. U. S. F. Com., XXIII (pt. 1), 1903 (1905), p. 498. ' Jour. Mus. Godeffroy, VI (Heft, xi), 1877. p. 203, pi. 116, fig. a. 126 A Collection of Fishes from Samoa. Gill-opening forms free fold across broad isthmus. Rakers at least a dozen short points, about one-third of filaments and latter one-third of eye. Body covered with smooth skin. Head with median cutaneous keel or crest in largest individual only, arises opposite posterior edge of pupil and not quite to dorsal origin, its length 1.6 in head. The smaller individuals show a pair of short nuchal tentacles. Long, pointed, fleshy flap above eye 1.5 eye diameters, each edge with several (3 to 5) short tentacles. Row of pores behind eye along suborbitals, another down preopercle, third one on mandible. Spinous dorsal origin slightly in front of gill-opening edge, only last few spines graduated down short, as deep notch before soft dorsal, and fin largely uniform with entire edge. Soft dorsal with entire edge, large, uniform in height, last ray joined by membrane to upper caudal peduncle edge, but not to caudal fin. Anal origin about opposite tenth dorsal spine base, fin uniform in height, free from caudal peduncle behind, membrane behind each ray tip notched. Caudal rounded. Lower median pectoral raj^s longest, fin almost reaches anal. Ventral inserted slightly in front of pectoral origin, inner ray slightly longer, halfway to anal origin. Color in alcohol dull lavender-brown generally on back, lower and under surface pale or whitish. Trunk with a dozen dusky-brown vertical blotches somewhat arranged as if in pairs, joined above alternately on back with small, dusky, vertical blotches extending up somewhat on bases of dorsal fin. On head dusky clouded blotch behind eye, one on opercle, one on cheek below, one from edge of lower eye to mouth, and one on snout. Number of indistinct, small brownish ocelli, much less than pupil, on head. Crest dusky, mottled paler. Trunk mottled with scattered paler dots and obscure marblings. Fins largely grayish, spinous dorsal with membranes medianly anteriorly, or until seventh, with jet-black round blotch little less than eye, and on rest of fin gradually become pale dusky behind. Soft dorsal, except base, with many even oblique pale gray-blue lines, up and backward. Caudal with dark blotch little less tiian eye at bases of median rays, lower and submarginal part of fin dusky. Anal with long submarginal dusky band. Dark obscure transverse streak at pectoral base. One specimen, 50 mm. in length, from tide-pools near Double Point, just west of entrance to Pago Pago Harbor, March 20, 1917, and three specimens, 21 to 30 mm., same locality. Salaries rivulatus Ri'ippell. Eleven dark examples, largest 98 mm. Pago Pago. Enchelyurus ater (Giinther). Two from cove just south of Aua village. Jordan and Soale state that "Giinther describes the ventrals as reaching the anal, but in his figure the fins are much shorter." This is probably a variation with age. FIERASFERIDJE. Jordanicus parvipinnis (Kaup). One, same locality as last. VIII. LEODICID^ FROM FIJI AND SAMOA. By a. L. TREADWELL, Professor of Zoology in Vassar College. Eight plates and 68 text-figures. 127 CONTENTS. Page Introduction 129 Systematic descriptions 130 Arabella Grube 160 Arabella dubia Treadwell 160 Dorvillea Parfitt 166 Dorvillea aiistraliensis Mcintosh 166 Dovilleinae 166 Drilonereia Claparede 161 lumbricus Treadwell 161 paucidentata Treadwell 162 Leodice Savigny 130 Leodice aciculata Treadwell 143 antennata Savigny 136 aphroditois Pallas 134 armillata Treadwell 144 biformi-cirrata, Treadwell 148 coccinea Grube 142 crassi-tentaculata Treadwell 146 flava-punctata Treadwell 136 gracili-cirrata Treadwell 149 Buviensis Treadwell 138 tubicola Treadwell 139 viridis Gray 131 Page Leodice viridia Gray, var. vernalis Tread- well 133 Leodicidse 130 Leodicinse 130 Lumbrinereinse 157 Lumbrinereis de Blainville 157 Lumbrinereis brevicirrata Schmarda 158 japonica v. Marenzeller. . . . 159 sphisrocephala Schmarda. . . 158 Lysidice Savigny 154 Lysidice fusca Treadwell 154 parva Treadwell 155 Marphyaa Savigny 150 Marphysa calif ornica Moore 150 macintoshi Crossland 151 simplex Treadwell 151 Nicidion Kinberg 156 Nicidion fusca-fasciata Treadwell 156 Oenone Savigny 163 Oenone fulgida, Savigny 163 Onuphis Audouin et Milne Edwards 154 Onuphis holobranchiata v. Marenzeller. 154 Paramarphysa Ehlers 153 Paramarphysa teres Treadwell 153 128 LEODICIDAE FROM FIJI AND SAMOA. By a. L. Treadwell. INTRODUCTION. The annelids here described were collected by the writer in Fiji and Samoa in April, May, and June of 1920, on an expedition con- ducted by the Department of Marine Biology of the Carnegie Institu- tion of Washington, Dr. A. G. Mayor, Director. Owing to our short stay in Fiji, collecting there was limited to the immediate vicinity of Suva and was mainly done on the reefs on either side of the main entrance to the harbor and on the mud flats near the city front. Pt. MATATUUA •^VAlTOQI '"SAIL KOCK 'Steps Pt. T^ R 1 N c I N c. 1* e c F Map of Tutuila, American Samoa. In Samoa, a very thorough survey was made of the harbor of Pago Pago in Tutuila, so that so far as the Leodicidae are concerned this work can claim to be exhaustive for that locality. Only a few annelids from other families were collected and none equaled the Leodicidae in number of species or abundance. The Leodicidae occur especially in the reef rocks and are not represented in the mud of the upper harbor. Apparently this is due to the large quantities of fresh water which flow into the upper end of the harbor, making a brackish-water condition in which only a few Terebellids and Capitellidids can live. I found a similar condition at Fagaalu and 129 130 Leodicidce from Fiji and Samoa. on the opposite side of the island in the harbor of Leone. Collections were also made in the lagoon southwest of Nuuli and in the reef west of the island of Aunuu. The specimens were narcotized in a solution of MgSOi, 153 grams to the liter, killed in 10 per cent formalin, and preserved in strong alcohol. SYSTEMATIC DESCRIPTIONS. In an earlier paper (Treadwell, 1921a) I have discussed the anatomical features which are of most importance in the taxonomy of the Leodicidae, and later study has led to no conclusions dififerent from those there stated, unless it is to further emphasize the im- portance of the jaw in classification. I found the general character of the jaw remarkably constant in any one species, especially in the form of the plates and their color. MultipUcation of genera and subgenera on the part of taxonomists is a very unfortunate habit, but there would be some justification for making a subgenus of Leodice to include L. siciliensis Grube, L. caribcea Grube, L. paloloides Moore, L. viridis Gray, L. viridis var. vernalis Treadwell (see page 123), and L. duUa Woodworth, on the basis of their jaw structure. In the relatively large size of the mandible, the Umited tooth development on the proximal plates, and the pecuHar appearance of the distal ones, these differ from all others that I have seen. Students of this group have not agreed on the major classification of the animals included in it. In an earlier paper (Treadwell, 1921a) I followed what seemed to be the majority opinion and classified the Leodicidse as a family, with the subfamilies Leodicinse, Lumbrinereinse, and Stauronereinse. Chamberlin (1919a), in a paper which appeared after mine had gone to the editor, constructs the superfamily Leodicoidea, putting under it the families Leodicidae, Lumbrinereidse, Onuphididse, and Dorvilleidse ( = Stauronereidae, see page 166). In the present paper I shall follow my original arrange- ment. Family LEODICID-S. Annelida varying much in size in different species, with or without prostomial tentacles, nuchal cirri, eyes, and parapodial gills. Notopodium of parapodium rudi- mentary or apparently absent. Jaw of maxilla and mandible, the former of two or more rows of plates, mostly toothed. Subfamily LEODICIN.ff:. With dorsal and ventral parapodial cirri, with or without nuchal cirri and para- podial gills. Prostomium with from 1 to 7 tentacles and one pair of palps more or less fused with the prostomium. Genus LEODICE Savigny. Savigny, J. C, Systeme des Annelides 1930, p. 13. Prostomium 2- or 4-lobed, the lobing often obscure. With 5 tentacles and 1 pair of eyes. A pair of nuchal cirri rtn the second l)ody-somite. Parapodia begin on the Leodiddce from Fiji and Samoa, 131 third body-somite. Gills on more or fewer of the body somites. Jaw of maxilla and mandible, the former of forceps, 2 pairs of toothed plates, and 1 unpaired plate. Man- dible of symmetrical halves joined anteriorly to form a cutting edge. One or two pairs of anal cirri. Leodice viridis Gray. Plate 1, figures 1 to 7; text-figures 1 and 2. Palolo viridis Gray, Stair, 1847, pp. 17-18. Paloh viridis Macdonald, 1858, pp. 237-239, pi. 41. Lysidice palolo Quatrefages, 1865, p. 379. Lysidice mridis Ehlers, 1864-1868, p. 367, pi. 16, figs. 17, 18. Lysidice viridis Collin, 1897, pp. 164-174. (Reprint pp. 1-11.) Pahlowurm Friedlander, 1898, pp. 337-357. Eunice viridis Ehlers, 1898, pp. 1-16. Eunice viridis Woodworth, 1907, pp. 3-21, pis. 1-3. Fully mature specimens were not found, and no measurements can be given of the completely grown individuals. One specimen measured, after preservation, 270 mm. in length and contained about 450 somites. The prostomium (plate 1, fig. 1) is noticeably 2-lobed, and when expanded is a trifle wider than the peristomium. Dorsally it is colored a yellowish brown with many minute yellow spots, the anterior margins and the ventral surface being coloriess. The tentacles are colorless, blunt-pointed, and more or less wrinkled, the median one reaching as far as the anterior border of the fifth somite, the inner paired to the middle of the third, the outer paired to the second. Ehlers (1898, p. 5) is in error in describin g the tentacles and cirri as jointed. The eyes are large. The peristomium is a little longer than the prostomium ; its anterior margin bends around on either side so as more or less to inclose the bases of the tentacles. It is colored much like the prostomium, the pigment extending over the lateral faces but leaving the ventral surface uncolored. The second somite is colored much like the first, but on its anterior border has an uncolored band a little wider than the bases of the nuchal cirri. The nuchal cirri are without color and extend about as far as to the anterior border of the prostomium. The color is continued as far as the region of somites 18 to 20, but the anterior colorless band in each somite becomes successively broader, so that behind the region of somite 20 the only body-color is that given to it by the contents of the intestine or, in the posterior portion, the color of the sex organs. In very young animals, where the sex products have not formed in sufficient quantities to produce a color, this posterior region is colorless. In the posterior third of the body, except for the extreme posterior end, each somite has a median ventral black spot. Woodworth (1907, plate 1) gives colored figures of the species. I was unable, owing to the time of my visit to Samoa, to collect the epitokous ends, but Woodworth's figure 3 corresponds very well with my observations. There are two pairs of anal cirri, the ventral ones much the smaller (plate 1, fig. 2). Throughout the median region the gills appear as single filaments larger than the dorsal cirri, to whose bases they are attached (plate 1, fig. 3), and in life are a bright red color. In a specimen 250 mm. long the gills first appeared at the region of somite 137 and continued to about 180 somites from the pygidium. The first and last of the series are the smallest and are not very prominent, but through the middle region, where the gills are largest, they are prominent because of their color. The first parapodium has very prominent dorsal and ventral cirri and a very small setal lobe, the latter with vertical, parallel, anterior and ventral lips. Two acicul*. one much darker than the other, extend into the setal lobe, and there is a small tuft of setae containing both simple and compound forms. On subsequent parapodia there is a relatively great increase in size of the setal portion, and at the same time a shifting of parapodial position, so that they come to lie higher on the lateral face. The first parapodia, because of their ventral position, are partly hidden from a dorsal view 132 Leodiddce from Fiji and Samoa. and appear to be no larger than the others. The tenth parapodium (plate 1, fig. 4) has a prominent setal lobe with its anterior lip asymmetrically bifid and with a rounded posterior lip. There is a single very large black acicula. Dorsally in the seta tuft are a few long, slender, simple setse and ventral to these a dense tuft of compound ones with heavy basal portions and relatively short but stout terminal joints. The dorsal cirrus is Very slender, the ventral one short, carried on the end of a pad-like swelling. I could find no trace of needle aciculse. Throughout the anterior region of the body tliis ventral pad-like swelling, which appears at the region of somite 10, is continued and the pads on the two sides of each somite, together with the flattened ventral surface, make up a sole-like ventral region which is in marked contrast to the rounded dorsal region. The pads disappear in the gilled somites, but the flattened ventral surface persists. A gilled parapodium (plate 1 , fig. 3) has a pointed setal region with a single large acicula. The gill is attached to the base of the dorsal cirrus and when fully developed is much larger than the cirrus. As is well known, this species develops at the approach of the breeding-season a posterior epitokous region, and consequently the form of the posterior end depends on the degree of development of the epitokous portion. The swarming occurs in October and November, and my collections were made in April, May, and June, so that the epitokous modifications had appeared only to a limited extent. According to figures given by Friedlander (1898, p. 344) and Woodworth (1907, plate 2, fig. 10), the epitokous portion is much narrower than the atokous, as if a slirinking in diameter occurs at this time. Tliis is contrary to the conditions found in Leodice fucata Ehlers (the Atlantic Palolo, in which a swarming occurs), to L. paloloides Moore, to L. cariboea Grube, and to L. viridis var. vernalis (see page 133), where, when the sex products are formed, the posterior region is much broader than the anterior. Swarming has never been observed in these latter species, but (except for this question of absolute width) the structural modifications are quite as they are in the true Palolo. I am indebted to Lieut. Commander R. C. Reed for specmiens of swarming ends collected in Tutuila, Samoa, in November 1920. In these the setse appeared to be longer than in the atokous phase, but careful measurements showed that the absolute length is the same, though because of the narrowing of the body diameter, they extend to a greater distance from the surface. A parapodium from the epitokous region is shown in plate 1, figure 5. The setal lobe is pointed and has a large conical ventral cirrus attached near the end of the lobe. The dorsal cirrus is attached much farther back from the apex, and while conical is much narrower than the ventral one. The aciculse are all of one kind, straight and bluntly rounded at the apex. They may be nearly colorless, as happens in the case of the smaller ones, or contain brown pigment, which may be very dense but never appears black. The setse are of two kinds, simple and compound, and are similar in form tliroughout the body. The simple ones (text-fig. 2) are long and sharp-pointed, minutely denticulated along one edge. The compound ones (text-fig. 1) have relatively heavy basal joints which are dentic- ulated at their apices, the terminal joints small, with equal-sized apical and subapical teeth covered by a hood which is minutely serrated along its border. The maxillae (plate 1, fig. 6) are dark, with the carrier much hghter than the re- mainder. The two halves of the carrier are closely united throughout most of their extent, the basal ends rounded and relatively broad. The forceps has a heavy basal portion narrowing very abruptly to form the fang at about the middle of the plate. The proximal plates are large, extending back to the carriers; the left one has 3, the right one has 2, indistinctly marked-off teeth. The forceps and proximal plates are very dark in color, with a whitish incrustation along the cutting edges. Distally are 2 plates on the right side and 3 on the left. These are very irregular in outline and their apparent form and size depends on the position from which they are viewed, more than Ls the case in the majority of Leodicid maxillie. Their general appearance is shown in figure 6 of plate 1. As compared with the maxilla, the mandible is very large (plate 1, fig. 7) and its lateral margins are much rolled. To the naked eye or under Leodicidce froin Fiji and Fiamoa. 133 very low magnification the mandible is an intense white and, as in the case of L. caribaa (Treadwell, 1921, p. 49), when protruded from the mouth, forms an easily recognized diagnostic character of the species. Under higher magnification the center shows dark. Lcodice viridis belongs to the group of the Leodicidse of which L. siciliensis is a representative species, all distinguished by the small development of the gills, the peculiar jaw apparatus and in most species by the formation of an epitokous posterior end possibly in all cases connected with a swarming. Swarming has actually been seen only in L. viridis and L. duhia (Woodworth, 1907), but it seems possible that it occurs in the other species as well. (See Treadwell, 1921, p. 47). The only other known case of swarming among the Leodicidse is that of L. fucata, which occurs gener- ally in the West Indian region, where it has been reported on by Mayor (1902) and by Treadwell (1921a, p. 43-47, pi. 4, figs. 5 to 10, text-figs. 127 to 135). L. fucata is not, however, a member of the siciliensis group. As my collecting was done some months earlier than the swarming period, I was unable to make any observations on this phenomenon and can add nothing to the literature, which is well summarized by Woodworth (1907). In collecting at the spring tides of April, May, and June successively, where low water made it possible to get near the outer edges of the reefs, I found indications of a gradual change toward the epitokous condition in the change of color due to the developing se.x products, but these changes were comparatively slight. Dr. Mayor very kindly collected the species in July and reported that there was little change in color from the June condition. My collections were all made on the reefs in and near Pago Pago Harbor in Tutuila, Samoa. The animals were to be found in rocks at all distances from the shore, but were larger and evidently more mature the farther from shore they were collected, my largest specimens being obtained at as near the edge of the reef as it was possible to go. This led to the suggestion (Treadwell, 1921, pp. 199, 200) that the rate of development may depend on the environment and that those living near shore find the conditions so unfavorable that they grow very slowly and possibly never mature. In all respects except size, these resemble those from farther out, so that there is no question as to the identity of species. I made careful studies of those localities where both the native Samoans and residents at the U. S. Naval Station told me the swarm is most numerous in October and November, but found no place where they are as abundant in the rocks as they should be to supply the enormous number of epitokous ends which appear at the swarming. It seems to me probable that the largest indi- viduals and the greatest number of individuals are to be looked for on the edges of the reefs, where, on account of the surf, I was unable to collect. Leodice viridis var. vernalis, new variety. Plate 1, figures 8-11. A considerable number of a small Leodice were collected in Suva Harbor, Fiji. Many were in the epitokous condition, with bright green eggs in the posterior part of the body. These belong to the siciliensis group and I at first took them for the Palolo, though puzzled by their sexual condition. As this was in AprU, and the first swarming would be in October, the mature condition of the eggs was hard to under- stand. Later the true Palolo was collected in Samoa, but it was not until a more careful examination of my collections was made after my return from the expedition that I detected the differences between the species and the variety. (In my report, (19216, pp. 199, 200) I erroneously confused the two.) The variety does not appear in my Samoan collections, but as I was intent on collecting the largest individuals, I may have passed it over on the assumption that it was the young of the true Palolo. I do not, however, think that it occurs in Samoa, as if it had been there and as fully mature as were the Fijian forms I could hardly have failed to notice it. The living animal has at the anterior end an intense greenish-brown color with much iridescence which is continued with a gradual diminution in intensity to the 134 LeodicidcB jrom Fiji and Saynoa. region of somite 50. The prostomium is broader than tlie peristomium (plate 1, fig. 8) and decidedly 2-lobed, its dorsal surface dark in color, the anterior margin and regions lateral to the eyes being colorless. The tentacles are colorless, the median one 4 to 5 times as long as the prostomium and pointed at the apex. The tentacles are more sharply pointed and have longer cirrophores than in L. viridis. Preserved material retains the coloration of the anterior end, so that for about the first 40 somites both dorsal and ventral surfaces are dark brown. The anterior somites do not have the colorless band on their anterior borders which are present in L. viridis (compare fig. 1 and fig. 8, plate 1). The parapodia are uncolored, as are the nuchal cirri. In the epitokous portion there is on either side of the dorsal surface in each somite a dark spot at the base of the parapodium (plate l,fig. 11). Apparently these spots do not extend to the very posterior end, but I could not determine this with certainty in the material at my disposal. They may also be found on a few of the posterior atokous somites. A parapodium from setigerous somite 10 is shown in plate 1, fig. 10. The prominent pad-like swelling which carries the ventral cirrus begins at about this region and extends for about the first quarter of the length of the animal. The gills have about the same arrangement that they have in L. viridis, but are more slender, extend into the epitokous region, and are relatively more promi- nent. A small L. viridis may be distinguished from one of the variety of the same size by the fact that the gills in viridis would be much smaller than in the variety. An epitokous parapodium (plate 1, fig. 11) has a small dorsal cirrus with the long gill attached near its base. There are two pairs of anal cirri quite similar to those of L. viridis, the jaws, except for size, are exactly like those of the species, and the setae and aciculse are similar to those of the species. The type is in the American Museum of Natural History. Leodice aphroditois Pallas. Plate 1, figures 12 to 17; text-figures 3 to 7. Nereis aphroditois Pallas, 1788, p. 229, pi. 5, figs. 1-7. Eunice aphroditois Ehlers, 1864-68, p. 306, pi. 15, figs. 23-29. Eunice aphroditois Mcintosh, 1885, p. 282, pi. 38, figs. 16, 17; pi. 20a, figs. 8-10. Eunice aphroditois Crossland, 1904, p. 288. Eunice aphroditois Augener, 1913, pp. 267-270. Eunice aphroditois Fauvel, 1917, pp. 215-220, pis. 7, 8. References to a considerable literature concerning this species and a related or more probably identical form L. kinbergii wdll be found in the works cited above. As the material at my disposal did not enable me to attempt this question of the synonymy, it does not seem necessary to undertake any discussion of the subject. Two specimens were collected at Pago Pago, Samoa, the smaller one 220 mm. long, with a prostomial width of 6 mm. and a larger one 450 mm. long, for the part which was preserved, a considerable portion of the posterior end having been lost. The prostomial width is 8 mm. Text-Figures 1 to 7. 1 and 2. Leodice viridis. 1, com- pound seta X 260; 2, simple seta X 220. 3 to 7. Leodice aphroditois. 3, com- pound seta X 220; 4, pectinate seta X 220; 5, detail of simple seta X 220; 6, acicula X 220; 7, ventral acicula X 220. Leodiddxx jrom Fiji and Samoa. 135 The smaller individual in life was very dark brown, almost black. The prosto- mium was lighter brown in color, the peristomium greenish brown, with longitudinal black markings, and very iridescent. The tentacles were greenish brown, with white bands not regularly arranged and not uniform on the different tentacles. The re- mainder of the body was dark purplish-brown, becoming purple at the pygidium. There is one pair of stout purple anal cirri uncolored at their apices (plate 1, fig. 13). The larger individual was quite uniformly dark brown, with a greenish tint at the auterior end. The tentacles were faint green, with a darker tip, as were the dorsal cirri. The cirri were not banded. The fourth and fifth setigerous somites were lighter in color than the others, but did not show a distinct "collar." In alcohol, both specimens are kwvn, though the larger is the darker; and more color is retained in the tentacles. Both show numerous purple lines and streaks running longitudinally on the dorsal surface. The prostomium (plate 1, fig. 12) is wider than the peristomium and very noticeably 4-lobed. The tentacles are all of about the same length, about as long as the peri- stomium. The latter is slightly wider anteriorly than posteriorly, with straight margins, and is as long as the following 5 somites. Somite 2 is very short and the nuchal cirri are shorter than the peristomium. The first parapodium has the usual form, with a large dorsal and a smaller ventral cirrus, with a very small setal lobe. Two good-sized aciculaj extend into the dorsal cirrus, and there is one in the setal lobe. The tenth parapodium (plate 1, fig. 14) has a gill of 16 branches, a very hea-s^ dorsal cirrus, which is longer than the gill, a small acicula in the dorsal cirrus, and a very large one in the setal portion. A para- podium from the posterior end of the body (plate 1, fig. 15). still shows the relatively very large dorsal cirrus and, in addition to the acicular equipment of the anterior ones, there is a ventral hooked acicula. In the smaller specimen the gills begin on the sixth setigerous somite (the entire body having 180) and extend through about 120 somites. In the larger specimen they arise on the fifth somite on the right and sixth on the left, but as the posterior end is lost I am unable to give their extent. The basal portion of the gill is thick, giving it a heavy appearance during life, but the branches are relatively small and short. The compound setae (text-fig. 3) are stout with a heaA^ shaft and a 2-hooked terminal joint. In the anterior somites they are arranged in a formidable vertical row, but diminish in number in the posterior. somites. The pectinate setse (text-fig. 4) are small and slender, with about 10 teeth on the margin. The simple setae are very long and slender, with a slender basal portion, widening slightly just outside the para- podial margin, and beyond this tapering gradually to an acute point. Along orie edge is a marginal wing, having very fine denticulations. A detail of the shaft is shown in text-figure 5. The dorsal aciculse are blunt-pointed (text-fig. 6), the ventral ones 2-hooked, with an uncolored apex and a very dark shaft (text-fig. 7). The maxillae (plate 1, fig. 16) have a short carrier, with long, slender forceps. The proximal paired plates have each 5 teeth, the right paired with 9, the left with 3, the unpaired with 6. All plates of the maxilla are very black. The shafts of the mandibles are slender, but widen decidedly toward the cutting edge, and are very black in color. The beveled portion is covered with a white incrustation (plate 1, fig. 17). Although this species has received much attention, it seems worth while to add the above description, because, wliile I have no doubt as to the accuracy of the identifica- tion, the various descriptions which have been written vary so much from each other and from the specimens from Pago Pago that this must be a very variable species, and it seems desirable to record as far as possible these variations. Ehlers's figures are not very satisfactory, especially of the jaws, and the figure he gives of the simple seta shows much more of a broadening in the shaft than I have seen. He gives two figures of pectinate setae, with 7 teeth in one and 20 in the other. He figures no ventral acicula. He states that the gills are longer than the dorsal cirrus, which is not true in 136 LeodicidcE from Fiji mid Samoa. my specimens. Mcintosh figures the nuchal cirri as much shorter than in Ehlers's or in mine, but his figure of the compound seta agrees with that of mine. Augener records a specimen of this species from Samoa, in the collections of the Gbttingen Museum, but gives no description of it. Fauvel describes the gills as longer than the dorsal cirrus, while Crossland states that in spite of the large number of their branches, the gills reallj' cover only a small portion of the dorsal surface. This agrees with my specimens from Pago Pago. Leodice antennata Savigny. Eunice antennata Crossland, 1904, pp. 312-318, pi. 22, figs. 1-7, text-figs. 56-60. Eunice antennata Augener, 1913, pp. 270-274. Eunice antennata Fauvel, 1917, pp. 225-228, text-figs. 20o, 206. Eunice antennata Fauvel, 1919, p. 377. Figures are given by Crossland, and discussions of the possible synonymy of the species \A\\ be found in the first three of the above references. Two individuals were collected in Pago Pago Harbor, in rock near the landing in front of Cook's Hotel. In life they are brownish green in color, but rather translucent, so that the contained blood modifies the tint very decidedly. The dorsal surface of the prostomium is uncolored except for a purple band around the base of the median tentacle and a similar one around the bases of the inner paired tentacles. From each of these latter a band runs toward the median line, uniting with a broader greenish band wliich runs toward the anterior margin. On either side of this stripe is a colorless spot. The tentacles and all cirri are articulated, and on the tentacles and anal cirri, but not on the dorsal, are brown bands in the constrictions. From the eighth somite posteriorly a black spot is present in each somite near the dorso-lateral margin, and the smaller of the two shows traces of dorsal white spots toward its posterior end. None of the color remains in the preserved material. The animals are much more active than is usual in this genus, squirming much as does Nereis when captured. Crossland (pp. 313, 314) mentions the green color as an occasional variation, possibly in relation to environment, and also comments on the activity of the animals when handled. The larger of the two specimens is 65 mm. long, and has a peristomial width of 1.5 nun. The other individual is about one-third smaller. The larger one contains immature eggs, so m.ust be adult. The first gill, of 2 branches, is on the fourth seti- gerous somite. The number of branches rises to 6 in the region of somites 15 to 20, through the middle of the body it drops to 2, and at the extreme posterior end of the body rises again to 4. Only the last 2 or 3 somites are free from gills. The jaws are very delicate, only their margins colored. The proximal paired plates have 6 teeth on the left and 8 on the right, the distal paired plates have 10 on the left and 8 on the right, and the impaired has 9. The distinguishing features of this species are the articulated tentacles and cirri, the median tentacle being long (in the larger of my two specimens it reached somite 8) ; the peculiar arrangement of gills whereby the number of filaments decreases throughout the middle of the body to increase again at the posterior end; and the fact that the ventral acicula has a trifid apex. Crossland's text-figure 60 (p. 317) shows this arrange- ment of gills, but the dorsal cirri are represented as non-articulated. As this is not in agreement with figures 1 and 7 of his plate 22, it is probably an error in the drawing. Fauvel (1919, p. 378) says that the tridentate aciculse are rare in specimens from Mada- gascar, but thej' are mentioned as distinctive in his Australian specimens (Fauvel, 1917, p. 226, figs. 20a, 206), and they are present in my Samoan material. Leodice flava-punctata, new species. Plate 2, figures 1 to 7; text-figures 8 to 11. Several specunens, none entire, were collected in Pago Pago Harbor, Samoa. The general appearance of the animals would indicate that they are immature, but the fact that several contain eggs indicates that they are adults. One individual Leodicidw from Fiji and Samoa. 137 (in 3 pieces) measures altogether 50 mm. in length, has a prostomial width of \.lh mm., and contains about 250 somites. The specific name is given because of the yellow spotting which is common in Leodice, but the spots are unusually prominent in this species. The general body-color of the anterior end is dark brown, darker on the prostomiura than elsewhere and gradually lightening posteriorly, the color disappearing entirely in the region of from somites 40 to 60. In the peristomium the color is continued on to the ventral surface, but in all other parts it is limited to the dorsal. Scattered over the surface of this browTi are numerous yellow dots. In the type a prominent un- colored band extends along the median dorsal line of the prostomium and is continued on to somite 2 (plate 2, fig. 1). Somites .3, 4, and 5 have each a prominent uncolored spot toward the posterior dorsal border and somite 6 is uncolored. In other individuals the.se colorless spots are either only faintly indicated or are absent. Toward the posterior end of the body each somite is marked by a very narrow purple line across its poste- rior dorsal margin. One specimen was regenerating a pygidium which had one pair of short, colorless anal cirri. The tentacles are light brown except for the apices, which are uncolored, as are the nuchal and other cirri. The prostomium (plate 2, fig. 1) is noticeably bifid, and when expanded is a little wider than the peristomium. The latter is a trifle wider than long and slightlj' concave along the lateral mar- gins. Somite 2 is about one-third as long as 1 and sharply marked off from it. The tentacles are shorter than the peristomium, with inconspicuous cirrophores, and the eyes are so sur- rounded by pigment as to be scarcely visible. The nuchal cin-i are slender and much shorter than the peristomium. The first parapodium (plate 2, fig, 2) has a bilobed setal portion, with heavj' cirri, the ventral cirrus being especially large. There is a single acicula, also eleventh parapodium (plate 2, fig. 3) Text-Fiqures 8 TO 16. 8 to 11. Leodice flava-punctata. 8, simple seta X 285; 9, compound seta X 285; 10. ventral acicula X 285; 11, dorsal acicula X 250. 12 to 16. Leodice suviensis. 12, simple seta X 250; 13, compound seta X 250; 14, pectinate seta X 250; 15, dorsal acicula X 68; 16, ventral acicula X 68 needle aciculse in the dorsal cirrus. The has a very prominent setal portion, with a dense tuft of compound setae ventrally and a smaller tuft of simple setse dorsally. There is a needle acicula in the dorsal cirrus. The dorsal cirrus is slender, but the ventral one is short and thick and merges gradually into the ventral pad-like swelling characteristic of the anterior parapodia in this genus, but especially prom- inent here. A later parapodium from behind the middle of the body (plate 2, fig. 4) has a conical setal portion with a prominent ventral swelling, carrying the short, thick ventral cirrus on its outer end. The dorsal cirrus is very slender, and the gill arises from the body-wall dorsal to its base. There is a single pointed acicula in the middle of the setal lobe, and a hooked one near its ventral surface. A few small needle aciculjE extend into the dorsal curus. In a small specimen the gills arise as a single filament on the thirteenth setigerous somite and in the great majority of later somites there are two and three branches 188 Leodiddce from Fiji and Samoa. (plate 2, fig, 5), with an increase posteriorly of the number of two-branched gills, but at about somite 40 the number is reduced to one. About 100 of the posterior somites of the specimen are without gULs, but I have no information as to the original number of somites in the entire animal. On the right side of somite 16 the gill has 7 branches, and there are 6 on the left of somite 21. This large number of branches, limited to only a single somite, is a very unusual condition. In an individual twice the size of the one just described I could find no gill with more than 6 filaments. The filaments remain relatively large to the end of the series and the blood-vessel in each is especially prominent. The simple setse (text-fig. 8) are very slender, only slightly broadened toward the end and taper to acute apices with very minute denticulations along one border. The compound setae have small terminal joints, the subapical tooth the larger, and with small denticulations along the end of the basal portion (text-fig. 9). Figure 9 is drawn from a seta from the anterior end of the body. In the posterior region the compound seta? have longer terminal joints. The pectinate setse have about 20 slender teeth, the one at one end of the row longer than the others. The dorsal aciculse (text-fig. 10) are bluntly rounded at the apex and dark-colored to the very tip. The ventral ones have the tip uncolored, with bluntly rounded teeth covered by a hood (text-fig. 11). The forceps and margins of all plates of the maxilla are dark brown, while the re- maining portions are much lighter. The carriers (plate 2, fig. 7) are short, the forceps long and much curved. This curvature is not adequately represented in the figure, for the forceps are drawn as pointing upward. The proximal paired plates have 5 teeth on the left and 4 on the right; the distal paired have 9 on the right and 5 on the left, 2 of these being much smaller than the others. The unpaired has 8. Beyond the paired plates are rounded pigment patches in the chitin. The mandibles (plate 2, fig. 6) are rather small and slender, the beveled portion being marked with pigment on the outer and inner margins and with concentric Unes on the surface. In structure of gill this species resembles ChamberUn's L. lita (1919o, pp. 240-244, pi. 54, figs. 6-10; pi. 55, figs. 1-7), but in general body-coloration, form of the peri- stomium, and character of jaws the two are unlike. The type is in the American Museum of Natural History. Leodice suviensis, new species. Plate 2, figures 8 to 13; text-figures 12 to 16. A single specimen, collected in rock exposed at low tide on the west side of Rat Passage, in Suva Harbor, Fiji. It measures after preservation 370 mm. in length, has a prostomial width of 4 mm., and at somites 9 and 10 is 9 mm. wide. To the naked eye the Uving animal appears very dark, nearly black, while under a hand lens the color is seen to be dark purple, with numerous dirty-white spots over the surface. The tentacles are dark green, uncolored at the tips. All of the cirri have uncolored tips. In preserved material the color is a dark brown, with numerous yellow spots over the surface and a considerable iridescence. The prostomium is rather small, 2-lobed, the tentacles smooth, short, and tapered gradually toward the apices. The unpaired tentacle extends as far as the anterior border of somite 3, the inner paired tentacles are about three-quarters as long as the median, the outer paired about half as long. The outer paired have their bases of attachment noticeably farther forward than the inner ones. The eyes are in the usual position (plate 2, fig. 8). The peristomial width is to its dorso-median length about as 5 to 3, and its antero- lateral border has a prominent lip on either side. Somite 2 is about one-quarter as long as somite 1, the nuchal cirri extending to about the middle of somite 1. The ventral surface of the anterior region of the body is a little lighter in tint than the dorsal, but otherwise is colored like the dorsal. In the type the pygidium is apparently LeodicidoB from Fiji and Samoa. 139 regenerating; it has one pair of very short anal cirri colored like the anterior cirri and tentacles. The gills arise as 2 branches on the right side and as 3 branches on the left side of somite 11 and extend throughout approximately 160 somites. The tenth parapodium has on the right side a gill mth 5 filaments (plate 2, fig. 9). From setigerous somites 20 to 60 the gills are long enough to meet across the dorsal surface and may have as many as 9 filaments, but throughout the greater number of the posterior gills the number of filaments is reduced to 1. The first parapodium has a thick dorsal cirrus, slightly smaller than the nuchal cirri, but otherwise similar to them in form. There is a very small setal lobe and a thick ventral cirrus. Parapodium 10 has a rounded postsetal lobe, the presetal with a concave margin (plate 2, fig. 9, posterior view); there are 2 very black aciculsc. The dorsal cirrus is finger-shaped, ^vith the 5-branched gUl arising at its base. The ventral cirrus is conical, on the end of a very prominent pad-like swelling. I was unable to find any needle aciculse in the dorsal cirrus. The fiftieth parapodium (plate 2, fig. 10) is not very different in general outline from the tenth, though the ventral cirrus is smaller. This difference is exaggerated in the drawing because the cirrus is partly under the ventral lobe and does not entirely appear. The dorsal cirrus is short and seems to arise from the base of the gill, the latter being so much greater in diameter than the cirrus. There are 2 dorsal acicula) and 1 ventral one. The gill has 8 filaments arising from a base which is very thick at the point of attachment, but narrows toward the apex. A posterior parapodium (plate 2, fig. 11) is broadly rounded in profile, with very little distinction between anterior and posterior hps. The cirri arc small. There arc 2 aciculse, 1 dorsal and 1 ventral, both very black. In the tenth parapodium there is a tuft of simple setae A^dth a few pectinate, and a ventral tuft of compound ones, this latter tuft being much stouter than the other. A few of the simple setae were bilimbate, but this did not show in all cases. In the fiftieth parapodium the pectinate setse show an increase in number over the conditions found in the tenth, with a corresponding decrease in the number of simple ones. In posterior somites the pectinate setse are more numerous than the simple ones and have extremely long stalks extending beyond the apex of the dorsal cirrus. A simple seta from the tenth parapodium (text-fig. 12) is very slightly widened and curved toward the end. In the one drawn the margins are smooth, in others there is a marginal fin. The compound setae have rather heavy stalks, the terminal joints with apical and subapical teeth covered by hoods with smooth margins (text- fig. 13). The pectinate setae have about 16 teeth, the terminal one at one end of the row being longer than the others (text-fig. 14). The dorsal aciculse are bluntly rounded at the apex (text-fig. 15); the ventral ones have terminal and subterminal teeth, the latter the larger; the apices are hooded (text-fig. 16). The maxilla (plate 2, fig. 12) is very dark, showing in the translucent portions a lighter bro\\Ti. The carrier is large relative to the forceps. The proximal paired plates have 5 teeth on the right and 5 on the left, the distal paired have 8 on the right and 6 on the left; the unpaired has 7. The mandible is lighter brown in color than is the maxilla, but has longitudinal brown stripings in each of the basal halves. The beveled portion has concentric fines (plate 2, fig. 13). The type is in the American Museum of Natural History. Leodice tubicola, new species. Plate 3, figures 1 to 6; text-figiires 17 to 23. One entire specimen collected on rocks near Breaker Point, Pago Pago Harbor. It was in a tube which had a membranous foundation covered with debris, the tube having a general zig-zag outline T\ith blindly ending branches. The tentacles and cirri uo LeodicidcB jrom Fiji and Samoa. are long and sharp-pointed. Thero arc no color characters to be noted. The animal is entire, is 25 mm. long, and its greatest width is 1.5 mm. There are about 110 somites. The gills begin on setigerous somite 12 and extend to within about 10 somites of the pygidium. There is never more than one filament, which in most of the somites is quite similar in size and length to the dorsal cirrus. Toward the posterior end these filaments are shorter than the dorsal cirrus. In somite 50 (plate 3, fig. 4) the filament is also shorter than the dorsal cirrus, but a little stouter. The prostomium (plate 3, fig. 1) is prominent, deeply incised, almost as long as the peristomium, but narrower. The tentacles are not quite twice as long as the pro- stomium, the longest barely reaching the anterior border of somite 4. The median and inner paired are almost of the same length, the outer paired arc considerably shorter than these. They are all smooth and slender, with acute apices. There is a considerable distance between the bases of the median and the inner paired, so that a large part of the dorsal sur- face of the prostomium is uncov- ered. The eyes are inconspicuous. The peristomium is on its dorso- median line about as long as the three following somites, its lateral margins very nearly straight, the small lip on either side not very prominent. Somite 2 is very short, the boundaries between it and somites 1 and 3 being very indis- tinct. The nuchal cirri are slender, about half as long as the first somite. The tenth parapodium (plate 3, fig. 3) has a slender dorsal cirrus into which needle aciculse extend, the setal lobe long, but with an un- usually short dorso-ventral diam- eter. The anterior Up of the setal lobe is rounded; the posterior Hp is a cirrus-like protrusion extending beyond the anterior. Between the two a large acicula with a bent apex comes to the surface and extends beyond the end of the posterior lip. Dorsal to the acicula is a tuft of simple setae, 4 in number in the parapodium dra\\Ti; ventral to it a tuft composed of two kinds of compound seta. No pectinate setse appear in the tenth parapodium, but they are in the eleventh. I am uncertain as to their exact distribution. Where there are so few of each kind of seta in a parapodium it is not possible to be sure that non-appearance may not be due to accidental loss and exact data concerning their distribution seems impossible to procure, if indeed it is a matter of any especial importance. The ventral cirrus is short, acute on the apex, and carried on the end of a pad-like swelling. The fiftieth parapodium (plate 3, fig. 4) has a conical setal portion, the slender dorsal cirrus arising from a common base with the gill, which is a little shorter than the cirrus but stouter. The ventral cirrus is finger-shaped. I could find no needle acicula in the dorsal cirrus. The most noticeable feature of this parapodium is the very large 19 Text-Figures 17 to 25. 17 to 23. Leodice tubicola. 17, compound seta X 250; 18, compound seta X 250; 19, simple seta X 250; 20, pectinate seta X 600; 21, postero-dorsal acicula X 250:22. postero-ventral acicula X 250; 23, anterior acicula X 250. 24 and 25. Leodice aciculata. 24, ventral acicula X 250; 25, compound seta X 250. LeodicidcB from Fiji and Samoa. 141 ventral acicula with sharply hooked apex, which comes to the surface near the setae tuft. The single dorsal acicula is straight and bluntly rounded at the end. Dorsally in the parapodium drawn are 3 pectinate setse with very long shafts and a tuft of simple setae with verj'- much curved stalks. Ventrally is a single compound seta of the usual type. The posterior parapodia retain the long dorsal cirrus found in the anterior somites and apparently do not differ essentially from the fiftieth. As the animal was allowed to remain in its tube, the posterior end was not well preserved, and it is diflBcult to be certain about the details of the last parapodia. There are two pairs of anal cirri (plate 3, fig. 2), the dorsal pair much longer than the other. Leodice tuhicola has the three kinds of setae characteristic of this genus, but differs from the majority in that there are two kinds of compound seta. In the anterior somites both sorts appear. One (text-fig. 17) has a terminal joint shaped like a knife- blade, with perfectly smooth edges; the other (text-fig. 18) has the bidentate terminal joint which is more characteristic of Leodice. As stated above, the number of these setae is so small in any somite that accidental loss might easily remove all of any one kind, and thus the result of accident be interpreted as absence, so that I have not attempted to study their distribution or to determine how far posteriorly the first kind of compound setae extend. They are certainly absent on the fiftieth parapodium. The simple setae (text-fig. 19) are long and in the posterior somites are much more curved than in the anterior, but in other respects they are similar throughout the body. Each widens toward the end and tapers to a sharp point with a wing along one margin. The pectinate (text-fig. 20) have about 12 rather prominent teeth. In posterior som- ites their shafts are much longer than in the anterior. The aciculae from anterior somites (text-fig. 23) are slightly curved at the end, bluntly rounded. In posterior somites there are two kinds of aciculae. The dorsal ones (text-fig. 21) are straight, Tvith rounded ends, the ventral ones much larger (text- fig. 22), their apices bidentate, the subterminal tooth in each being especially large and sharp. The apex of each of this last form is covered by a hood. The position assumed by the ventral acicula is unusual in that instead of lying at an angle with the dorsal one the two are parallel, the ventral one coming to the surface near the middle of the setal lobe (plate 3, fig. 4). The maxilla is light brown, with apices of forceps and toothed niargins of plates darker. There is a dark-brown band at the junction between the carrier and each half of the forceps; along the line of junction between the two halves of the carrier and at the base of the carrier, which is prolonged into two dark-colored tooth-like processes (plate 3, fig. 5). The proximal plate on either side has 3 teeth, the right distal paired has 8, the left distal paired has 2, the unpaired has 8. Distal to each series of paired plates is a patch of pigment and a thin plate with a recurved corner. The mandible was broken in removing and only one half is drawn. This (plate 3, fig. 6) has a small beveled surface not very sharply marked off from the shaft and carries at one side a horn-like protrusion. The shaft is noticeably marked with concentric lines. Crossland (1904, pp. 303-310, plate 21, figs. 1 to 8, text-figs. 52-55) described Eunice {Leodice) hibifex from Zanzibar, which is similar to L. tuhicola in character of tube and in the possession of two kinds of compound setae. While Crossland does not give measurements, this species was evidently much larger than L. tuhicola and differs from it in nearly every respect. With these larger specimens of Eunice (Leodice) tuhifex, Crossland collected some smaller individuals which he regards as the young of the same species. These he says were about one-third the size of the full-grown ones, one of "head" and 50 somites measuring 35 mm. in length; another of "head" and 35 somites was 13 mm. long. While these are larger than my single specimen of L. tuhicola, they agree with it so closely in the character of jaws and setae (the only characters Crossland gives) that I regard them as belonging to the same species, and either my L. tuhicola is a young 142 Leodicidce from Fiji and Samoa. Eunice (Leodice) tubifex or Crossland's small specimens are not tubifex, and I would adopt the latter explanation. It is evident that a comparison of Crossland's figure 7, plate 21, with his text-figure 53, showing the jaw of the small and the large individual respectively, indicates that they belong to distinct species. Figure 7 shows the proximal paired plates of the small animal. Crossland in the text gives the formula for these as 4-4, but it seems to me that the figure shows 2 teeth on the right and 3 on the left, all teeth very large, while his text-figure 53 shows these plates to have 7 on the left and 6 on the right, all teeth very small. The forms of the carriers are also quite unlike. Having made a considerable number of comparisons of the young and adult jaws in other species of Leodice, and having found that the general form is usually quite the same, I am very doubtful if such a jaw as is figured in figure 7 could ever become trans- formed into that of text-figure 53. Again, a comparison of Crossland's figures 8o to 8e, showing the setiB and aciculae of the small, with his figures 6a to Gd, showing the same structures in the large, shows very decided differences between the two. On the other hand, the figures of the small animals referred to above agree quite closely witli the figures I have given of L. tubicola. The most important differences are that I could not find marginal striations along the edge of the unusual form of compound setffi, that the structure of the carrier and forceps as given in his figure 7 do not agree with mine, and that his text-figure 55 shows a gill with 4 branches, while in tubicoln 1 never found more than 1 filament. Points of agreement are the form of the mandible and the general arrangement of the maxilla (Crossland's figure of the distal plates is not dear, but he gives 8-8 as the formula for their teeth) ; the form of the setai and of the aciculffi, especially the large hooked ventral acicula, which is exactly like that of tubicula. Crossland states that some of these small animals were sexually mature, which he interprets as meaning that sexual maturity appears before the animal has reached the adult structural condition. I would regard it as an adult condition, and until better evidence is presented for their distinction will include tlie small specimens of tubifex with lubicola. The type is in the American Museum of Natural History. Leodice coccinea Grube. Eunice coccinea Grube, 1878, pp. 153-155, pi. 9, fig- 1. , ^ ^ ^, ^ ^^ Eunice coccinea Crossland, 1904, pp. 297-303, pi. 20, figs. 6 and 7, text-figs. 4b to 51. In life the whole anterior end is dark green and very iridescent, the prostomium a little lighter green than the peristomium, the tentacles a lighter green than either. The tentacles have more or less purplish pigment around their bases and are a lighter green than the prostomium and have uncolored apices. The nuchal and dorsal cirri are colorless, except for a faint greenish band around the middle. The anterior somites are dark green, but at about the region of somite 30 the color begins to lighten and pos- terior to this the green is soon lost, the body-color being a light brown with numerous small yellow spots. The whole posterior region of the body is light yellowish browTi, still with the yellow spots, but at the extreme posterior end a purplish tint appears which becomes most intense at the pygidium. There is one pan- of anal cu-ri which are rather stout and colored an intense purple, but with uncolored tips. In the preserved material the bases of the cirrophores of the tentacles and a narrow ring around the base of each tentacle are dark purple, the tentacles and anterior cirri are green with uncolored tips, while later cirri are uncolored. The third setigerous somite is much lighter in color than any of the others. The peristomium is very dis- tinctly marked dorsally by anastomosing longitudinally arranged purplish lines, and this is continued but very faintly over the succeeding 2 or 3 somites. The quasi- articular condition of the tentacles mentioned by Grube is showii only by wrinkles. The single specimen in my collection does not agree with Grube's figure 1, plate 9, in that it has much shorter tentacles, the prostomium is more decidedly bifid and the third setigerous somite is uncolored. Grube states, however, that an African specimen Leodicidce from Fiji and Samoa. 143 had the whole sixth somite (fourth setigerous?) uncolored, so he evidently found some variations in this respect. Crossland's material was largely from the Maldives and evidently showed a considerable range of variability in form and color, for he identi- fied the species as coccinea, although some individuals showed as great a difference from Grube's description as does my Samoan specimen. Since the Samoan material agrees with Crossland's description, I have identified it as of this species. A single specimen was collected on Aua Reef in Pago Pago Harbor, Samoa. The body is 230 mm. long and has a peristomial width of 3 mm. The gills begin on somite 6 and extend over a distance of 40 mm. There are approximately 300 somites. Leodice aciculata, new species. Plate 3, figures 7 to 13; text-figures 24, 25. The general body-color is yellowish brown, somewhat lighter on the ventral surface, but otherwise with no noticeable difference in the two areas. Numerous small yellow spots are scattered over the entire dorsal surface. Toward the posterior end the general color becomes lighter, with a decided pearly luster which is more prominent on the ventral surface. A characteristic feature is an irregular banding and blotching of the tentacles and cirri with a brown pigment. On the tentacles there are several (5 or 6) of these bands which seem to extend entirely around, with many other shorter patches, which are very irregularly arranged. In alcoholic material the banding on the tentacles remains, but that on the nuchal and dorsal cirri may disappear. The pygidiumand somites immediately in front of it have a decided purple color in alcohol. Irregularly distributed colorless patches occur on the dorsal surface, these patches being of various sizes. The fourth setigerous somite has a colorless dorsal band, which varies in width in different individuals. The ventral surface of the prostomiura and peristomium are colorless, this appearing in a dorsal view as a whitish margin. A specimen 190 mm. long has about 250 somites and a peristomial width of 3 mm. The prostomium (plate 3, fig. 7) is 2-lobed, narrower than the peristomium. The tentacles are rather short and thick, the median extending as far as the second somite, the inner paired about as long as this, the outer paired shorter. The large eyes are in the usual position. As stated above, the tentacles are banded with brown, with the apices uncolored. The peristomium has straight margins, broadening at the anterior end, forming rather a prominent lip. The distinction between the first and second somites is most noticeable on the dorsal surface and is obscure elsewhere. The nuchal cirri are situated at the very anterior end of somite 2 and extend only to a little over half the length of the peristomium. They are slender and in life are banded with 1 irown. Somite 3 is about as long as somite 2 and there is very little change in diameter until the extreme posterior end. The apices of all parapodia are uncolored. The first has a very large dorsal cirrus, into which extend two aciculae which are unusually large as compared with the needle- like aciculaj usually found in this position. The ventral cirrus is thick and heavy, the setal portion very small. The tenth parapodium (plate 3, fig. 10) has a much greater dorso-ventral diameter, the dorsal cirrus smaller than in the first and provided with three aciculae. The setal portion has a presetal and postsetal lobe, the latter the longer, and the aciculae com(^ to the surface between them. The ventral cirrus is also smaller than in the first para- podium, but is carried on the end of a rounded swelling, which gives it the general appearance of being larger. A gilled parapodium (plate 3, fig. 13 of the sixtieth) shows a still greater reduction of the cirri, the setal portion remaining about as before. Two relatively large aciculae extend into the dorsal cirrus, and two especially large ones occur in the setal portion. Toward the posterior end the parapodia (fig. 9, the twentieth from the pygidium) are more nearly conical, the distinction between the anterior and posterior setal lips is less marked, and the cirri are very small. There is one pair of stout anal cirri (plate 3, 144 LeodicidcB from Fiji and Samoa. fig. 8). The gills begin from the fourteenth to the twentieth setigerous somite. In one individual there was a 2-branched gill on the left side of the seventh and no more until the twenty-first, but this was exceptional. They arise as a single branch, but become more complicated in the immediately following somites. One individual had a 2-branched gill on setigerous somite 18, a 4-branched one on 19, and a 3-branched one on 20. Throughout the greater part of the body the gills have 5 branches, the number becoming reduced to 1 or 2 toward the posterior end, but they are relatively long. The last gill is not more than 20 somites from the pygidium. The gills (plate 3, figs. 9 and 13) arise from the base of the dorsal cirrus and are large as compared with it. The blood-vessel is also prominent. The pectinate setse have a very slender stalk, the apex widening to form a broad and rather flat plate, carrying about 20 teeth. Proximal to each tooth is a small, highly refractile spot. The compound setae (text-fig. 25) have a small terminal joint, with a hood whose margin is finely denticulated, the basal joint being rather large. The terminal joint has a blunt apical and a sharp-pointed subapical tooth. The simple setffi are very long and slender, tapering gradually to a sharp point, and with a narrow wing on either side. As stated above, the aciculse are unusually large, especially through the median region, as is sho'WTi in plate 3, figure 13. The dorsal one of the two has a bluntly rounded apex and the ventral one is bifid (text-fig. 24). The aciculse of the dorsal cirrus are also unusually large. The jaw apparatus is very dark brown in color. The maxilla (plate 3, fig. 11) has a very short carrier, the forceps being long and slender and not much curved. The right proximal plate has 4, the left 3 teeth. The right paired has 5, the left has 7, the unpaired has 4. The unpaired plate is unusually small. The mandible (plate 3, fig. 12), has slender shafts widely separated, the beveled surface nearly round in out- line. On the beveled surface of the mandible and on the ends of the teeth of the maxilla is a whitish incrustation. This species was first collected in rocks outside the entrance light in Suva Harbor, Fiji. The surface of the rocks is much channeled by boring echinoids, and in the ridges left between these channels, L. aciculata occurs in large numbers, being the most abun- dant Leodicid that I found in Fiji. It was later collected in Samoa, a few individuals occurring in the rocks in Pago Pago Harbor, but was more abundant in rocks from the reef at Aunuu Island. Only a very few, however, were collected in Samoa. In a collection of Hawaiian annelids sent me for identification in July 1921, by the U. S. National Museum, was a single specimen of this species labeled as collected at Waikiki Beach. It was larger than any others I had seen, measuring 350 mm. in length, but was poorly preserved, so that this is only an approximate measurement. The coloring is more intense than in those from Samoa (possibly owing to greater age) and the brown bands on the cirri persist after preservation. The type is in the American Museum of Natural History. Leodice armillata, new species. Plate 3, figures 14 to 19; text-figures 26 to 29. The living animal is reddish brown, with the second setigerous somite uncolored and a row of white dots, one to each somite in the mid-dorsal line. The prostomium is colored like the remainder of the body, but has an uncolored patch on either side of the median tentacles. The tentacles are articulated, with greenish-brown pigment in the interarticular grooves, the median tentacles reaching as far as the fifth somite. The inner paired tentacles are almost as long as the median, the outer paired are much shorter. All dorsal cirri are uncolored; the anal cirri two pairs, one very short and colorless, the other pair much longer, colored light brown, but with uncolored tips. Neither is articulated. Leodicidce from Fiji and Samoa. 145 In general appearance this species is very similar to Leodice biformi-cirrata, if individuals of the same size are compared, but differ decidedly from larger individuals of the latter species. In L. biformi-cirrata the cirri and gills are large relatively to the somite; in L. armillata behind the middle region the somites are short, thick, and much rounded, with relatively very short parapodia. In small L. biformi-cirrata gills are absent from a considerable number of the posterior somites and there is one pair of articulated anal cirri; in L. armillata, even in the small individuals, the gills are con- tinued to within 10 somites of the py- gidiura and there are two pairs of non- articulated anal cirri (plate 3, fig. 15). The prostomium (plate 3, fig. 14) is bilobed, rather narrower than the peri- stomium. The tentacles are all monili- form without evident cirrophores, with about 25 joints in the median, 18 in the inner paired, and not more than 10 in the outer paired. The relative lengths of these tentacles in preserved material is indicated in figure 14. The eyes are prominent and lie in the usual position. The anterior border of the peristomium is produced into a narrow collar-like structure which protrudes to a short distance over the prostomium. The width of the peristomium is about twice that of its length; toward its posterior dorsal surface it has a prominent white spot, the remainder of its surface being tinted light brown. The animal figured was 2 mm. wide at the peristomium, 60 mm. long., and had 160 somites. The second somite is about one-third as long as the first, its nuchal cirri without artic- ulations and shorter than the peri- stomium. , , i. .J The gills begin with 1 filament on the sixth setigerous somite. On the left side of the body the eighth somite has a gill with 2 filaments, while on the right side the seventh and eighth each has 1 filament and the ninth has 3. Throughout the remainder of the anterior two-thirds of the body the number of filaments in each gill varies from 2 to 3. The number is reduced to 1 in the posterior one-third, not more than 10 of the terminal somites being without gills. The tenth parapodium (plate 3, fig. 16) has a relatively small setal lobe with only a slight distinction between the anterior and posterior lips, while the ventral cirrus is supported on a swelling that is almost as large as the setal portion. The dorsal cirrus is long and slender, shaped much like the gill which rises at its base, the diameter of the gill being only a very little smaller than that of the cirrus. There are 2 straight aciculie, and needle aciculse appear in the dorsal cirrus. A posterior parapodium (plate 3, fig. 17) is small as compared with the vertical height of the somite. Each has a rounded post-setal lobe with two dorsal aciculse and a ventral one, all nearly colorless or very faint yellowish brown. The dorsal aciculse have curved apices, one more curved than the other (text-fig. 28), while the ventral one is bidentate (text-fig. 29). The compound seta (text-fig. 26) has rather a stout basal shaft with a few indistinct denticulations along the longer edge. The terminal joint is small, the apex bidentate. There are only a few of these compound setae in each somite and their basal portions 3^ 30 Text-Fiqures 26 to 33. 26 to 29. Leodice armillata. 26, compound seta X 185; 27, simple seta X 185; 28, dorsal acicula X 185; 29, ventral acicula X 185. 30 to 33. Leodice crassi-tentaculata. 30, simple seta X 250; 31, compound seta X 250; 32. dorsal acicula X 250; 33. ventral acicula X 250. 146 Leodiddce from Fiji and Samoa. are long, extending beyond the end of the dorsal cirrus. The simple setse are also long, slender, and sharp-pointed, extending beyond the cirrus, with lateral striations along two sides (text-fig. 27). There were only 4 of these in the parapodium drawn. A very small tuft of pectinate setse lie close to the base of the simple ones; they are of the usual form but very small, with about 10 teeth. A parapodium from farther forward in the body (the thirty-fifth) differed in no essential respects from the one just described, but had a larger number of compound and fewer of the other two kinds of setse, and no ventral acicula. The maxilla (plate 3, fig. 18) is colored brown, darker between the halves of the carrier; the base of the forceps, and the terminal portion of each half of the forceps. The teeth are very clear-cut and prominent. The proximal paired plates have 6 teeth on the right and 5 on the left, the right distal paired has 10, the left distal paired has 8 (the plate was rolled so that I was unable to get a clear view of its margin and this number may not be quite accurate); and the unpaired has 6. The mandible (plate 3, fig. 19) is rather more dehcate than the maxilla, colorless except for a pigment patch between the halves and a line on either side near the margin of the beveled por- tion. From these colored spots concentric lines run inward over the surface. At the outer anterior angle of each side is a horn-like cyhndrical extension of the plate, which has a peculiar whitish tint, and is quite unusual for this genus. Leodice armillata was collected on Aua and on Utile reefs in Pago Pago Harbor. Samoa. The type is in the American Museum of Natural History. Leodice crassi-tentaculata, new species. Plate 4, figures 1 to 5; text-figures 30 to 33. A single specimen, collected on Utile reef in Pago Pago Harbor, in a loose coral rock lying on the sand near the shore. It has a general color-resemblance to L. biformi-cirrata, but can be distinguished from that species by the difference in the tentacles. While in L. biformi-cirrata the tentacles are articulated and not especially large, in L. crassi-tentaculata they are relatively enormous, being the largest as com- pared with the size of the entire animal that I have ever seen in this genus. The prostomial width of the specimen was 2 mm., the greatest body- width 3 mm. The specimen was in two pieces; the anterior piece 145 mm. in length with about 158 somites; the shorter piece had about 100 somites and was 60 mm. long. The extreme posterior end with pygidium was not found. In preserved material the anterior somites are mottled dorsally with yellowish brown on pearly white; the sixth somite has more white than any other somite, and there is a brilliant iridescence. This color gradually weakens away from the anterior region, and disappears entirely behind somite 50, the remainder of the body being a dingy yellowish gray. Anteriorly the ventral surface is iridescent, but has none of the markings of the dorsal surface. The gills first appear as a single filament on the left side of somite 34 and on the right side of somite 30, On the left side the second gill is 2-branched, while on the right it is the fifth which has the first of the 2-branched gills. This 2-branched condition is continued throughout the greater part of the specimen, but in the posterior portion of the smaller fragment there is but one branch. Gills continue to the very end of the specimen, losing, as above stated, in the posterior somites one of the gill filaments, but there is no diminution in the length of the remaining filament. The anterior gills are shorter than the dorsal cirri, but with the progressive decrease in the size of the cirri posteriorly and their own absolute increase in length, posterior ones are very much longer than the cirri. The prostomium is deeply bilobed, each lobe subdivided incompletely into a dorso- median and a ventro-lateral portion, forming the quadripartite lobing found in many Leodicids, The tentacles are very large, covering, when lying straight out in front, Leodicidce from Fiji and Samoa. 147 the greater part of the dorsal surface of the prostomium (plate 4, fig. 1). The median tentacle is over 5 times as long as the peristomium, the inner paired more than half as long as the median, the outer paired about as long as the peristomium, all very thick and heavy. In the preserved material no tentacle shows any color. The eyes are prominent. The lateral margins of the peristomium are nearly straight and parallel to one another, with the anterior lip on either side much in evidence. The peristomial length is about one-third less than its width and longer than the combined length of somites 2 and 3. In life the constriction between the anterior 3 somites is not very sharply defined and somite boundaries are further obscured by the dorsal mottling with pigment. The nuchal cirri are not quite as long as the peristomium. Throughout the anterior region the dorsal cirri are especially prominent, being both long and thick. Behind the region of somite 20 they are more slender, but remain long, and in the gill region they become successively smaller beyond the region of somite 50. In the posterior portions they are shorter than the gills and are sharp-pointed. As is common in this genus, the first parapodium has a small setal lobe \\ath large cirri, though these latter are relatively smaller in L. crassi-tentacidata than in most species. The tenth parapodium (plate 4, fig. 2), has a prominent setal lobe with dense tufts of simple and compound setse. The anterior lip is vertical, with a dorsal protrusion; the posterior lip is rounded. The apex of the setal portion has a ventro- anterior and a dorso-posterior rounded swelling. The compound setae arise between the anterior lip and the former of these swellings, while the simple setae arise between the latter and the posterior hp. Three heavy aciculae reach the surface between the two swellings. The dorsal cirrus is long and symmetrically narrowed toward the apex, the ventral cirrus short and thick on a rounded swelling. A tuft of needle aciculae extends into the base of the dorsal cirrus. The fiftieth parapodium (plate 4, fig. 3) is much smaller than the tenth, and has fewer setae. There is one dorsal and one ventral acicula, the latter hooked at the apex. The figure shows the rounded post-setal lobe, the anterior one being vertical. The dorsal cirrus is slender and sharp- pointed, larger than the base of the gill which arises from the basal portion of the cirrus. At some distance from its point of origin the gill divides into two nearly equal branches. There is a tuft of needle aciculae in the dorsal cirrus. The ventral cirrus is conical, on the end of a rounded swelling. Parapodia from the posterior end of the specimen in general outline and setal components are not noticeably different from the fiftieth, but the gill is 1-branched. The simple setae are of varying lengths, but in the tuft they are arranged so that the longest lie at the dorsal part of the tuft. Apart from length differences, they are all alilce, each (text-fig. 30) curved and tapering to a sharp point at the apex with a wing along the concave and convex edges. The compound setae (text-fig. 31) have their basal portions with no denticulations along the terminal edge, the terminal joints with terminal and subterminal teeth covered by a hood with smooth margin. The pectinate setae are very few in number, even in the posterior somites where the number in other species frequently exceeds the number of the other kinds. Each pectinate seta is very slender and delicate, with about 20 terminal teeth, but these were very difficult to demonstrate. The aciculae from somite 50 are very black (except for their extreme tips), the dorsal ones with bluntly rounded tip (text-fig. 32), the ventral ones with 2 teeth, of which the ventralmost is slightly the larger (text-fig. 33). The maxilla (plate 4, fig. 4), is dark-colored, especially at the ends of the forceps and the plates. The carrier is small, each half rounded on the margin; the forceps heavy relative to the carrier. The proximal paired plates have 4 teeth on the left and 5 on the right, the distal paired have 8 on the right and 5 on the left, the unpaired has 7. Dark pigment-patches lie just bej'^ond the distal plates and a very thin plate with one corner bent lies on either side of them. The mandible (plate 4, fig. 5) was 148 Leodicidce jrom Fiji and Samoa. broken in removing and onlj' one half is drawn. Each half is colorless except for faint lines along the line of contact of the two, and dark lines are at the median margin and outer angle of the beveled portion. From these patches of pigment concentric lines extend over the surface. The type is in the American Museum of Natural History. Leodice biformi-cirrata, new species. Plate 4, fignjres 6 to 11; text-figures 34 and 35. Collected both at Suva in Fiji and in Pago Pago Harbor, Samoa. A specimen from Samoa is about 95 mm. long and has a peristomial width of 2 mm. The body contains about 112 somites. The general body-color in life is an irregular splashing of white on a yellowish-brown background, the white being especially prominent on the anterior- dorsal face of the peristomium and on the fourth setigerous somite, which is entirely white, and a brownish dorso-median patch toward the posterior end. In some ten- tacles there is brown pigment in the constrictions between the joints. The tentacles are distinctly articulated with a small basal joint, the second joint being the longest of any, while at the apices they are moniliform. In a Suva specimen the median tentacle has 16 joints, the left inner paired one has 14, the right inner paired one has 11. It seems probable that accidental injuries are responsible for variations in this respect. The form of the peristomium and prostomium is indicated in figure 6, plate 4. The nuchal cirri are shorter than the peristomium and more or less wrinkled, but without articulations. The anal cirri are artic- ulated (plate 6, fig. 7) and have brown pigment between the joints. On one specimen the gills begin as a 2-branched organ on the fourth setigerous somite, become 4-branched on setigerous TEXT-FiauREa 34 somite 5, 6-branched on setigerous somite 6, 7-branched on se- ^^^^d 35. tigerous somite 8, and 6-branched on setigerous somite 12. ^^^^ .^^ biformi^irrata. From setigerous somites 15 to 20 the number varies between 5 34^ compound seta and 6, but later decreases, the usual number being 3, though x 185; 35, acicula there are exceptionally 5. The last gill has one filament and X 185. is on the fifth para podium in front of the pygidium. The first parapodium has large cirri, but shows no especial characteristics. The tenth parapodium (plate 4, fig. 8) has a stout setal lobe with rounded posterior lip, 2 very heavy dark aciculse which extend beyond the end of the posterior hp, and needle aciculse in the dorsal cirrus. The dorsal cirrus is large and more or less wrinkled, but is not at all articulated, and a 4-branched gill arises near its base. Just inside the body-wall on the dorsal surface of the parapodium is a black pigment spot and there is a smaller brown one near the ventral surface. The dorsal spots can be seen in a surface view of the entire animal. The ventral cirrus is ovate on the end of a pad-like swelling. The thirty-fifth parapodium is very similar to the tenth in general appear- ance, but the ventral pad has disappeared and the ventral cirrus is much larger, extend- ing to a considerable distance beyond the setal lobe. The dorsal cirri and gills are as in the tenth, but a ventral hooked acicula has made its appearance. In posterior somites the parapodia change very little in their general character, but the ventral pigment spot disappears and there is a gradual decrease in the number of gill branches. The compound seta (text-fig. 34) has a heavy basal portion and a relatively small terminal joint, the latter with two large teeth. The simple setae vary in length but all liave clearly seen denticulations along one edge. Some are nearly straight, others are much longer and curved. The pectinate setse are delicate with about 20 terminal teeth, the terminal one at one end being the longest. The maxillae are extremely delicate and were very easily broken, so that I was not able to get an entire one mounted for study. Figure 9 of plate 4 shows the right forceps LeodicrdcB from Fiji and Samoa. 149 with the right proximal and distal paired plates, the former with 5, the latter with 8 teeth; figure 10 shows the left half of the forceps, the left proximal paired, and the unpaired plates. The proximal has 5 large teeth. The unpaired plate has 6 teeth. The forceps are slender and have very small carriers. It is not easy to understand how such deUcate jaws as these could function in chewing and they may have been abnormal, though they are alike in all of the specimens I have. The mandibles have slender shafts which are darker and evidently much harder than the maxUlary plates. They are dark brown, with the outer portion of the shaft next to the beveled portion lighter in tint. The beveled portion is covered by a whitish incrustation (plate 4, fig. 11). The hooked acicula is heavy, with an especially large subterminal tooth (text-fig. 35) . This species shows some points of resemblance to Leodice {Eunice) tentaculata Quatrefages as described by Fauvel (1917, pp. 209-215, text-fig. 17), but differs in that the dorsal and nuchal cirri are not articulated, the number of filaments is much less, and the jaws have not the same structure. The type is in the American Museum of Natural History. Leodice gracill-cirrata, new species. Plate 5, figures 1 to 8; text-figures 36 to 38. A slender, easily broken form, first found in loosely constructed tubes made of stones and shells and fastened to the under side of rocks in Suva Harbor, Fiji. Later they were found in cavities in the dead stag-horn coral, and in this latter locality were without definite tubes. The body of the living animal has a pearly luster with a good deal of iridescence, a pink tint anteriorly because of the blood, and the middle region decidedly purple. Toward the posterior end there is a large dark pigment spot on either side in each somite. The pygidium has pigment spots. The whole pos- terior portion may be colorless (except for the dark spots), in which case the pygidium is pink. A specimen of average size measured after preservation 220 mm. in length, with a peristo- mial width of 2 mm., and this general width was maintained for the greater part of the body, but it tapered toward the pygidium. The prostomium (plate 5, fig. 1) is rather short and obscurely 4-lobed. The median tentacle extends to the posterior border of the fifth somite, the inner paired to the anterior border of somite 3, and the outer paired are a trifle shorter than the inner. In preserved material the median tentacle shows a slight trace of articula- tions, the inner paired are jointed for their terminal half, the outer paired for about the terminal third. The peristomium is as long as the three following somites, is noticeably narrowest along its anterior margin, its lateral margin decidedly rounded. Somite 2 is about one-third as long as somite l,the nuchal cirri are long and slender, extending fully to half the length of the outer paired tentacles. The pygidium (plate 5, fig. 2) carries two pairs of anal cirri, the larger dorsal ones being stout at the base, but narrow rapidly and are very long; tlie ventral ones are very small and slender. The tenth parapodium (plate 5, fig. 3) has a long dorsal cirrus which frequently has an appearance of jointing due to superficial wrinkling, and a broadly lanceolate ventral cirrus. Two aciculse are in the setal lobe and there is a tuft of needle aciculae in the dorsal cirrus. The fiftieth parapodium (plate 5, fig. 4) has a rounded post-setal lobe 37 ^ Text-Figures 36 to 39. n 39 36 to 38. Leodice gracili-cirrata. 36, compound seta X 250; 37, pectinate seta X 250; 38, acicula X 260. 39, ventral acicula of Marphysa simplex X 250. 150 Leodicidce from Fiji and Samoa. with biiiooth aciculac in the setal portion and needle acicula^ as before. The dorsal cirrus is long and slender, longer than the gills. The A^entral cirrus is rather long, finger-shaped. A posterior parapodium (plate 5, fig. 5) has a setal lobe with a rounded outline and no elongated postsetal portion. The dorsal cirrus is very long and slender, the ventral one sharply pointed. There are 2 straight dorsal aciculse and 2 ventral hooked ones. I could find no needle aciculse. Inside the body-wall is a large black pigment spot just dorsal to the bases of the acicute. The gills in one specimen begin as a single very long and slender filament on the third setigerous somite and the number of filaments increased to 2 on the fifth. Gills extend to about the region of the one-hundred and twenty-fifth from the pygidium. The largest number of filaments I could find in this specimen was 7. The dorsal cirrus has much the appearance of a gill filament, but is a little longer than any of the latter. The posterior gills are of but a single filament and very small. The compound setae (text-fig. 36) are similar throughout the body; the basal portions have narrow shafts much broadened at the apices, the terminal margins minutely serrated. The terminal joints have apical and subapical teeth, and a hood with minute marginal denticulations. The simple seta is long and slender, with a very slight increase in width toward the end and tapers gradually from this wider portion to the extremely acute tip. The pectinate seta (text-fig. 37) has the usual form with the terminal teeth the largest and about 9 other teeth along the edge. The acicula (text-fig. 38) is heavy, Avith a large subapical and a smaller apical tooth, both covered by a hood. The jaw apparatus varies in color with the size of the individual, being much darker in the larger and presumably older specimens. In the one figured (plate 5, fig. 6) the maxilla was light brown, with darker transverse bands on the base of the forceps. In a larger specimen the whole maxilla was one-third larger than the one figured and very much darker, the terminal two-thirds of the forceps being nearly black, and there was much dark pigment on the plates. In even the smallest specimens the shafts of the mandible are black (plate 5, fig. 7). The carrier of the maxilla is small, the terminal portions much curved. The right proximal paired plate has 7 teeth, the left one has 6, the right^ distal paired plate has 9, the left has 5, and the unpaired has 7. There is on either side a small rectangular plate lateral to the distal paired. The mandible has black shafts and a beveled portion Avhich is not A'ery sharply marked-off from the shaft. On either side is a thin chitinous plate with an irregular margin. The type is in the American Museum of Natural History. Genus MARPHYSA Savigny. J. C. Savigny, 1820, Systeme des Annelidas, p. 13. Similar to Leodice in most characters, but without nuchal cirri. Marphysa can always be distinguished from Leodice by the lack of nuchal cirri. Other characteristics, which are usually but not always present, are the relatively small size of the carriers of the maxilla, the presence of compound setse with long terminal joints, at most only finely denticulated along one edge, instead of the toothed distal joint covered by a hood, which is found in Leodice. A frequent feature of Marphysa is also the pectinate seta; of the posterior end, which, instead of the fine teeth of those farther forward, have only a few very heavy ones. Marphysa californica Moore. Plate 4, figures 12 to 14; plate 6, figure 1. Marphysa californica Moore, 1909, pp. 251 to 253, pi. 7, figs. 13 to 18; pi. 8, figs. 19, 20. Collected in sandy mud outside of mangroves in a lagoon a short distance southwest of Nuuli. In life the prostomium is a little wider than the peristomium, and its dorsal surface is greenish in color, the margins uncolored. The peristomium and the first few somites are dark green, the peristomium dotted with minute white specks, but LeodicidGB from Fiji and Samoa. 151 posterior io the sixth to eighth somite the green color disappears and the whole body has a flesh-color due to the blood in the body-walls. The gills are bright red and are prominent in the living animal. In the preserved material all color is lost, and the whole body has a milk-white appearance, with much iridescence at the anterior end. The anterior end of the body is rounded in cross-section, posteriorly it becomes very flat, as is common in this genus. An entire specimen after preservation is 160 mm. long, 1.5 mm. wide at the pcri- stoniium, and 3 mm. wide in the widest part. This is considerably smaller than Moore's type, but about the size of his cotype, assuming that his "3 mm." refers to the body- width. The prostomium (plate 6, fig. 1) is laterally rounded and deeply bilobed, so that if the median constriction were deepened each half would form nearly a circle. The tentacles are slender, the unpaired the longest, the others successively shorter, but all at least t^ice the length of the prostomium. The eyes are very small, situated between the bases of the inner and outer paired tentacles. Moore's specimens lacked the pos- terior end. The pygidium of the Samoan specimen has 2 pairs of anal cirri, one pair much larger than the other, both situated ventral to the large oval anus (plate 4, fig. 12, ventral view). The parapodia are as described by Moore. The setal iobe becomes more and more pointed toward the posterior somites and the ventral cirrus is very large and thick, fused to the setal lobe for the greater part of the length of the latter (plate 4, fig. 13). I found 4 aciculse in anterior somites, 3 in the forty-fifth (plate 4, fig. 14), and 2 in the one-hundredth parapodium, which is essentially in agreement with Moore's description. The only lack of agreement is in the character of the ventral acicula, which comes to the surface just dorsal to the ventral cirrus. Moore did not find this in his type and in the cotype it is bifid and hooded. I find it present in all except the anterior parapodia, but its form is very unusual in that it is not bifid and hooded, but has a straight end, bluntly rounded at the apex, quite similar to the other acicute. The gills in one specimen begin as 1 short filament on the twenty-fifth parapodium, on the twenty-sixth there are 3 very short filaments, and from here the number in- creases gradually to a maximum of 6. The filaments are slender and arise from a base which is very thick at the point of attachment and gradually narrows with the formation of each successive filament. The last gill has only one filament and is on the twentieth somite from the pygidium. The number of somites is approximately 300, so that gills extend over about 250 somites. I can add nothing to the description Moore gave of the setae or jaws except to make a slight change in nomenclature of the maxillary plates. Moore's figure 19, plate 8, is the left half of the maxilla (erroneously referred to in the text as the right). His III is the unpaired plate and his IV the left paired according to the nomenclature I am employing in this paper. Marphysa macintoshi Crossland. Marphysa macintoshi Crossland, 1903, pp. 137-138, pi. xrv, figs. 3 to 6; text-fig. 12. A number of specimens collected in Suva Harbor, Fiji, which I have identified as belonging to this species because of the peculiar form of the undivided prostomium, the character of the jaw apparatus, and the general form of the parapodia. In the Fijian specimens the gills had fewer branches and the dorsal cirri were longer than indicated in Crossland's figure 6. Marphysa simplex, new species. Plate 5, figures 8 to 12; text-figure 39. One specimen collected in Suva Harbor, in association with M. macintoshi, and much like it in general appearance, but differing decidedly in the form of the prosto- mium and tentacles. While in macintoshi the prostomium is shaped like a broad hoof of a horse, with no trace of a median indentation, in M. simplex (plate 5, fig. 8) it is 152 Leodicidoc jrom Fiji and Samoa. decidedly bifid. The tentacles of simplex are twice as long as the prostomium, while in macintoshi they hardly reach its anterior border. The prostomium (plate 5, fig. 8) is about half as long as the peristomium and de- cidedly bifid, with a definite angle at the anterior and the outer posterior portion of each half. The tentacles are rather stout, smooth, tapering at the ends, the unpaired and inner paired ones nearly equal in length, the outer paired ones much shorter. In the preserved material their bases are colored much like the prostomium, but at the ends they are very dark brown, nearly black in color. No eyes are \dsible in the pre- served material. The prostomium is about twice as long as the second somite; the two somites together are about as long as they are wide. There is no color in the anterior region, but the surface is very iridescent. This shades into a dirty gray pos- teriorly and toward the posterior end there is a faint trace of purple, due apparently to the tint of the intestine. Anteriorly the body is rounded, posteriorly it is much flattened. The animal is 100 mm. long, Mith a peristomial width of 2 mm., a greatest body- width counting parapodia of 4 mm., and contains about 200 somites. It seems to be entire, but lacks anal cirri. The gills begin as a single short filament on somite 242, the number increases to 2 on about somite 30 (plate 5, fig. 10) there is later an increase to 3, and throughout the greater portion of the gill region the number is 4 (plate 5, fig. 11). The last gill is on about the twentieth somite from the posterior end, the decrease in number of filaments being very abrupt at the end. The parapodia are unusually uniform in size throughout the body. The tenth (plate 5, fig. 9) has a short anterior and a rounded posterior lip, with aciculse protruding from between the lips. The ventral cirrus is rounded, the dorsal cirrus pointed at the apex. The forty-fifth parapodium does not show the distinction between anterior and posterior setal lips shown by the tenth, and the dorsal cirrus is narrower (plate 5, fig. 10). A bifid gill arises from the dorsal surface and there are 2 dorsal and 2 ventral acicufe. A tuft of needle aciculae extends into the dorsal cirrus. A posterior para- podium (plate 5, fig. 11) differs from the forty-fifth, mainly in the greater gill develop- ment. It has one dorsal and one ventral acicula and needle aciculs in the dorsal cirrus. I could not find any of these needle aciculse in the dorsal cirrus of the tenth parapodium. The dorsal aciculae are straight with rounded ends, while the ventral ones have the usual form, with a terminal and a subterminal tooth and a hood (text-fig. 39). No compound setse were to be found. The simple setae are very long and slender, showing nowhere any indication of a broadening from the average width of the stalk, but tapering apically to a very slender point, the whole setae thrown into several curves. Along one margin there may be a series of very small denticulations. The simple setse are quite similar in form throughout the body, varying only in the number of curves, in length, and in the sharpness of the marginal denticulation. Pectinate setse occur sparingly in anterior somites, a parapodium from the region of somite 45 showing two, but I could find none in the parapodium drawn in figure 9. At the extreme posterior end they are more prominent. They are all of one kind, with a broad unsymmetrical end and about 25 very slender teeth. The maxilla is dark broAATi in color. The whole jaw apparatus was broken in removing it from the body, and I am unable to give all of the details of its structure. The carriers are broken (plate 5, fig. 12). The forceps are long and not much curved, the left proximal plate has 5 teeth, and apparently the right one has 5 or 6. The right distal paired plate has 8 teeth, the left paired has 2, the unpaired has 9. The forceps are dark brown, the other plates much lighter in tint, but colored dark brown along the edge where the teeth are. The mandible was too badly broken to describe, so that aside from the statement that along the cutting edge and on the shaft there is much dark pigment I am unable to make anj' statements concerning it. The type is in the American Museum of Natural History. Lcodicidcc from Fiji and Samua. 153 Genus PARAMARPHYSA Ehlers. Ernst Ehlers, Florida Anneliden, p. 99. Similar to Marphysa in every respect except that it lacks gills. Tlic individuals are usually small as compared with Marphysa and have delicate and comparatively soft jaw-apparatus. Paramarphysa teres, new species. Plate 6, figures 2 to 6; text-figures 40, 41. Two incomplete specimens were collected in Pago Pago Harbor. One specimen, incomplete posteriorly, is 75 mm. long, 0.75 mm. in diameter at its widest part, and contains over 200 somites. The prostomium (plate 6, fig. 2) is deeply bilobcd, hardly wider than the peristo- mium. The median tentacle is about twice as long as the prostomium, the inner paired about half as long as the unpaired, the outer paired rather more than half as long as the inner paired. On either side tliere is a prominent eye situated just outside the base of the inner paired tentacle. No pigmenta- tion is present in any portion of the body. In the preserved material the anterior ventral surface is much flattened, the parapodia prominent, and the dorsal cirri are relatively large. Behind about the region of somite 50 there is a change in the form of the body, in that the cross-section is nearly round and the dorsal cirrus be- comes very small. One of the two specimens was badly distorted as a result of drying and the prostomium of the other had been injured, so that neither was entirely normal, but neither showed any distinction between the first two somites, which together are longer than the prostomium. At the anterior border the prostomium has a median indentation, and its anterior diameter is a trifle Avider than iU posterior. Anterior parapodia (a drawing of the thirteenth is shown in figure 3, plate 6) have relatively rather prom- inent dorsal cirri, with a vertical anterior and a rounded posterior lip, and a dense tuft of setae arising between the two. There is a single acicula situated near the dorsal surface of the parapodium. The acicula has a very dark base, but is much lighter in color near the apex. Both simple and compound setse were present in the parapodium drawn. A number were broken, but apparently the arrangement is that there is an anterior vertical row of compound setse extending to the dorsal surface of the parapodium. Posterior to the dorsal end of tliis row is a smaller tuft of simple setae. A posterior parapodium (plate, 6, fig. 4, from the region of the two-hundredth somite) is conical, with no evident distinction between anterior and posterior lips, and with quite similar dorsal and ventral cirri. The ventral cirrus is located a little nearer the apex of the parapodium than is the dorsal, but otherwise they are very similar. In the one drawn there were 2 simple and 2 compound setse, the simple lying ventral to the compound. Simple setse from the region of the two-hundredth parapodium are very slender, of uniform width to about one-third of their length outside the body- wall; at this point they bend at an angle of about 45° and taper to a very fine point. The compound setse are ver}'- small and slender, the terminal portion ha^dng an apical and a subapical tooth and a hood (text-fig. 40). In the thirteenth parapo- dium the compound setse are similar to those in later somites, while the simple ones \\-iden slightly, bend toward the apex, becoming very narrow and curved beyond the bend (text-fig. 41). Text-Figures 40 to 44. 40 and 41. Paramarphysa teres. 40, compound seta X 250; 41, simple seta X 250. 42 to 44. Lysidice fusca. 42, compound seta X 250; 43, pec- tinate seta X 500; 44, simple seta X 250. 154 LeodicidcB from Fiji and Samoa. In both of the specimens at my disposal the jaw apparatus was incomplete, appar- ently due to injury. The carrier is large, the forceps have heavy bases and slender apices, and the proximal pau-ed plates each has 3 teeth. I could find no other plates (plate 6, fig. 5). The mandible is composed of slender halves rather widely separated (plate 6, fig. 6). The type is in the American Museum of Natural History. Genus ONUPHIS Audouin et Milne Edwards. Audouin et Milne Edwards, 1834, p. 151, pi. 3a, figs. 1 to 5. Prostomium with 7 appendages arranged in 3 rows, the anterior row of two "frontal tentacles" or "frontal palps," short and rounded at the ends. Other appendages on ringed cirrophores. Anterior parapodia produced so as to extend in front of the prostomium. The gills are pectinate or simple. With one pair of nuchal cirri. Omiphis may be distinguished from Diopatra by the gills, which in the latter genus are spirally coiled, and from HyalincBcia by the possession of nuchal cirri lacking in Hyalinoecia. Onuphis holobranchiata v. Marenzeller. Onuphis holobranchiata v. Marenzeller, 1879, p. 24-26, pi. 4, fig. 1. Onuphis holobranchiata Crossland, 1903, p. 135, pi. 14, fig. 2. Onuphis holobranchiata Augener, 1913, p. 283-284. A single incomplete specimen lacking the posterior end was collected in Suva Harbor, Fiji. What remains of the body is 30 mm. long, 2 mm. in greatest width, and contains 69 somites. The color is a uniform dark brown, with brilliant iridescence on the dorsal anterior surface. In all details of surface structure this agrees closely with V. Marenzeller's description and figures. In the jaws there are only slight discrepan- cies, the carriers having more globular outlines than v. Marenzeller described. Augener suggests that this species may be identical with Johnson's Onuphis (Northia) ekgans and iridescens (1901, pp. 406 to 408, plates 8, 9, figs. 77 to 92) . In gill structure 0. holobranchiata agrees more closely with Johnson's species elegans than with irides- cens. I have compared the Suva specimen with one of 0. elegans which I collected at Friday Harbor, Washington. Neither has the 3-jointed inner paired tentacles figured by Johnson, but my Friday Harbor specimen agreed in other details, especially as to the jaws, with Johnson's description, and differed in jaw structure from the Suva specimen. I think they are distinct species, though closely related. v. Marenzeller's specimens came from the east coast of Enosima Island, and Augener's from Sharks Bay, Freycinet, estuary between Baba Head and Cararong "Halb-Inseln," in 7 to 11 meters. Genus LYSIDICE Savigny. J. C. Savigny, Systems des Ann^lides, 1820, p. 13. With a jaw apparatus like that of Leodice, but with only 3 tentacles and no nuchal cirri. The mandible is usually very large as compared with the maxilla. Generally rather small in size. Lysidice fusca, new species. Plate 6, figures 7 to 13; text-figures 42 to 44. Collected both at Suva in Fiji and at Pago Pago Harbor in Samoa. The animal lives in the porous rocks in association with Niddion, and in Samoa is as abundant in the porous surface of the dead rock as is Niddion in the West Indies. Because of the way it twists its body into the intricate cavities of the rock, unbroken specimens are difficult to secure. There is considerable variability in both size and color. The specmaen whose maxilla is drawn in figure 12 was 2 mm. in body diameter, and this would be about the maximum size. The most characteristic coloration is that figured (plate 6, fig. 7), Leodicidw from Fiji and Samoa. 155 in which there is a sharply defined dark patch on the dorsal surface of I he prostomiuni and the first 3 somites are dark brown with numerous white spots. Somites 4 and 5 are uncolored and the remainder of the body is like the anterior region, except that the colors are lighter. In most of the body each somite has a narrow darker band across its anterior margin. The pygidium (plate 6, fig. 8) is darker than is most of the body, with a narrow very dark band across the posterior margin of each somite. The third body somite may show less pigment than the first and second and there may be a little pigment on the fourth and fifth. Preserved material shows at best only a trace of this coloration and it may be entirely lost. The prostomium (plate 6, fig. 7) is deeply bilobed and is wider than the peristomium. The tentacles are uncolored, the median one a little longer than the prostomium, the two lateral ones extend just beyond its anterior margin. The kidney-shaped eyes are very dark and prominent. The first parapodium (plate 6, fig. 9) has a heavy ventral and a more slender dorsal cirrus; the tenth (plate 6, fig. 10) has a very large setal lobe and the ventral cirrus is on the end of a heavy pad-like structure; in posterior parapodia (plate 6, fig. 11) the setal lobe is conical and the ventral cirrus very small. In all parapodia the dorsal cirrus is slender. The anterior parapodia have each one large acicula, to which a ventral hooked acicula is added in later somites. There are no needJe aciculse in the dorsal cirri. There is one pair of stout anal cirri (plate 6, fig. 8). The setae are all very small. The tenth parapodium has a dense tuft of compound sctff, each with a shaft serrated at the apex; the terminal portion has a terminal and a subterminal tooth covered by a hood with serrated margin (text-fig. 42) . The pectinate Betse are small, each with about 12 teeth (text-fig. 43). The simple setse (text-fig. 44) are short, curved toward the ends, and with serrated convex margins. In the posterior parapodia the pectinate setse are much larger than anteriorly and form a prominent dorsal tuft on the setal lobe, while the other setse remain as in anterior somites. The maxilla (plate 6, fig. 12) is dark brown in color, the color deepening toward the inner margins, though in the forceps the extreme margin is colorless. The carriers are conical, almost as long as the forceps. The basal portion of the forceps is about as long as the free fang. The right proximal paired plate has 3 teeth, the left has 4. The right distal paired has 5 teeth, the left has 2. The unpaired plate has 3 teeth. On either side is a small accessory plate. The mandible (plate 6, fig. 13) is relatively rather large, each half with a decidedly rolled edge which is darker than the remainder. (Note that it is drawn under about half the magnification of the maxilla.) Most of the mandible is light in color, but there are numerous dark bands (plate 6, fig. 13). The type is in the American Museum of Natural History. Lysidice parva, new species. Plate 6, figures 14 to 17; text-figures 45, 46. So far as mji- material goes, the animals of this species are of small size, the type, which was the only complete individual in my collection, being 70 mm. long and at no place more than 0..5 mm. in width. Other individuals were larger, but none more than 1 mm. in body-\Nidth. The anterior region of the living animal is colorless, except for faint indications of brownish spots on the parapodia on either side. Behind the colorless anterior region is one where there is a prominent dark spot on the dorsal surface of each parapodium. Farther back the body-color is lemon-yellow, due appar- ently to the intestine seen through the colorless body- wall. In preserved material the color is a uniform brown. The prostomium (plate 6, fig. 14) is about as long as the peristomium and is bilobed, this lobing being more noticeable in some individuals than in others. The median tentacle is twice as long as the prostomium ; the lateral ones extend for about one-third of their length beyond the prostomial border. The eyes are not very large, but are very distinct, dark brown in color. The peristomium is nearly twice as long as somite 156 Leodicidce frotn Fiji and Samoa. 2, the line of separation between the two being very indistinct. Later somites continue in gen- eral the diameter of the first two, with a narrow- ing and a flattening toward the posterior end, which carries one pair of rather heavy anal cirri. Each anterior parapodium (plate 6, fig. 15) has a single acicula which is very dark brown in color except for the very apex, contained in a rounded setal lobe. The dorsal cirrus is long and slender, the ventral one much shorter. Posteriorly the only change in the parapodium is that the setal lobe becomes more pointed. The setae are very small (note the scale of text-fig. 45); the simple ones broaden near the end and narrow to a very fine point (text-fig. 46) ; the compound ones have inconspicuous teeth on their hooded terminal] oints (text-fig. 45). The jaws are extremely small and very easily broken. The carrier (plate 6, fig. 16) is larger than the forceps, being broader and about of the same length. The forceps has slender fangs. Each proximal paired plate has lobings, hardly to be called teeth. I was unable to get satisfactory preparations of the distal plates, and beyond the statement that there are two on one side and one on the other I can say nothing of their structure. The whole maxilla is pale yellow in transmitted light, with a dark transverse band at the junction between carrier and forceps. The mandible is larger than the maxilla (note the difference in the scale of magnification in the drawings), and is very thin and delicate (plate 6, fig. 17). In general form and in the structure of jaw and setse this species is closely related to Lysidice tortugw of the Gulf of Mexico (Treadwell, 1921a, p. 85, figs. 298 to 304). The type is in the American Museum of Natural History. Genus NICIDION Kinberg. J. G. Kinberg, Annulata Nova, 1864, p. 564. Similar to Leodice in having 5 tentacles and a pair of nuchal cirri, but without gills. The parapodial development, especially in the median and posterior regions, is very sUght. They might be mistaken for young Leodice in which the gills have not appeared, but can be distinguished from these by the feeble parapodial structure. Nicidion fusca-fasciata, new species. Plate 7, figure 5; text-figures 47 to 50. Collected near the Governor's wharf and on Aua Reef in Pago Pago Harbor, Samoa. The prostomium is entirely colorless, a little wider than the peristomium, and has a very shallow anterior median notch. The unpaired and the inner paired tentacles are yellowish brown in color, except for the extreme tips, which are without color. The outer paired are entirely colorless. The median tentacle is as long as the first 3 somites, the others from the inner to the outer become progressively shorter. The eyes are prominent. The peristomium is colorless and is longer than the two following somites. Somites 2 and 3 are entirely colored, except for a small white patch on either side of the mid-dorsal line on the anterior margin of each somite, and in somite Text-Figures 45 to 51. 45 and 46. Lysidice parva. 45, compound seta X 560; 46, simple seta X 250. 47 to 50. Nicidion fusca-fasciata. 47, simple seta X 250 ; 48, simple seta X 250; 49, compound seta X 250; 50, pectinate seta X 250. 51. Compound seta of Lumbri?iereis sphmrocephala X 250. Leodicidce from Fiji and Samoa. 157 2 the small nuchal cirri, which are entirely colorless and very small, so that they look like two small white patches, one on either side of the somite. A very narrow band of the posterior margin of somite 3 is colorless. Somite 4 has on either side on the dorsal surface a triangular pigment patch with the apex pointed dorsally; somite 5 has a similar but larger patch and on somite 6 and 7 these patches from opposite sides have united in the mid-dorsal line. These are yellowish hrovm like those on somites 2 and 3, but are lighter in tint than on those somites. Behind somite 7 the pigment patch begins in front of the parapodium, bends around it so as to lie dorsal to it, and then extends dorsally to meet its fellow, leaving the posterior half of the dorsal surface of the somite uncolored. On either side is a colorless patch similar to those on somites 2 and 3. Behind the region of somite 20 these patches disappear and the band of pigment becomes entire, bifurcating at its ends so as to go on either side of the para- podium. The pigment disappears behind the region of somite 50, the posterior end being, so far as I can tell, without color. The first parapodium has relatively very large dorsal and ventral cirri and a large acicula (plate 7, fig. 6). There is an antero-ventral and a postero-ventral setal lobe. The tenth parapodium (plate 7, fig. 7) has a slender dorsal and heavy ventral cirrus, 2 postsetal and one presetal lobes, and a single large acicula which is colorless at the base and end but dark in the middle region. A posterior parapodium (plate 7, fig. 8) consists mainly of a rounded setal lobe with very small dorsal and ventral cirri and 2 black aciculee, the ventral one hooked at the end. The simple setse of the first parapodium are not very long, are very sHghtly widened toward the end, and very sharp (text-fig. 47). In later somites these setse are much longer and not at all widened (text-fig. 48). The compound setse are very small, with a minute terminal jomt carrying a pointed terminal and a rounded subterminal tooth, the two covered by a hood (text-fig. 49). Anteriorly the pectinate setse (text-fig. 50) are small, with about 12 teeth. Posteriorly they are larger and with more numerous teeth. The pygidium was not present in any of my specimens. The mandible (plate 7, fig. 9) is thin and only very faintly colored, except for dark bands at the base of the forceps and between the forceps and carrier. The carrier is small, the forceps long and not much curved; the right proximal plate has 6 teeth, the left has 5; the right distal pau-ed plate has 9 teeth, the left has 4; the unpaired has 6. The mandibles (plate 7, fig. 10) are long and slender, the halves only slightly united, and have practically no pigment. The type is in the American Museum of Natural History. Subfamily LUMBRINEREIN^. Ventral cirrus absent, dorsal cirrus rudunentary or foliaceous. No appendages or evident gills, but with anal cirri. The maxillary plates are all paired, but the two of the same pair may or may not be symmetrical. Dorsal lobes, probably functioning as gills, have been described in Lumbrinereis branchiata (TreadweU, 1921a, pp. 94, 95, plate 8, figs. 5, 6; text-figs. 333-343), but it is doubtful whether these could he regarded as homologous with the gills of the other Leodicidae. Genus LUMBRINEREIS de BlainvUIe. H. M. de Blainville, Dictionnaire des Sciences Naturelles, 1828, p. 46. Body elongated, without prostomial appendages or parapodial cirri. The first somite interrupted ventrally by a forward extension of somite 2 to form the posterior border of the mouth. Maxilla of short carriers, forceps, and 3 pairs of toothed plates. Mandible about as long as maxilla, the two halves more or less fused. Setae compound, simple, and hooked. 158 LeodicidcB from Fiji and Samoa. Lumbrinereis sphaerocephala Schmarda. Text-figure 51. Notocirriis sphaerocephala Schmarda, 1861, p. 116. Lumbriconereis sphcerocephala Ehlers, 1904, pp. 33-34, pi. v, figs. 3 to 11. Lumhriconereis sphcerocephala Augener, 1913, p. 288. A single specimen was collected at low tide in Suva Harbor, Fiji. The living animal is more like Oenone than Lumbrinereis in its general appearance, for it has a soft body which secretes much mucus and it moves much more slowly than is usual in Lum- brinereis. In life the prostomium varies in form according to the degree of contraction, but is always nearly hemispherical. The greater part of the dorsal surface of the prostomium is colored dark greenish-brown, leaving only the margin uncolored. On the dorsal surface of the peristomium are two pigmented bands with sharply defined pigment rows lying in the broader one. Similar cross-bands appear in following somites, but gradually decrease in size posteriorly, so that behind the one-hundredth somite the dorsal surface of each somite is marked only by many minute spots of a golden color. The ventral surface has a yellowish tint, but no pigment. Interseg- mental constrictions are uncolored. Preserved material as far back as somite 30 is reddish brown with darker bands in the middle of each somite, while behind this the whole animal has a decidedly greenish tint. In the structure of the head region, the parapodia, and the jaws my specimen agrees with Ehlers's description, except that the prostomium is more rounded than is shown in Ehlers's figures 3 and 4. The teeth in the setsB are more sharply defined than Ehlers represents them (text-fig. 51), but this may have been merely an error made by his artist. Ehlers's material was collected at Pitt's Island and Chatham Island; Schmarda's at Auckland, New Zealand; Augener's at Station 1, Sharks Bay, northwest of Middle Bluff, in 7 to 8 metres; Station 26, Sharks Bay, Sunday Island, 5.5 meters; Station 56, Koombaua Bay, 6 to 7 English miles southwest of Bunbury, 14.5 to 18 meters. Benham (1915, p. 227) records a single specimen identified as of this species from east of Babel Island, Bass Strait. Lumbriconereis brevicirra Schmarda. Lumhriconereis brevicirra Schmarda, 1861, p. 117. Lumbriconereis brevicirra Ehlers, 1904, pp. 35-36, pi. iv, figs. 13 to 20; pi. v, figs. 1 and 2. Lumbriconereis brevicirra Augener, 1913, p. 288. Two specimens were collected at Rat Passage, in Suva Harbor, Fiji, in sand at the bottom of a shallow pool on the surface of the reef at low tide. Later some were col- lected in mud near the Carnegie Library at Suva. In the preserved material the two from Rat Passage are uniformly dark gray in color, with only very faint pigment bands in some of the posterior somites. In life each had a median and two lateral pigment patches on both the dorsal and ventral prostomial surfaces. The median patches are circular in outline, the lateral ones are nearly linear. In the specimens from the mud the whole body-color is much lighter and the prostomial pigment patches are very prominent. In these latter specimens each somite throughout the body has a transverse reddish-orange pigment band, covering more than half of the dorsal and ventral surfaces. There are two pairs of anal cirri, one of which, in one specimen, was bifid. The Fijian specimens agree with Ehlers's figures and description in the form of the prostomium and in the nearly complete fusion of somites 1 and 2. They all show more of the longitudinal plications on the ventral surface of somite 2 where it extends forward to form the boundary of the mouth than Ehlers shows in his figure 13, and I saw no trace of the everted nuchal organs Ehlers shows in his figure 14. The parapodia have longer posterior cirri than Ehlers describes, but the structure of setae and aciculae correspond Mith his description. In the jaw apparatus the maxilla is as Ehlers Leodicidce from Fiji and Samoa, 159 describes it, but the mandible is quite different. Ehlers shows the mandible as constricted near the middle of the shaft, which widens at either end. My specimens have a mandible much more like that of L. sphcerocephala as given in Ehlers's plate v, figure 11. This mandible which Ehlers figures is very different from anything which has been described for this genus, and it seems to me legitimate to question whether it was a normal specimen. Schmarda recorded the species from Port Jackson; Ehlers from Chatham, Waitangi; Augener from Sharks Bay, 2 to 4.5 meters, Cockburn Sound, Port Royal, 14.5 to 18 meters, Albany, Princess Royal Harbor and Oyster Harbor. Lumbrinereis japonica v. Marenzeller. Plate 7, figures 1 to 4. Lumbrinereis japonica v. Marenzeller, 1879, pp. 29-30, pi. v, figs. 3 to Sd. Two specimens were collected in Pago Pago Harbor, on the under surface of loose coral rock at Utile reef. This is an unusual position for this genus, which is essentially mud-dweUing. The larger individual is 250 mm. long and composed of about 300 somites. The pygidial region is regenerating and contains about 15 very short and narrow somites. There are two pairs of short, stout, unequal anal cirri, v. Maren- zeller's specimen lacked the posterior end. The second of my specimens is only about two-thirds as large as the other one, and only the anterior half is retained. The prostomium (plate 7, fig. 1) is about as wide as it is long, with a blunt-pointed apex. The peristomium is a little wider than the prostomium, and is about twice as long as somite 2, from which it is separated by a poorly defined constriction. Ven- trally, as is characteristic of this genus, somite 2 extends forward through an interrup- tion in somite 1 to form the posterior border of the mouth. This ventral prolongation is longitudinally plicated. The parapodia have prominent posterior lobes and, as stated by v. Marenzeller, each parapodium has a rudimentary dorsal cirrus into which a tuft of seta extends (plate 7, fig. 2, taken from somite 100). Posteriorly there is an increase in the length of the posterior parapodial lobes, but aside from this the parapodia are of the same form throughout. The maxilla (plate 7, fig. 3) is dark brown, almost black, but with occasional lighter brown areas, especially along the margins of the carriers. The carrier is rather long and the margins have a frayed appearance. The forceps is slender. The left proximal plate has 4 teeth, the right proximal has 5. Apparently the 2 terminal teeth of the right plate are more or less broken. The second and third pairs of plates have respectively 2 teeth and 1 tooth on a side. In the figure these are shown as inverted, a position assumed during dissection. The teeth should lie on the inner side of each plate. The mandible (plate 7, fig. 4) is narrow at the base, but broadens at the anterior end. It is marked by concentric lines both in the shaft and in the beveled portion, the latter with very darkly pigmented ends. V. Marenzeller gives no drawing of the prostomium, but his figures of parapodia and jaw agree so well with those of the Samoan specimens that I have no hesitation in assigning them to this species, v. Marenzeller figures three forms of setae, simple winged, simple hooded, and compound hooded, the last two agreeing in form of apex and hood. In the smaller of my two specimens I find the form and distribution of setse exactly as v. Marenzeller described them, but the larger specimen lacks the compound ones. It would be desirable to examine a series of varying sizes to determine if these setse are lost with increasing age or size, but the material now in hand is not suflftcient for this purpose. Moore (1903, p. 454) records this species from Sagami and Suruga Bays, Japan. V. MarenzeUer's was from the east coast of Eno-Sima Island. 100 Leodicidce from Fiji and Samoa. Genus ARABELLA Grube. A. E. Grube, Die Familien der Anneliden, etc., 1851, p. 45. Body elongated, slender, without prostomial appendages or gills, with rudimentary dorsal and anal cirri. No compound, pectinate, or hooked setse. Eyes often present on the prostomium. Maxilla with long slender carriers, forceps, and 4 or 5 pairs of toothed plates which may or may not be symmetrically arranged on the two sides. Mandible well developed, the shafts pointed and well separated. All jaw apparatus very dark in color. Arabella may easily be distinguished from Lumbrinereis, which it resembles very closely in general appearance, by the fact that the first somite forms the posterior border of the mouth, instead of the second, as in Lumbrinereis. In structure of the jaw Arabella resembles Drilonereis, but the maxillary plates are larger and may be more numerous, and the mandible is always more developed than in that genus. Arabella dubia, new species. Plate 7, figures 11, 12; plate 8, figures 8, 9; text-figure 52. The type is from Pago Pago Harbor, Samoa, and is about 70 ram. long, with a prostomial width of less than 0.5 mm. An incomplete specimen found in mud in Suva Harbor, Fiji, is much larger but is badly preserved, so that measurements are of little value. It is possible that the type may be immature. The body is very irides- cent. One preserved specimen from Pago Pago shows transverse brown bands on the somites, while the other is colorless. The prostomium (plate 7, fig. 11) is elon- gated sugar-loaf in form, its length about one- third greater than its width. There are two paiis of eyes, the inner pair lying close to- gether near tlie posterior border of the pros- tomium, and easily seen; the second pair are much less easily seen and lie one on either side, at a little more than half the distance from the corresponding inner eye to the pros- tomial border. The first somite is a little wider than the prostomium, but about two- thirds as long. Its anterior margin is notice- ably recurved. The parapodia are well developed from the beginning. They have the form usual in this genus, with a rounded setal lobe and a finger- shaped posterior one. There is a single, rather prominent, yellow-colored acicula, rounded at the apex. The setse (text-fig. 52) are broadened and bent toward the apex, this bent region striated. Along the convex margin is a narrow wing. A denticulation along the margin of the wing is very marked in some and barely discernible in others very small. There are two pairs of stout anal cirri (plate 7, fig. 12). The maxilla is jet-black \sdth long carriers, not all of which are shown in the figure (plate S, fig. 8). On the basal portion of the right forceps are 7 prominent teeth, with smaller denticulations at the base. In the forceps figured, the right half has a bifid apex which apparently is not present in the other specimen at my disposal and is probably an individual 54 53 55 TEXT-FiauBE3 52 to 56. X 52. Simple seta of Arabella dubia 250. 53 to 55. Drilonereis lumbricus. 53, simple seta X 185; 54, acicula X 185; 55, acicula X 185. 56. Simple seta of Drilonereis paiici- dentata X 500. The number of setae in each somite is Leodicidce from Fiji and Samoa. 161 abnormality. The left half of the forceps has a shorter base and longer fang than the right; the base has 4 well-marked teeth. The first pair of plates are asymmetrical, the right one much shorter than the left, and its margin has 7 rounded teeth, while the left one is longer than the forceps, with at least 10 large, sharp teeth. The second, third, and fourth pairs of plates are symmetrical, the second and third each has 4 and 3 teeth respectively, while each one of the fourth pair has one tooth. The mandible (plate 8, fig. 9) is large as compared with the maxilla, its halves fused for more than half their length. In character of prostomium, parapodia, and setae this species is closely alUed to A. attenuata Treadwell (1906, p. 1172, fig. 62), but differs in the character of the jaw. In the structure of the jaw it approaches more closely to A, munda Chamberlin (19196, pp. 258-259), but differs in the number of teeth on the maxillary plates. The type is in the American Museum of Natural History. Genus DRILONERIES Claperede. E. Clap6r6de, Les Annelides Chaetopodes, etc., 1870, p. 399. Body elongated, slender, without prostomial appendages. Parapodia with rudi- mentary dorsal cirri, but frequently in anterior somites the parapodium reduced or practically absent. No compound or hooked setae. Prostomium always very much flattened dorsoventrally, so that vertical diameter at the base is little greater than at apex. Maxilla with long, slender carriers, forceps, and three or four pairs of plates, the latter feebly developed as compared with Arabella. Mandible absent or rudi- mentary. Drilonereis can usually be distinguished from Arabella, to which it bears a very close resemblance, by the peculiar flattened prostomium (see plate 7, fig. 14). Drilonereis lumbricus, new species. Plate 7, figures 13 to 15; plate 8, figure 10; text-figures 63 to 65. Individuals of this species are large for the genus Drilonereis, measuring 150 mm. in length, with a prostomial width of 1 mm. and a body diameter of 2 mm. in the widest part. Apparently the diameter becomes smaller toward the posterior end, but the single specimen at my disposal was too badly preserved posteriorly to be certain on this point. The prostomium (plate 7, fig. 13) has an oval outline as seen from above and is narrower at the posterior end, where it fits into the anterior margin of the peristomium. In preserved material the prostomium is bent ventrally so as to make an angle with the main axis of the body. On its median dorsal line is a relatively deep depression extending nearly the whole length of the prostomium. The peristomium is short on the dorsal surface, but extends forward on either side, so that the lateral length is more than double that of the dorsal. Ventrally it is thrown into a number of folds (plate 7, fig. 14). Near the posterior margin on the dorsal surface is a depression, the nuchal organ. The anterior somites for about one-quarter of the whole body are smooth, highly iridescent, and greatly resemble an earthworm in general appearance. Behind this region the body-color is a dirty brown, but this may have been in part due to imperfect preservation. Apparently the pygidium is very narrow. The first setae arise in a tuft on the side of somite 3, but the first appearance of anything that could be called a parapodium is on somite 30. Behind this region parapodda are clearly to be seen, but are never very prominent. Anteriorly each (plate 7, fig. 15) has a posterior lobe, a single stout acicula, and a tuft of simple bilimbate setae curved at the end. Owing to poor preservation no satisfactory preparation of the parapodia could be obtained, and the figure from the forty-fifth somite is the best I could get. Farther posteriorly the acicula becomes relatively smaller, the shafts of the setae elongate, the bilimbate 1 62 Leodicidce from Fiji and Samoa. setae become more numerous, and a second, smaller form of seta appears. In a pos- terior parapodia there are 6 of each form of seta. The bilimbate sets (text-fig. 53) have long shafts, are noticeably curved toward the apices, with a striated wing which is wider on the convex side of the curve. It was not always possible to see this JDilim- bate structure, but this may have been due to the position the seta assumed in the preparation. The second form of seta (text-fig. 54) may have a base nearly or quite as broad as the bilimbate, but they narrow rapidly and terminate in a very slender sharp apex which barely protrudes from the surface of the setal lobe. The acicula is very large (plate 7, fig. 15, and text-fig. 55). The jaw (plate 8, fig. 10) is jet-black. The carriers are long and slender, the forceps has a heavy basal portion with teeth on the inner margin of each half, the terminal portion strongly hooked. The proximal pau-ed plates, each with four teeth, lie inside the curves of the forceps. There are two pairs of distal plates, each with one tooth. The mandible is represented by a pair of black plates lying in the wall of the pharynx considerably in front of the maxilla. There is a small plate attached to the ventral surface of the maxilla. This is really darker in color than is indicated in the figure, where it is shaded lightly so as to be more readily seen. One specimen collected in Suva Harbor, Fiji. The type is in the American Museum of Natural History. Drilonereis paucidentata, new species. Plate 7, figures 16, 17; plate 8, figure 11; text-figure 66. Two individuals were collected, both very slender and very much elongated. One was found in Suva Harbor, Fiji, and one in Pago Pago Harbor, Samoa. The following description and figures are taken from the Samoan mdividual, which is incomplete but has about 450 somites. Its prostomial width is not over 0.25 mm. and the average somite length is 0.3 mm. I have designated as the type the specimen from Suva, which is nearly twice the size of this but also lacks the posterior end. The prostomium (plate 7, fig. 16) is relatively rather large and is bluntly rounded. It is only a little narrower than the average body somite and is about as long as somite 1. As is characteristic of this genus, the prostomium does not, as in Lumbrinereis , thicken from the apex toward the base, but is of nearly uniform thickness throughout, the vertical diameter being about half that of the first somite (compare fig. 14 of D. lumbricm). In life the body has a yellowish tint which is most noticeable toward tlie anterior region, while posteriorly the intestinal contents give the body a gray tint. Preserved material shows a transverse brown band in the middle of each somite, but this is apparently due to coagulated blood and appears in so many of the Lumbriner- einse as to have little diagnostic value. _ The parapodia begin on somite 3 and are at first very small. They increase m size posteriorly but never become very prominent. Each has when fully developed a rounded setal lobe and a finger-shaped posterior lobe. Between the two on the ventral surface a heavy acicula protrudes to a considerable distance beyond the surface of the parapodium (plate 7, fig. 17). The setae are all of the same kind, differing only in the length of the shafts. Toward the end each seta broadens and bends and narrows rapidly to an acute tip. There is little distinction to be made between a central shaft and a wing (text-fig. 56). , j • The jaws (plate 8, fig. 11) are jet-black. The maxilla has 2 long, slender earners; the basal portion of the forceps is short and without teeth on the inner margins; the terminal portion is relatively large. Each proximal paired plate has 5 teeth, each of the second pair has 1 long and 2 short teeth, while each of the third pair has 1. A dark triangular plate is attached to the ventral face of the carrier, but I saw no man- dible. The type is in the American Museum of Natural History. Leoclicidce from Fiji and Samoa. 163 Genus OENONE Savigny. J. C. Savigny, Systfeme des Ann^lides, etc., 1820. Prostomium with 3 short tentacles which may be covered by the anterior border of the peristomium. Two lobes of the dorsal surface of the peristomium may be pro- truded so as to cover the prostomium or be retracted into pits. Two pairs of prostomial eyes. The dorsal cu-ri are flattened plates. Maxilla with long, slender carriers and 2 series of toothed plates which may or may not be symmetrical on the two sides. Mandible short and broad. Setae all simple, in a vertical row between the two lobes of the parapodium. Oenone fulgida Savigny. Text-figures 57 to 64. Aglaura fulgida Savigny, 1820, p. 55, pi. 5, fig. 2. Oenone lucida Savigny, 1820, p. 56, pi. 5, fig. 3. Aglaurides fulgida Fauvel, 1917, p. 240-254, pi. 6, figs. 52 to 55. For a diagnosis of the species and a full Uteraturelist see the paper by Fauvel. Collected at Aua Reef and near the governor's wharf at Pago Pago Harbor, Samoa. The living animals are yellowish brown in color and very iridescent, but have no Text-Figure3 57 to 62. .57 to 59. Oenone fulgida horn Samoa. 57, anterior end X 5; 5S, tenth parapodium X 5; 59, one-hundredth parapodium X 5. 60 to 62. Oenone diphyllidia from Tobago. 60, anterior end X 5; 61 tenth parapodium X 5; 62, one-hundredth parapodium X 5. special pigment markings. They are very active in confinement and will crawl out of uncovered dishes. The prostomium is broadly rounded, with 3 tentacles and evertible nuchal organs which appear when the animal is moving. There are two pairs of eyes, the outer ones larger than the inner. After preservation the body-pigment turns to a brownish purple which is darker in some individuals than in others and in all cases is of a lighter tint ventrally than dorsally. In some cases in the preserved material this pigment completely obscures the eyes, while in others the smaller pair only are invisible. It seems probable that this condition is responsible for confusion in the classification of this and related species, some of which have been described as having only one pair or no eyes. The peristomium is wider than the prostomium and usually is a Uttle wider than the first setigerous somite. In some individuals there is a faint constriction, especially noticeable on the lateral and ventral areas, which obviously marks the boundary between the first and second somites. By a number of writers the possession of one or two apodous somites has been regarded as of unportance in the separation of genera. If this distinction really occurs it would be important as indicating that the second apodous somite of one genus is homologous with the first setigerous of another. In my study of the Leodicidae I have found so much variability between closely allied 164 LeodicidcB from Fiji and Samoa. species of the same genus in the degree of distinctness of the first two somites that I very much doubt whether there is anywhere a Leodicid with only one apodous somite. While this assertion is perhaps going beyond the evidence in such a genus as Onuphis, it seems to me obvious that these lateral and ventral markings in Oenone indicate clearly a fusion of somites and that a classification which separates this genus from others because of the (apparently) single apodous condition is erroneous. The appearance of the ventral prostomial surface varies with the amount of pro- trusion of the palpal lobes, but it usually shows some longitudinal wrinklings. In other respects there Is nothing to be added to the descriptions already given by Fauvel and others. The genera of the Lumbrinereinse with foliaceous dorsal cirri are somewhat confused, The first genera and species were described by Savigny (1820, pp. 55, 56, plate 5, figs. 2 and 3) as Aglaura fulgida and Oenone lucida. The type of the latter species and genus was afterward found to be an immature in- dividual belonging to Aglaura fulgida, so that the latter was regarded as the valid genus and species. Ehlers (1864-1868, pp. 407, 408) showed that the name Aglaura was preoccupied and proposed instead the generic name Aglaurides. Savigny's original description of Oenone might be interpreted to read that the lack of tentacles was characteristic of the genus, and Ehlers redefined Oenone as having no tentacles and 1 apodous somite, while Aglaurides has 3 tentacles and 2 apodous somites. In other respects the two genera are alike. Fauvel (1917, pp. 240 to 257) accepts Aglaurides as the valid name, arguing that since Oenone was founded on an evident error it should disappear from the literature, and I followed this procedure in a paper on the West Indian Leodicidse (Treadwell, 1921a, p. 116). In a personal letter Dr. Chamberhn pointed out that this procedure is contrary to the rules of nomen- clature in that when Aglaura was discovered to have been preoccupied, Oenone, the oldest recorded synonym, should take its place. The criticism is valid and I have used Oenone accordingly. Gravier (1900, p. 222) accepts Aglaurides as defined by Ehlers and retains Oenone to include species without antennae or nuchal organs. Augener (1913, pp. 290, 291) and Chamberlin (1919a, p. 326) use Oenone as the only valid genus for this group, but neither writer seems to attach much importance to the tentacles and apparently they confuse the tentacles with the nuchal organs. Chamberlin (1919a, p. 337) speaks of "obscure antennal nodules — apparently subject to retraction like true antennal organs," and Augener (p. 290), in speaking of the "sogenannten3Fuhler,"says"Diese FuWer sind ohne zweifel keine fuhlerartigen Anhange des Kopfes in gewohnlichen Sinne sondern als Nackenorgane aufzufassen." Chamberlin (1919a, p. 335, plate 62, figs. 2 to 5) describes as Oenone telura a species without tentacles, with peculiarly shaped somites around the mouth, and with a maxillary apparatus quite unlike any thus far figured in other species. Chamberlin states that the specimen had apparently been dried, which would lead to a distortion of the soft parts, and its resemblance in some details to Oenone fulgida would lead to a suspicion that the drying had been responsible for the apparent lack of tentacles. Dr. Chamberlin kindly offered to reexamine the specimen to be certain on this point, but it proved to be inaccessible and so the matter can not be determined. So far as I know, this is the only recorded case of an ^^ Oenone" without tentacles, and since the presence or absence of tentacles Text-Figures 63 and 54. Jaws of Oenone. 63, jaw of Oenone fulgida from Samoa X 40; 64, jaw of Oenone diphyllidia from Tobago X 40. Lcodicidcc from Fiji and Sofnoa. 165 is obviously of generic importance, a new genus should be created for the reception of this species telvra. I have collected Oejioyie in Bermuda, Tobago, and the Dry Tortugas in the Atlantic and in Samoa in the Pacific. In a description of the West Indian species I suggested (Treadwell, 1921a, pp. 118, 119), that the species described by Ehlers (1887, pp. 109- 111, plate 34, figs. 1-7) as Oenone diphyllidia of Schmarda (1861, p. 120, plate 32, fig. 256) was really Oenone fulgida Savigny, and that those of my own collections were Aglaurides {Oenone) diphyllidia Schmarda, which might be identical with Aglaurides (Oenone) symmetrica of Fauvel (Fauvel, 1917, p. 252). Augener (1913, p. 290), who apparently had access to Ehlers'd collections, reported that he had identified as Oenone fid^ida specimens from south Australia, and that a comparison of these with Ehlers's West Indian specimens showed that the two are identical. As will be noted imme- diately, the jaw structure is important in this connection, and it is perfectly possible that Augener did not examine this organ, but made his comparisons entirely from surface features. I have made a careful comparison of my West Indian with my Samoan specimens and find that they agree in every respect except the form of the jaws. Text-figures 57, 58, and 59 are outline drawings of the anterior end, the tenth and the one-hundredth parapodium of the Samoan Oenone, while text-figures 60, 61, and 62 are corresponding drawings of specimens from Tobago in the West Indies. In other details, such as size, coloration, and habits, they are alike. It seems obvious that, so far as these structures are concerned, there are no differences of specific value between the animals from the two localities. Reexamination of my Bermuda material shows that the acicula I figured (Treadwell, 1921a, text-fig. 452) was not typical and that the aciculse agree with those drawn in Fauvel's figure 52. On the other hand, there are decided differences in the form of the jaw, as shown in text-figures 63 and 64. The jaw shown in text-figure 63 agrees in all essentials with Fauvel's description if due allowance is made for a slight difference in the position of the plates. In profile the teeth look large and sharp, but when rolled so as to be seen in full face they have the appearance shown by Fauvel, and in one specimen the second and third plates of the right-hand series had more teeth than here represented. The small plate adjoining the right-hand end of the carrier seems to me to be attached to the larger one and its teeth are to be seen only in strong reflected light, appearing then as bright spots. My Samoan specimens are evidently Oenone fulgida Savigny, which is synonymous with Oenone diphyllidia Ehlers, while the ones I have seen from the West Indies are 0. diphyllidia Schmarda. It seems probable that this latter species is synonymous with 0. symmetrica Fauvel. If my suggestion that all of the Leodicidse have two apodous somites is correct, and if it be remembered that in Oenone either one or both pairs of eyes are not to be seen in preserved material, the distinction usually made between Halla and Oenone (see Gravier, 1900, p. 322, and Chamberlin, 1919a, p. 326) would appear of doubtful accuracy. I have examined a specimen bought of the Naples Zoological Station as Halla parthenopeia and two specimens identical with this belonging to the American Museum of Natural History but without data, and find that they are certainly the species described by Ehlers (1864-1868, p. 408, plate 17, figs. 25-34) as Cirrobranchia parthenopeia. A second pair of eyes apparently escaped Ehlers's attention and he does not mention the protrusible lobes which distinguish his genus Aglaurides from Cirro- hranchia. Dissection of these specimens, however, shows the lobes lying under the peristomial border exactly as in Oenone. It seems probable that Ehlers really had a species of Oenone and probable that Halla was originally described from a member of this genus. The matter could, I suppose, be settled only by reference to the original type specimen, but I am skeptical as to the validity of Halla. Cirrobranchia, as Cham- 166 LeodicidcB from Fiji and Samoa. berlin and others have noted, is synonymous with Jlalln, which has precedence, if Halla is a valid genus. Subfamily DORVILLEIN^. The characteristic genus of this family was named Staurocephalus by Grube (1855, p. 97), but Verrill (1900, pp. 647, 648) showed that the name was preoccupied and renamed it Stauronereis, making the subfamily Stauronereina^. Chambcrlin (1919flr, pp. 338, 339) showed that Stauronereis was preoccupied by Dorvillea, given by Parfitt (1866, pp. 113, 114, with 5 figures) to a new genus Dorvillea. The name of the prin- cipal genus should then be Dormllea and the subfamily renamed accordingly. Chamberlin is in error in referring to a swarming like that of Palolo in Dorvillea. Mayer (1902) corrected an error in an earlier paper and pointed out that this swarming species is Leodice fucata Ehlers. Genus DORVILLEA Parfitt. Parfitt, E, 1866, Zoologist, 2d series, pp. 113, 114. Prostomium rounded, pentagonal or quadrangular, with two more or less articulated tentacles and elongated palps which may be spirally contorted. Body with relatively few somites, parapodia with dorsal and ventral cirri but without gills. Four anal cirri. Maxilla of 2 or more rows of toothed plates on either side, the rows all united at the base but diverging in a V-shape. Mandible bifurcated, with slender shafts, the margin often prolonged laterally into rows of plates. Dorvillea australiensis Mcintosh. Plate 8, figures 1 to 7; text-figures 65 to 68. Staurocephalus australiensis Mcintosh, 1885, pp. 232, 233, pi. 36, fig. 6; pi. 17a, figs. 9 and 10. Staurocephalus australieusis Treadwell, 1906, p. 1173, figs. 63 to 66. Dorvillea {Staurocephalus) australiensis was described by Mcintosh from a posterior fragment of a single individual. Treadwell identified with this a species from Hawaii and figured the prostomium with appendages in which the tentacles are shown as unsegmented. Augener (1913, pp. 293-296) and Benham (1915, pp. 209-212, plates 41, figs. 58 to 66) described specimens of this genus from the Australian region as S. australiensis, though all of the tentacles in their specimens had strongly articulated tentacles. I have reexamined the Hawaiian specimen (now No. 5463 in the U. S. National Museum) and find that the tentacles certainly are not strongly articulated, though they show a jointing toward the end. The only structures in which direct comparison is possible between Mcintosh's and Benham's specimens are in the parapodia and setse. The difference between the two figures of the parapodia (Benham, plate 41, fig. 62, and Mcintosh, plate 36, fig. 6) might be due to imperfect preservation of the material or to the fact that they represent parapodia from different regions of the body, but this explanation does not hold for the setff. Benham figures the terminal joints of the compound setse as each having a stout subapical tooth and a denticulated margin to the hood, and he describes the simple setae as "long, curved, capilliform" with fine serrations along the upper convex margin, while in Mcintosh's figures (plate 17a, figs. 9, 10) the terminal joint of the compound seta has a small subapical tooth and no serration along the margin of the hood. The simple seta has a serrated edge and terminates in a bifid extremity. In the form of the setal lobes and setse Benham's specimens differ from the Hawaiian and the Samoan species. He states that the mandibles are without denticulations, while in the Samoan specimens they are denticulated (plate 8, fig. 7). The form of the paragnaths is quite unlike in the two cases. Augener's specimens from Australia had on the parapodia "am Ende 3 blatt-formige Lippen-einc vordere oberc und eine LeodicidcB from Fiji and Samoa. 167 hintere mediane," while Mcintosh's description of the type is that "superiorly the free edge of the foot presents two prominent mammillse, between which the bristles of the region emerge." This is not in entire agreement with his figure, but is quite in accord with the conditions found in the Samoan and Hawaiian specimens and quite unlike those described by Augener. Augener followed Benham in his identification of his specimens with those of Mcintosh and Tread well, explaining the differences as due to poor preservation. It seems to me evident that the species described by Benham and Augener are not the same as the ones I have seen from Hawaii and Samoa, which I regard as belonging to Mcintosh's species. Several specimens were collected on Aua reef in Pago Pago Harbor in rocks near the upper end of the harbor and one on the reef at Aunuu. They are very sensitive to changes in the water and most of them died before they could be got to the laboratory. In the living animal the prostomium has a faint pink color on the anterior margin, but this fades out on the dorsal surface. The palps are colorless or only faintly tinted with pink. The peristo- mium is pink, this color being most intense along the posterior edges of the antero-lateral depressions and on the "caruncle." The first 6 somites show a pink color which is most intense along the anterior margins, but behind the region of somite 6 the pink color disappears and is replaced by a decided yellow. No color persists in the preserved material. The prostomium is rounded (plate 8, fig. 1), not more than one-third the diameter of the peristomium. The palps are large, thick at the base, unjointed, and with rather blunt tips. The tentacles are less than half the diameter of the palps at the base and not more than three-quarters as long as they, and are jointed only toward the apices. There are two pairs of eyes — the dorsal ones the smaller, the larger ventral ones visible from the dorsal surface only through the translucent bases of the tentacles. Both pairs of eyes are very black. The peristomium is rectangular in outHne, but with a deep depression on either side on the dorsal surface. The median ridge between these two depressions is continued forward to join with a knob which belongs to the prostomium, the whole forming a sort of caruncle. The first parapodium (plate 8, fig. 2) has a rounded setal lobe, with lips equal in length. The dorsal cirrus is 2-jointed, the basal joint much the longer. The ventral cirrus is rather large, joined for about half its length to the setal lobe. There is a heavy acicula in the setal lobe and a tuft of needle setae in the dorsal cirrus. There are two tufts of setse, the dorsal ones simple and the ventral ones compound. In the tenth parapodium (plate 8, fig. 3) the parts are all larger, but the relative forms arc about as before, except for an increase in the vertical diameter of the setal lobe. There are a dorsal and a ventral tuft of setse with acicula? as in the first parapodium. The pygidium is rounded, with one pair of long anal cirri (plate 8, fig. 4). The dorsal setae are long and curved, with minute denticulations along the convex margin and very small terminal teeth at the end (text-fig. 65). Ventral to these is a row of stouter setse with serrated convex margin and apical and subapical teeth covered by a hood (text-fig. 66). Ventral to these are compound setse having smooth basal joints; the terminal joints have each a large apical and a smaller subapical tooth, the whole covered by a hood (text-fig. 67). The maxilla has the form characteristic of this genus, composed of two rows of plates on either side, with from 35 to 40 plates in each row, the rows from opposite Text-Figures 65 to 6S. DorviUca auslraliensis. 65, dorsal simple seta X 250; 66, ventral hooded seta X 250; 67, compound seta X 250; 68, one half of maxilla X 12.5. 168 Leodicidce from Fiji and Sainoa. sides united in pairs by a V-shaped connection (text-fig. 68 shows one-half of the jaw;. In the basal portion of each row the plates are closely united, looking like a series of vertebra. At about the middle of the series (plate 8, fig. 6) the plates reach their greatest development. Each has a bifurcated base and a rather heavy fang. Along one edge of the fang is a plate-like protrusion with a row of teeth along its margin. In the view drawn the teeth appear as if on the margin of the fang, but in reality they are on a plate set at an angle to the fang and pointed away from the observer. At the apex of the row the plates become smaller and the teeth disappear (plate 8, fig. 5). The inner row is a little shorter than the outer. The mandible (plate 8, fig. 7) is black, each half with teeth along the anterior margin and small plates lateral to the apex. Apparently the number of these plates is not constant. BIBLIOGRAPHY. AuDouiN, J. v., et A. Milne-Edwards. 1834. R^cherches pour servir k I'histoire du littoral de la France. Ann^lides. v. 2, pp. 1-290, pis. 1-8. AuGENER, H. 1813. Die Fauna Siidwest Australiens, Ergebnisse der Hamburger sudwestaustralischen Forschungreise 1905. Polychaeta 1. Errantia, pp. 65-304, pis. 2, 3, 42 text-figs. Benham, W. B. 1915. Report on the Polychaeta obtained by the F. I. S. Endeavor on the coast of New South Wales, Victoria, Tasmania, and South Australia. Zoological results of fisliing experiments carried on by the F. I. S. Endeavor 1900-1904. Pt. 1, pp. 173-237, pis. 38-45. Blainville, H. M. de. 1828. Dictionnaire des Sciences Naturelles. T. 67, Articles Nereides et Vers. Chamberlin, R. V. 1919a. The Annelida Polycha;ta, pp. 1-514, plates 1-80. Reports on an exploration off the west coast of Mexico, Central and South America, and off the Galapagos Islands, in charge of Alexander Agassiz, by the U. S. Fish Commission steamer "Albatross" during 1891, Lieut.-Commander Z. I. Tanner, U. S. N., commanding, xxxviii. Reports on the scientific results of the expedition to the tropical Pacific, in charge of Alexander Agassiz, by the U. S. Fish Commission steamer "Albatross," from August 1899 to March 1900, Commander Jefferson F. Moser, U. S. N., commanding, xx. Reports on the scientific results of the expedition to the eastern tropical Pacific in charge of Alexander Agassiz, by the U. S. Fish Commission steamer "Alba- tross" from October 1904 to March 1905, Lieut-Commander L. M. Garrett, U. S. N., commanding, xxxi. 1919b. Pacific Coast Polychaeta collected by Alexander Agassiz. Bull. Museum of Comparative Zoology of Harvard College, vol. 65, No. 6, pp. 251-270, pis. 1-5. ClaperIide, E. 1870. Les Ann6hdes ch^topodes du golfe de Naples. Mem. Sec. Phys. et Hist. Nat. de Gen6ve, vol. 20, pp. 305-542, pis. 1-8. Collin, Ant. 1897. Bemerkungen liber den essbaren Palolowurm, Lysidice viridis (Gray). Anhang zu Kramer; Ueber den Bau der Korallenriffe und die Planktonvertheilung an den Samoanischen Kusten. Crobsland, Cyril. 1903. On the marine fauna of Zanzibar and British East Africa, from collections made by Cyril Crossland in the years 1901 and 1902. The Polychaeta; Part ii, pp. 129-144, pis. 14 and 15, text-figs. 12-15; Proc. Zool. Soc. London, 1903, vol. 2. 1904. On the marine fauna of Zanzibar and British East Africa, from collections made by Cyril Crossland in the years 1901 and 1902. The Polychaeta; Part iii; with which is incorporated the account of Stanley Gardiner's collection made in the Maldive Archipelago in the year 1889, pp. 287-330, plates 20-22, text-figs. 43-46. Proc. Zool. Soc. of London, 1904, vol. 1. Ehlers, Ernst. 1864-68. Die Borstenviiirmer (Annelida chatopoda) nach systema- tischen und anatomischen imtersuchungen. pp. 1-748, pis. 1-24. 1887. Florida AnneUden. Reports on the results of dredging under the direction of L. F. Pourtales during the years 1868-70, and in the Caribbean Sea (1878- 79) in the U. S. Coast Survey Steamer "Blake." Memoirs Museum of Com- parative Zoology at Harvard College, 15; pp. vi +1-328, pis. 1-60. 1898. "Oeber Palolo Eunice viridis (Gray). Nachrichten der K. Gesellschaft der Wissenschaft zu Gottingen, Math-phys. Klasse, Hft. 4, pp. 1-16. 1904. Neusselandische Anneliden. Abhandlungen der K. Gesellschaft der Wissenschaften zu Gottingen, Math-phys. Klasse, Neue Folge, Bd. 3, No. 1, pp. 1-79, pis. 1-9. ^ Fauvel, Pierre. 1914. AnneHdes polychaetes de San Thom6 (Golfe de Guin^e) receu- illes par M. Chas. Gravier. Archiv. de Zoologie Exp^rimentale et G^neralc, T. 54, fasc. 5, pp. 105-155, pis. 7, 8. 1917. Ann61ides polychetes de I'AustraHe m6ridionale. Archiv. de Zoologie Exp^rimentale et G6n6rale, T. 56, fasc. 3, pp. 159-277, pis. 4-8, text-figs. 1-29. 1919. Ann6Hdes Polychetes de Madagascar, de Dibouti et du Golfe Persique. Archiv. de Zoologie Exp^rimentale et G6n^rale, T. 58, pp. 315-473, pis. 15-17, text-figs. 1-11. Friedlaender, Benedict. 1898. tleber die so-genannteu Palolowurm. Biologisches Centralblatt, Bd. 18, pp. 337-357. 169 170 Leodicidm from Fiji and Samoa. Gravier, Ch. 1900. Contribution h I'^tude des Ann^lides Polych&tes de la Mer Rouge. Nouvelles Archives du Museum de Hist. Nat. 4 S6rie, vol. 2, fasc. 11, pp. 137-282, pis. &-11. ^ .^ Grubb, Adolph-Eduard. 1851. Die FamiLen der Anneliden, mit Angabe ihrer Gafc- tungen und Arten. pp. 1-164. 1855. Beschreibung neuer oder wenig bekannter Annebden. Archiv. f. Natur- gesch. 21, p. 97. . , ^ ,. , , 1878. Annulata Semperiana. Beitrage zur Kenntmss der Anneudenfauna der Phibppinen. M^moires de I'acad^mie imp^riale des sciences de St. P^tersbourg, 7th series, vol. 25, pp. 1-300, pis. 1-15. Johnson, H. P. 1901. The Polychseta of the Puget Sound region. Proceedings of Boston Society of Natural History, 29, pp. 381-437, pis. 1-19. KiNBEBO, J. G. H. 1864. Annulata nova (Eunicea). Ofvers af K. Vetensk. Akad. Forhandlung, vol. 21, pp. 659-574. Macdonald, J. D. 1858. On the External anatomy and Natural history of the genus of annelids named Palolo by the Samoans and Tonguese, and Mbalolo by the Fijians. Trans. Linnean Society of London, 22, pt. 3, No. 16, pp. 237- 239, pi. 41. McIntosh, W. C. 1885. Report on the AnneUda Polychseta collected by H. M. S. Challenger during the years 1873-1876. Report on the Scientific Results of the Voyage of H. M. S. Challenger, vol. 12, pp. i-xxvi + 1-554, pis. 1-55, 1A-39A. V. Marenzeller, Emil. 1879. Sudjapanische Annebden. Denkschriften der Mathe- matisch-Naturwissenschaftlichen Classe der Kaiserhchen Akademie der Wissenschaften. Vienna, Bd. 41. pp. 1-45 (of separate), pis. 1-6. Mater, A. G. 1902. The Atlantic Palolo. Science Bull., Museum Brooklyn Inst. Arts and Sciences, vol. 1, No. 3, pp. 93-103, 1 pi. Moore, J. Percy. 1903. Polychseta from the coastal slope of Japan and from Kam- chatka. Proceedings of the Academy of Natural Sciences of Philadelphia 55, pp. 401-490, pis. 23-27. 1904. New Polychseta from CaUfornia. Proceedings of the Academy of Natural Sciences of Philadelphia 56, pp. 484-503, pis. 37, 38. 1909. Polychaetous AnneUds from Monterey Bay and San Diego, California. Proceedings of the Academy of Natural Sciences of Philadelphia, 61, pp. 235- 295, pis. 7-9. 1911. The Polychaetous AnneUds Dredged by the U. S. S. Albatross off the Coast of Southern California in 1904. III. Euphrosynidse to Goniadidse. Proceedings of the Academy of Natural Sciences of Philadelphia, 63, pp. 234- 318, pis. 15-21. Pallas, P. 1788. Marina varia nova et rariora. Nova Acta Acad. Scientiar. Imper. Petropolitanae, Tom. 2, p. 229, pi. 5, figs. 1-7. Parfitt, Edward. 1866. Description of a Nereis new to science. Zoologist, series 2, vol. 1, pp. 113-114, 5 text-figs. QuATREFAQEs, M. A. DE. 1865. Histoire naturelle des Annelds marins et d'eau douce, T. 2, pp. 1-794, pis. 1-20. Savignt, J. C. 1820. Syst&me des Ann^lides principalement de celles des cotes de 1' Egypt et de la Syrie. ScHMARDA, LuDwiQ K. 1861. Neue wirbellose Thiere beobachtet und gesammelt auf einer reise urn der Erde 1853 bis 1857. Band 1, Hft. 2, pp. 1-164, pis. 16-37. Stair, J. B. 1847. An account of Palolo, a sea-worm eaten in the Navigator Islands; with a description by J. E. Gray. Proceedings of the Zoological Society of London, part 15, pp. 17, 18. Treadwell, a. L. 1906. Polychsetous Annelids of the Hawaiian Islands collected by the steamer Albatross in 1902. Bulletin U. S. Fish Commission for 1903, pt. in, pp. 1145 to 1181. 1921a. Leodicidse of the West Indian Region. Carnegie Inst. Wash. Pub. No. 293, pp. 1-129, pis. 1-9, text-figs. 1-^67. 19216. Report on the anneUds of Puget Sound, Fiji, and Samoa. Carnegie Institution of Washington, Year Book No. 19, pp. 199-200. Verrill, a. E. 1900. Additions to the Turbellaria, Nemertinea, and Annelida of the Bermudas. Trans. Connecticut Academy of Sciences, vol. 10, part 2. pp. 695-670; pi. Ixx. WooDWORTH, W. McM. 1907. The Palolo worm, Eunice viridis Gray. Bull. Museum Comparative Zoology at Harvard College, vol. 51, No. 1, pp. 3-21, pis. 1-3. IX. P0LYCH.ET01JS ANNELIDS COLLECTED AT FRIDAY HARBOR, STATE OF WASHINGTON, IN FEBRUARY AND MARCH 1920. By a. L. TREADWELL, Professor of Zoology in Vassar College. Thirty-seven figures. 171 POLYCHAETOUS ANNELIDS COLLECTED AT FRIDAY ; OF WASHINGTON, AND MARCH 1920. HARBOR, STATE OF WASHINGTON, IN FEBRUARY By a. L. Treadwell. The following is a partial report of the results of an investigation on the Pacific annelids, conducted during the season of 1920 under the auspices of the Department of Marine Biology of the Carnegie Institution of Washington, Dr. A. G. Mayor, Director. It com- prises descriptions of certain new species of polychsetes collected by the writer in the vicinity of Friday Harbor, State of Washington, in February and early March 1920, with notes on some previously described. Grateful acknowledgment is made to Professor T. C. Frye, Director of the Puget Sound Biological Station, who put the resources of his laboratory at my disposal. Family SYLLID-S). Autolytus varius Treadwell. AtUolytus varius Treadwell. 1914, New Syllidae from San Francisco Bay, collected by the U. S. S. Albatross. Univ. of Calif. Pub. in Zoology, vol. 13, No. 9, page 237, figs. 4 to 7. The original description was taken from a single preserved specimen. The first ones obtained in 1920 were collected by Mr, J. Little in the herring trap near the ship- yard at Friday Harbor and were swimming at the surface. Later, others were found in the same locality. The anterior end (fig. 1) agreed in structure with the original description, except that the tentacular cirri are more nearly equal to the antennae in size than was there stated. The palps are very small, are directed ventrally, and are not visible from the dorsal surface. They are fused for more than half their length, the free portion being in the form of a blunt cone. The smaller eyes (black) are shown as lying directly over the (stippled) larger ones. The first five somites are not very sharply marked off from one another, but two sets of longitudinal lines extend along the dorsal surfaces of somites 1 to 4. The inner ones mark off median areas which are suc- cessively wider from before backward. Lateral to each is a fainter line running parallel to it. These areas are very clearly outlined, but not much elevated above the general surface. The natatory setse are excessively delicate and are blunt-ended, though the end may turn shghtly so as to be seen in profile, and then look as if sharp-pointed. Behind the brood pouch are 60 or more somites gradually narrowing toward the py- gidium and with successively smaller cirri. There are two anal cirri larger than the dorsal cirri immediately in front of them. The animals vary greatly in size, for while most were 30 mm. long, one of only 10 mm. was carrying a brood pouch with young. In the living animals are two color phases: (1) Light green body-color with a white brood-sac containing eggs in cleavage stages. The coloration is most intense through the median body-region, and is there most marked as a dark-green line crossing each somite toward its posterior end, continued on to the posterior face of the parapodium to the point of attachment of the dorsal cirrus and covering the basal joint of the cirrus. The distal part of the cirrus is white. (2) The color varies through shades of reddish brown, the brood-sac being bright salmon. In these the brood-sacs contained larvae. This color difference is quite independent of the color of the larvae them- 173 174 Polychcetous Annelids Collected at Friday Harbor, selves, but may perhaps be influenced by their state of development. I was, how- ever, unable to discover any constant relation between depth of coloration and degree of development of the larvae. The oldest larvae I found were 0.75 mm. long and had two setigerous somites (fig. 2). The anterior margin of the head is rounded and has two pairs of sharp spines, one ven- tral to the other. There are two pairs of eyes, the ventral ones being represented in the figure by stippling, as showing through the tissues of the head. There are two bands of ciUa around the head, and behind these a tuft of cilia on either side. It is possible that here, also, is a continuous band, but I was unable to determine whether this is the case. Two setigerous somites follow these, with very minute setae apparently like the compound ones previously described. A band of cilia surrounds the body near the posterior end. FamUy PHYLLODOCIDiE. Eteone maculata, new species. A single specimen of over 300 somites, the posterior end regenerating. The total length was 75 mm., width of prostomium 0.5 mm., greatest body-width including the parapodia 3 mm. The prostomium (fig. 3) is rather broadly rounded at its anterior margin, with the lateral margins sloping slightly outward posteriorly, the diameter at the posterior margin being thus about one-third greater than at the anterior. In the dorsal median line of the posterior margin is a blunt, backwardly directed lobe. There are 4 small triangular tentacles. Dorsally the first somite is longer on either margin than in the mid-line. It bears on either side 2 conical, rather thick, cirri of which the ventral ones are the larger. The second somite is about two-thirds as long as the first, its parapodia being placed more ventrally than is the case in succeeding somites. The flattened dorsal cirri characteris- tic of this family are very small on somite 2 and gradually increase backward, reaching their full size on about somite 10. Throughout they are rather thick, but never very large or prominent. In alcohol the general body-color is yellow, with the cirri, especially posteriorly, a distinct golden color. Beginning with somite 4, each somite except 9 bears on its dorsal surface a prominent spot or spots. If there is but one it usually lies in the median line; if two, they tend to lie one on either side, just dorsal to the parapodium. If three, one may lie in the median line and one over each parapodium. It seems probable that these are highly variable in distribution, but material is lacking for the determination of this point. Similar spots, though much smaller and more variable in their distribution, occur on the ventral surface. No eyes are visible on the pre- served material. The first parapodium (fig. 4) has the characteristic form, but the parts are all very small (cf. fig. 4 with 5). There is a rounded setal lobe and a ventral cirrus situated postero-ventrally to the setal portion and united to it for almost its entire length. The twentieth parapodium (fig. 5) has an entire anterior and a bifid posterior lip to the setal lobe. As before, the ventral cirrus is largely fused with the setal portion, while the large dorsal cirrus is more nearly free. The setae are all compound and similar, with long basal joints (fig. 6). The terminal joint varies somewhat in width, but in other respects those of a bundle are all alike. Each is flattened, tapers to an acute point, and may be more or less bent. Collected at Friday Harbor. Type in the American Museum of Natural History. Eteone tuberculata, new species. The type specimen is 106 mm. long, with a width of 1 mm. at the posterior margin of the prostomium. The width of body, including parapodia, is 4 mm. The prostomium (fig. 7) has a rounded anterior margin between the bases of the tentacles, but the lateral areas, where the tentacles are attached, are straight and State of Washington, in February and March 1920. 175 parallel to one another. Behind the bases of the tentacles each lateral margin slopes gently outward, so that the posterior margin of the prostomium is about equal to its antero-posterior diameter. The posterior dorsal surface has a distinct tubercle situated in a rather deep backward indentation of the anterior margin of the peristomium. This tubercle may possibly be regarded as homologous to the third tentacle found in the genus Eulalia, but in character of the setse this specimen differs widely from the latter genus. The tentacles are short-lanceolate and equal in size. On its anterior margin the peristomium is wider than the prostomium and widens still more posteriorly, so that its posterior margin is about twice as wide as the pro- stomium. The anterior margin is incurved, forming a bay in which the tubercle on the prostomium lies, and the 4 equal-sized tentacular cirri are on the lateral surfaces. Ventral to these the peristomium is noticeably wider than it is on the dorsal surface. The second somite is about half as long as the first, and succeeding somites increase in width up to about the ninth. The mouth is bounded dorsally by the prostomium, laterally and ventrally by the first somite. The dorsally directed anus has a rounded lip and there is one pair of short and rather stout anal cirri. The first parapodium is on somite 2 and has rounded postsetal and presetal lobes and a well-developed ventral cirrus (fig. 8), the dorsal cirrus being absent. There is a single acicula, and a row of compound setae with rather short terminal joints. In later parapodia the dorsal cirrus appears as a rather thick but not very large lobe. It is small on somite 2 but increases in size farther back. A well-developed parapodium is shown in figure 9. The presetal lobe is notched, while the postsetal is rounded. In preserved material all of the dorsal cirri are deep brown in color, being especially in the smaller specimens much darker than the general body-color. There is a sub and a supra tuft of cilia, the former (fig. 10) with a slender basal portion enlarged and toothed at the end. The terminal joint is flat, and at the base is as broad as the apex of the basal portion, but it rapidly narrows to an acute apex with a row of minute den- ticulations along the concave margin. The supra-acicular bundle is composed of similar setae, but the terminal joint is longer and more slender. Collected at Friday Harbor. Type in the American Museum of Natural History. Family LEODICIDiE. Lumbrinereis zonata Johnson. Lumhriconereis zonata Johnson, 1901. The Polychseta of the Puget Sound Region. Proc. Boston Soc. Nat. Hist. 29; pp. 408 to 409, pi. 9, figs. 93 to 100. This was the commonest annelid in the Friday Harbor collections. Measurements of living annelids, especially members of this genus where there is very great power of contractility, are of little value, but I found that preserved specunens vary in length from 100 to 300 mm. with a prostomial width of not more than 1.5 mm. and a body- width of 2 mm. The general body-color in life is dark brown with a marked iridescence, and it can be distinguished from the only other species of this genus that I collected in this locality by the fact that in the other (L. cervicalis, see below) there is a noticeable brown nuchal band, easily seen with the naked eye, and the body is a lighter brown. Under the hand lens (in L. zonata) a transverse band of brownish-gray dots may be seen running across the middle of each somite, and these may be the foundation of the dark transverse bands which appear in preserved material and were the reason for Johnson's specific name. This transverse dark band, appearing only in preserved material, occurs after preservation in a number of Lumbrinereids, so that its value as a specific character is small. The prostomium is not quite as acutely pointed ante- riorly as in Johnson's figures and the anterior peristomial line is straight rather than concave. The following additions are made to Johnson's description. The mandibles (fig. 12) are transparent, except where marked along their cutting edges with dark pigment. This is darkest at the outer angles and diminishes in 176 Polychcetous Annelids Collected at Friday Harbor, amount toward the middle line. The halves are united for almost their entire length. In the maxilla (fig. 11) the carriers are rounded basally but narrowed at about their middle. The forceps are rather slender and only faintly curved. The right proximal plate has 5 teeth, the left 4, the second pair each has 2, the third pair each has 1. All parts of the maxilla are jet-black and there are collections of black pigment in the chitin outside of the plates which are not shown in the figure. There are two pairs of anal cirri (fig. 13) approximately equal in size, the dorsal a little closer together than the ventral. Lumbrinereis cervicalis, new species. Individuals of this species are rather small as compared with others of this genus, the average size being 80 mm. in length and 1 mm. prostomial width. They are easily recognized in life by a noticeable dark-brown band crossing the dorsal surface of the first two somites (fig. 14) and by the yellowish body-color. Visible under a lens are minute brownish lines running longitudinally along the dorsal surface of the prosto- mium. These are variable in size and color. The prostomium in living animals is rounded and usually a little broader than long, though in preserved material the postero-anterior diameter may be the longer. The first two somites (fig. 14) show dorsally the pigment band, which does not extend on to the ventral surface (fig. 15). Throughout most of the body is a prominent brown spot just ventral to the parapodium in each somite, and other pigment spots are scattered irregularly over the dorsal surface. There are 4 equal-sized, short anal cirri. The parapodium has the form characterisitc of this species (fig. 16 is drawn from the tenth), and there are no noticeable differences in form, except that the posterior ones are more slender. In the anterior parapodia are both simple and compound setae. The dorsalmost are simple, each long, curved, and fine-pointed, with a wing along its convex margin (fig. 17). Toward tlie ventral surface are others of similar form, but shorter and less noticeably curved. The compound setie (fig. 18) lie in the middle of the tuft. The terminal joint has a number of minute teeth at the apex, which is covered with a hood. A continuation of this hood extends for a short distance down the basal portion, but at the joint between the basal and apical portions is a slight constriction, as if the hood were divided into two parts. Behind about the fortieth parapodium both these forms of setce disappear and their place is taken by a second form of simple seta resembling the compound ones of anterior somites if the two parts of these had fused (fig. 19). The stalk is long and the apex has a number of minute teeth, covered by a hood which extends for a short distance down the shaft. The maxilla is small and is dark brown in color (fig. 20). The carriers are sharp-pointed posteriorly and roughly triangular in form, with a notch in each on the outer face near the anterior end. The forceps are slender and gently curved. The right proximal plate has 5, the left 4 teeth, the second pair has 4 each, and the terminal pair 2. There is more or less pigment in the chitin bordering these plates which is not shoTVTi in the figure. The mandible (fig. 21) is uncolored, except for dark pigment patches on the outer anterior margins. Collected first at Boat Bay and later in Newhall's lagoon at Friday Harbor. It occurs also in hmited numbers in other localities in this vicinity. The tj'^pe is in the American Museum of Natural History. Onuphis stigmatis, new species. Collected in considerable numbers in False Bay, Friday Harbor, in sand exposed at low tide. One specimen was in a tube of sand grains, but most seemed to be lying loose. This was unexpected and it may be that I overlooked the other tubes, though a special effort was made to find them. The single complete specimen of the animal collected at this time was 80 mm. long, with a prostomial width of about 1.5 mm. In the living animal the basal joints of the tentacles show a very little pigmentation, but the remainder of the tentacles, the palps, and the nuchal cirri are colorless. A State of Washington, in February and March 1920. Ill broad band of dark-brown pigment extends across the peristomium and is continued anteriorly into a patch of the same color on either side (fig. 22). It is stated by some writers that the nuchal cirri in Onuphis are carried on the peristomium. This I am convinced is a mistake and I regard the apparent first somite as a fusion of somites 1 and 2. The apparently second somite is thus really somite 3. A very faint pigment line crosses the posterior fourth of this somite, while behind this two bands occur in each somite — one at the parapodial level and the other near the constriction between the somites, the latter being the smaller. This continues to about somite 20 and from here posteriorly the pigment gradually becomes less distinct, the patch near the inter- segmental constriction disappears, while remnants of the other persist as patches lying along the sides of the somites. They eventually disappear at about the region of the thirtieth somite. In the hving animal the red dorsal blood-vessel and the bright-red gills are noticeable features in the coloration. While these latter colors are lost in the preserved material, the course of the dorsal blood-vessel is marked by a row of spots due to coagulated blood. The pigmentation of the anterior region is retained, but the posterior end of the body is colorless in the preserved material. There are two pairs of anal cirri (fig. 23); one pair (the dorsalmost), very long and slender, the others very short. The prostomium (fig. 22) is rounded on its anterior margin with the frontal ten- tentacles rather widely separated, the latter elliptical in outline with a narrowed base. The cirrophores of the tentacles have not more than 3 basal rings, not very sharply marked off, and a terminal portion longer than the basal. These cirrophores extend nearly to the apices of the frontal tentacles. The outer paired tentacles have terminal portions about twice as long as the cirrophores, the other tentacles with the terminal portions as much as six times the length of the cirrophores. The terminal portions are slender, rounded at the apex, the unpaired a very little shorter than the inner paired. A very small and inconspicuous eye lies on either side postero-laterally from the base of the inner paired tentacle. The mouth is at the bottom of a broad pit, which is bounded dorsally by the two very prominent rounded palps, and ventrally by a 2-lobed prolongation of the first somite. In form and size these lobes are very similar to the palps, but unlike them, are not fused in the midline. The nuchal cirri arise on the anterior border of the (morphologically) second somite. They are conical in form and extend to as far as the terminal third of the cirrophores of the inner paired tentacles. Dorsally somites 1 and 2 are considerably shorter than the prostomium and are together about one-third as long as somite 3. Somite 3 is continued laterally into the first parapodium, of which only the dorsal cirrus is shown in the figure. This parapodium extends to a short distance in front of the anterior margin of somites 1 and 2. The parapodium of somite 4 also shows a forward extension, but this is not so apparent in later somites. Anteriorly the dorsal surface of the body is rounded and the parapodia have in consequence a ventral shifting, but behind somite 12 this rounding disappears and the parapodia assume a lateral position, while at the same time the dorsal median line shows a shallow groove. There is a large, dorsally directed^ anal opening. A ventral view of the pygidiura (fig. 23) shows the anal cirri, but the preservation was too poor to make it possible to draw the terminal somites and their relation to the pygidium. The first parapodium (fig. 24) has a very long cirrus-like postsetal lobe, with rounded presetal lobe. In the specimen figured the dorsal cirrus was bifid at the end, a quite exceptional condition, the usual cirrus being rounded at the apex, with a swollen base, and only a Httle longer than the postsetal lobe. The ventral cirrus is large, lanceolate, arising from a constricted base. There are 2 not very large aciculse, which taper slightly toward the apex until they reach the surface, when they suddenly narrow into very acute tips. A small tuft of needle aciculse extends into the base of the dorsal cirrus. There is a tuft of 5 or 6 very long "semi-compound" setae. The tenth parapodium (fig. 25) shows the ventral cirrus changed to the form of a vertically arranged, rather prominent pad, the postsetal lobe and the dorsal cirrus 178 Polychcetous Annelids Collected at Friday Harbor, retaining essentially their earlier character. A number of aciculse (4 in the one drawn) occupy the center of the setal lobe. One of these was drawn out into a slender, curved, acutely pointed tip, and it is probable that the others had originally this structure, which had been broken off. There is a tuft of needle setse in the dorsal cirrus. The setse form a dense tuft of four kinds: dorsalmost is a small tuft of very slender pectinate; next to these simple ones, long and very sharp-jx)inted, with the terminal portion very much flattened and provided with lateral wings; ventral to these, some of quite similar form, but with the terminal part broader and shorter than in the dorsal- most ones. Under high power they show a finely pilose character, as if fine hairs were thickly distributed over the terminal region. At the ventral end of the series is a bundle of compound setae. The thirty-fifth parapodium (fig. 26) has a conical setal portion with a rounded lobe on its anterior face. Dorsally it carries a gill which is larger than the dorsal cirrus and extends to beyond the dorsal midline. Long needle aciculae extend into the dorsal cirrus. The acicula* are of two kinds: (1) slender forms with the free end terminating in a very sharp point (fig. 27); (2) a much heavier form with apical and subapical teeth narrowing into a neck just proximal to the subapical tooth and then broadening into the shaft (fig. 28). Only one of each is drawn in figure 26, but there were two of one and four of 2 in the actual specimen. In place, the larger one looked as if hooded, but this was evidently due to the fact that it lay in a slight depression, so that it was partly covered by a transparent fold of the outer skin of the body. The semicompound setae (fig. 29) at first appear compound, but closer examination shows that what looks like a terminal joint is really not completely separated from the basal. There is a large terminal tooth and there are two unequal subterminal ones, with a hood covering them all, the apex of this hood thickened. The simple setae (fig. 30) enlarge rapidly toward the apex and then narrow to an acute point. The com- pound setae (fig. 31) are slender, with a smooth, pointed terminal joint without teeth or hood. The pectinate setae (fig. 32) have long shafts with about 12 very slender teeth. The gills begin on about the twentieth somite, the most anterior ones smaller than those farther back. There is never more than one branch (fig. 26). In the single entire specimen the last gill was on the twenty-fifth somite in front of the pygidium. The jaw apparatus is extremely soft and delicate and difficult to remove without injury. In the one figured the halves of the maxillae had fallen apart and the unpaired plate had turned over (fig. 33). Normally the inner margins of the carriers are in contact. Except for the dark pigmented patches indicated in the figure and the tips of the teeth, all parts are thin and transparent. The carriers are triangular in form, the forceps rather heavy and not much curved. Each proximal paired plate has 6 teeth, the distal paired has 10 on the right and 8 on the left. The unpaired has 7 well-developed teeth and 1 very small one. The mandible (fig. 34) has pointed lateral wings and faintly marked concentric lines on the beveled portion. Near the middle line each has a prominent dark band running lengthwise. Type in the American Museum of Natural History. Family SPIONID^. Polydora calilomica Treadwell. Polydora californica Treadwell, 1914, Polychsetous Annelids of the Pacific Coast in the Collections of the Zoological Museum of the University of California. Dniv. of Calif. Pub. in Zoology, vol. 13, p. 203, pi. 12, figs. 23 to 29. This species was abundant in Newhall's lagoon at Friday Harbor and was collected at False Bay. To the original description may now be added that of the pygidium, which was absent from the specimens from California. This has the form of a broad, very shallow funnel (fig. 35). State of Washington, in February and March 1920. 179 Family CIRRATULIDiE. Cirratulus robustus Johnson. Cirratulm robustus Johnson, 1901, Polychieta of the Puget Sound Region, Proc. Boston Soc. Nat. History, 29, p. 423, pi. 14, figs. 149, 150. Cirratulus cingulatus Johnson, loc. cit., p. 422, pi. 14, figs. 145 to 148. Living specimens are usually dark olive-green in color, but in some cases they may be almost black. The region in front of the tentacles is yellowish brown in most cases, but one specimen was nearly colorless. On the tentacles are bright scarlet spots looking like the eyes of a sabellid, arranged in an irregular row. The prostomium (fig. 36) has on either side a row of dark eye-like spots. The first somite has a curved anterior margin following the outline of the corresponding margin of the prostomium. Laterally and ventrally this somite is shorter than on the dorsal surface. A prolonga- tion of the second somite extends dorsally into the first. Later somites are much shorter than either of these. A comparison of Johnson's descriptions of his two species shows that in the general form of the body, the distribution of the cirri, the form of the setae, the general form of the head region, and the arrangement of the eye-spots the two are alike, the only differences being that in C. cingulatus the somites are ringed and the large cirri are on somite 8, while in robustus the somites are smooth and the large cirri are on somite 4. In this genus the first three somites are long and there is a tendency toward surface wrinkling. It seems to me that the difference Johnson described in tentacle position was an error, due to mistaking surface wrinkles for somite boundaries, and I ha\'e therefore combined the two species. Collected at Jurn Island and Minnesota Reef near Friday Harbor. Family OPHELIID.ffi, Ammotrypane brevis Moore. Ammotrypane brevis Moore, 1906. Proc. Acad. Nat. Sci. Phil. 58, p. 354, with 1 figure. Specimens of this genus collected at Friday Harbor, while one-third longer than the single specimen from Alaska on which Moore founded this species, agree with his description in every detail except two. The first is that the prostomium is not as definitely dorso-ventrally flattened as in Moore's description, a difference which might be due to the preservation, and the second is the structure of the pygidium. Moore notes: "If perfect, as it appears to be, the pygidium presents striking characters. The large, median spoon-shaped lobe of A. aulogasier is absent and represented only by a minute slender process. The lateral lobes are much larger, obliquely truncated above, and slightly indented at the end." In the Friday Harbor specimens the pygidium has a short hood, overhanging dorsally the anal opening, and continuing so far toward the ventral surface that the ventral margins are almost in contact. The margin of this lobe is drawn-out into cirrus-like processes. In all of the three individuals at my disposal the ventralmost of the processes are thick, elongate- elUptical in outline with constricted bases (fig. 37). Dorsal to these the margin of the hood bears a series of slender processes. In the three specimens there were, respect- ively, 4, 5, and 7 of these processes, varying in size, but always slender. A single slender cirrus arises from the ventral mid-line of the body, and may extend beyond the other processes as in the one figured, though in another it was shorter. In the third specimen it had evidently been lost. Collected at Friday Harbor. 180 PolychoBtous Annelids Collected at Friday Harbor, Figures 1 to 19. 1 and 2. Autolylus varius Treadwell. Fig. 1, anterior end, X 10; fig. 2, larva, X 52. :{ to 6. Eteone maculata new species. Fig. 3, anterior end, X 15; fig. 4, first parapodium, X 08; fig. 5, twentieth parapodium, X 45; fig. 6, seta, X 240. 7 to 10. Eteone tuberculata new species. Fig. 7, anterior end, X 10; fig. 8, first parapodium, X 68; fig. 9, later parapodium, X 68; fig. 10, compound seta, X 250. 11 to 13. Lumbrinereis zonata Johnson. Fig. 11, maxilla, X 41; fig. 12, mandible, X 41; fig. 13, anal cirri, X 18. 14 to 19. Lurnbriiiereis cenHcalis new species. Fig. 14, anterior dorsal view, X 10; fig. 15, anterior ventral view, X 10; fig. 16, first parapodium, X 47; fig. 17, simple seta. X 220; fig. 18, anterior compound seta, X 280; fig. 19, posterior simple seta, X 280. State of Washington, in February and March 1920. 181 Figures 20 to 37. Fig. 20. maxilla, X 28; fig. 21, 20 and 21. Lumbrinereis cervicalis new species (continued). 22 to 34. OnupSSrr^t^^new species, fig- 22. anterior dorsal view X 15; fig. ^ ventral view. X 15; fig. 24. first parapodium. X 68; fig. 25. ^^^^P^^.^P'l^;"^ X 68- fig. 26. parapodium from middle of body. X 45; fig. 27, acicuia. X 250 fig. 28, acicula, X 250; fig. 29. semi-compound seta. X 250; fig. 30. seta f^^^^ t«°t*if IJ" Dodium X 250; fig. 31, compound seta from tenth parapodmm, X ^50. Gg. 6^, pectinate seta. X 500; fig. 33. maxilla. X 45; fig. 34. mandible. X 45. 35. Polydora californica TieadweW. Pygidium, X 13. 36. Cirratulus robuslus Johnson. Anterior dorsal view, X 10. 37. Ammotrypane brevis Moore. Pygidium, X 20. TREADWELL PLATE 1 V- 4 "'' 11 - AHoertCo Baltimore Figures 1 to 7, Leodice vindis Gray. Fig. 1, anterior end X ' '■7%- F'^^''Z.!^JJ- F\J ^\ ^ % ^^^^isiili^-^' i 10 Figures 1 to 7, Borvillea australiensis Mcintosh. Fig. 1. anterior end X 6. Fig. 2, first para- podium X 41. Fig. 3, tenth parapodium X 41. Fig. 4, pygidiuni X 20. Fig. o, anterior plates of the maxillary series X 41. Fig. 6, median plates of the maxillary series X 41. Fig. 7, mandible X 41. Figures 8 and 9, Arabella duUa Treadwell. Fig. 8, maxilla X 68. Fig. 9, mandible X 68. Figure 10, maxilla of Drilonercis Inmhrictis Treadwell X 20. Figure U, maxilla of Drilonercis paucidentaia, Treadwell X oo. DEPARTMENT OF MARINE BIOLOGY OP THE CARNEGIE- INSTITUTION OF WASHINGTON ALFRED G. MAYOR, Directok PAPERS FROM THE DEPARTMENT OF MARINE BIOLOGY OF THE CARNEGIE INSTITUTION OF WASHINGTON VOLUME XVIII Published by the Caknegie Ii^stitution of Washtngton Washington, Novbmbkb, 1922 MBL WHUl LIBKAKY UH lani