fptl tegs *57X57 7 f//7 VOL. XXI JANUARY 1967 * NO. 1 PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION JOHN C. HOLDEN Late Cenozoic Ostracodes from Drowned Terraces in Hawaii JENS W. KNUDSEN Trapezia and Tetralia Crabs as Obligate Ectoparasites of Corals MARGARET D. KNIGHT Larval Development of Emerita rathbunae GEORG/ANDRA LITTLE Chromatophore Responses in the Hawaiian Ghost Crab ROBERT G. SCHWAB Responses of Polychoerus carmelensis to Temperature Changes RICHARD H. ROSENBLATT Osteology of the Congrid Eel Gorgasia punctata BENJAMIN C. STONE Flora of Romonum Island, Truk Lagoon B. C. ARNOLD Midge Gall of Myrsine australis H. L KRIVOY, C. G. JOHNSON, and R. Y. KOYANAGI Pseudoseisms Resulting from Military Exercises ALEXANDER MALAHOFF Gravity and Geological Studies of an Ultramafic Mass in New Zealand I I ' UNIVERSITY OF HAWAII PRESS o. Department John S. Gladys E. Baker Department of Botany University of Hawaii Sidney C. Hsiao Department of Zoology University of Hawaii George W. Chu Departments of Microbiology and Zoology University of Hawaii Charles H. Lamoureux Department of Botany University of Hawaii Maxwell S. Doty Department of Botany University of Hawaii John J. Naughton Department of Chemistry University of Hawaii Augustine S. Furumoto Hawaii Institute of Geophysics University of Hawaii Walter R. Steiger Department of Physics University of Hawaii Philip Helfrich Hawaii Institute of Marine Biology University of Hawaii Goro Uehara Department of Agronomy and Soil Science University of Hawaii Robert W. Sparks, Director University of Hawaii Press *.A ) ARD OF EDITORS A. Bushnell, Editor-in-Chief of Microbiology, University of Hawaii Howe, Jr., Assistant to the Editors University of Hawaii Press INFORMATION FOR AUTHORS Contributions to Pacific biological and physical science will be welcomed from authors in all parts of the world. (The fields of anthropology, agriculture, engineering, and medicine are not included.) Manu- scripts may be addressed to the Editor-in-Chief, PACIFIC SCIENCE, University of Hawaii, Honolulu, Hawaii 96822, or to individual members of the Board of Editors. Use of air mail is recommended for all communications. Manuscripts will be acknowledged when received and will be read promptly by members of the Board of Editors or other competent critics. Authors will be notified as soon as possible of the decision reached. Manuscripts of any length may be submitted, but it is suggested that authors inquire concerning possi- bilities of publication of papers of over 30 printed pages before sending their manuscripts. Authors should not overlook the need for good brief papers, presenting results of studies, notes and queries, com- munications to the editor, or other commentary. preparation of manuscript It is requested that authors following the style of Pacific Science described herein and exemplified in the journal. Authors should attempt to conform with the Style Manual for Biological Journals, Am. Inst. Biol. Sci. Washington. (Continued on inside back cover) PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION VOL. XXI JANUARY 1967 NO. 1 Previous issue published December 31, 1966 CONTENTS PAGE Late Cenozoic Ostracodes from the Drowned Terraces in the Hawaiian Islands. John C. Holden 1 Trapezia and Tetralia ( Decapo da , Brachyura, Xanthidae ) as Obligate Ectopara- sites of Pocilloporid and Acroporid Corals. Jens W. Knudsen 51 The Larval Development of the Sand Crab Emerita rathbunae Schmitt ( Dec a - poda, Hippidae). Margaret D. Knight 58 Chromatophore Responses in Relation to the Photoperiod and Background Color in the Hawaiian Ghost Crab, Ocypode ceratophthalma {Pallas). Georgiandra Little 77 Overt Responses of Polychoerus carmelensis (T urbellaria: Acoela) to Abrupt Changes in Ambient Water Temperature. Robert G. Schwab 85 The Osteology of the Con grid Eel Gorgasia punctata and the Relationships of the Heterocongrinae. Richard H. Rosenblatt 91 The Flora of Romonum Island, Truk Lagoon, Caroline Islands. Benjamin C. Stone 98 A Hitherto Unrecorded Midge Gall of Myrsine australis (A. Rich.) Allan. B. C. Arnold 115 An Unusual Example of Pseudoseisms Resulting from Military Exercises. Har- old L. Krivoy, Charles G. Johnson, and Robert Y. Koyanagi 119 Gravity and Geological Studies of an Ultramafc Mass in New Zealand. Alex- ander Malabo ff 129 NOTE A Noninjurious Attack by a Small Shark. David P. Fellows and A. Earl Murchison 150 Pacific Science is published quarterly by the University of Hawaii Press, in January, April, July, and October. Subscription prices: institutional, $10.00 a year, single copy, $3.00; individual, $5 00 a year, single copy, $1.25. Check or money order payable to University of Hawaii should be sent to University of Hawaii Press, 535 Ward Avenue, Honolulu, Hawaii 96814, U. S. A. Printed by Heffernan Press Inc., 35 New Street, Worcester, Massachusetts 01605. ^ ■ps«' - i i ■ Late Cenozoic Ostracodes from the Drowned Terraces in the Hawaiian Islands John C. Holden1 ABSTRACT: Late Cenozoic ostracodes from extensive submarine terraces in the Hawaiian Islands ranging in depth from 260 to 355 fathoms resemble, in part, modern shallow water faunas of the Hawaiian and tropical Pacific islands. Of the 35 species from the terraces, 13 are described as new. These are: Cytherelloidea monodenticulata, Bairdia kauaiensis, B. hanaumaensis, B. ritugerda, Hemicythere obesa, Mutilus oahuensis, M.(P) coales cens, Jugosocythereis venulosus, Quadracy- there hornibrooki, Loxoconcha batei, L. condyla, Cletocythereis bradyi , and Neocaudites terryi. The assemblage indicates an original shallow water environment for the terraces. Most of the extant species, which also occur as fossils from the terraces, live at depths less than 50 fathoms in present oceans, and only one is reliably reported as living deeper than 160 fathoms; several are known littoral forms. Submarine terraces occur at various depths on the flanks of the larger islands in the Hawai- ian archipelago. The central islands of Oahu, Molokai, Lanai, Maui, and Kahoolawe sur- mount a common ridge which rises abruptly from the Pacific Ocean floor from 2,500 fathoms to about 300 fathoms. The conspicuous break in slope at 300 fathoms marks the outer edges of extensively developed drowned terraces (Fig. A). Submarine terraces on the north side of Kauai at about the same depth are not as ex- tensively developed as those off the central is- lands. Deeper terraces are known down to 1,000 fathoms in the Hawaiian Islands (Menard et al., 1962) but have not yet been dredged. This study is primarily concerned with the ostracode remains contained in several dredge hauls taken from the submarine terraces within the central island complex and off Kauai. While this is primarily a taxonomic study, Tables 1 and 2 are given to evaluate paleoenvironments and thus unravel the obscure history of the Hawaiian Islands. Unfortunately, such evalua- tion is hindered by the lack of knowledge of the precise age of the Hawaiian fossil Ostracoda 1 U. S. Navy Electronics Laboratory, San Diego 52, California. Present address: Department of Paleon- tology, University of California, Berkeley, California. Manuscript received December 2, 1965. and of the ecology of living Pacific ostracodes included in this report. Cursory observations of Recent samples from Clipperton Island and New Caledonia reveal no gross faunal similarities to the Hawaiian fossil faunas. From the limited number of Re- cent samples from the Hawaiian Islands it can be seen that only a few of the fossil species are presently living in that area (Table 1). Some specimens from the Recent stations are illustrated for clarity. Marked faunistic differences occur between terraces, indicating either temporal or environ- mental distictions. For example, station T-12 at 308 fathoms does not contain Macrocypris gracilis or Loxoconchella honoluliensis, which are common at stations T-l, T-4, and T-7 at 310, 280, and 297 fathoms respectively, nor does it contain the common Mutilzis (?) coalescens and the abundant Mzztilus oahuensis, found at station AR at 260 fathoms. BRIEF HISTORIC REVIEW OF THE OSTRACODA in the tropical pacific: The first published work on tropical Pacific ostracodes was that of G. S. Brady (1868^) in "Fonds de la mer.” However, only 1 species was treated in the central Pacific. Ostracodes from Java and Hong Kong were also covered in this series. Of greater 1 2 PACIFIC SCIENCE, Vol. XXI, January 1967 Fig. A. A bathymetric chart of the areas immediately adjacent to the Hawaiian Islands, showing a major break in slope between the 1,000- and 3,000-ft isobath. Stations HA, in Hanauma Bay, and T-13, on Penguin Bank, contain Recent ostracodes exclusively; the remaining stations contain only fossils. import concerning Pacific faunas was the de- scription of the Ostracoda collected during the Challenger Expedition, also by G. S. Brady, published in 1880. This was primarily a taxo- nomic treatment of the ostracodes obtained from dredgings in many parts of the world by the Challenger Expedition during the years 1873-1876. Though an excellent work in certain respects, and well illustrated, it was little better than a reconnaissance. Brady again published in 1890 on the Os- tracoda from some South Sea islands (New Caledonia, Samoa, and Fiji). His entire collec- tions consisted of species found from the littoral zone to 6 fathoms. Though certainly not a com- prehensive study of South Pacific ostracodes, this, together with his 1880 work, constitutes the bulk of descriptive information available on this group in the tropical Pacific. J. Th. Kingma (1948) described 94 species, 40 of which were new, from Tertiary to Recent deposits of the Netherlands East Indies. He reported relatively few tropical Pacific island ostracodes living in the Java Sea. Shorter works in the Indo-Pacific realm include studies of late Cenozoic fossils by Le Roy (1939, 1941) and Doeglas (1931), and of Recent ostracode studies by Chapman (1902, 1910), Triebel (1954), Keij (1953, 1954), 1964), Fyan (1916), and Bold (19 46b, 1950). Major contributions have been made in the New Zealand and the South Australian regions; however, these faunas are distinct from those of the tropical Pacific region. For an historical discussion of that area the reader is referred to Hornibrook (1952). The history of ostracode study in the north Pacific-Japan area is treated thoroughly by Hanai (1959). age and paleoecology: Menard, Allison, and Durham (1962) interpreted the age of a single station (AR of this study) to be prob- able Late Miocene on the basis of the ratio of extinct coral species to living species and on vagaries in shape of specimens of Globtgeri- noides quadrilobatus. The hermatypic corals in- dicated an initial depth of only 10 m. Allison (personal communication) has since expressed Late Cenozoic Ostracodes — Holden 3 TABLE 1 Species Check List of Late Cenozoic Hawaiian Ostracodes from the Hawaiian Islands, Clipperton Island, and Easter Island Relative abundances are made with reference to the total population: 0-5% rare (R) ; 5-15% common (C) ; 15-40% abundant (A); 40-100% very abundant (VA). SPECIES EASTER ISL. HAWAIIAN ISLANDS EASTER ISL. CLIPPERTON ISL. INC O GENE RECENT EA-2 T-l T-2 T-4 T— 7 00 iL T-ll T-l 2 AR S-23 T-l 3 HA EA-I CL Cythereiioidea monodenticuiata n. sp. R Bairdia kauaiensis n. sp. A VA C A A VA A A A A L C Bairdia sp. R Bairdia hanaumaensis n. sp. R C R Bairdia ritugerda n. sp. C VA VA C C R Bairdia attenuate Brady, I860 R R C R C C R Bairdia expansa Brady, 1880 R 1 Macrocypris gracilis (Brady), 1890 R C A A R Propontocypris simplex (Brady), 1880 R R i d $ Mi 1 CJ as » 3 § "S O Pi 2 s : o m-h g G <5 £ h .5; co £uo 55 o H < csS P P > Ar < w 9 ^ N w oq ^ 'P CM SO f'l d SO 1 4 so I I | ITS OS OS 00 1^ OS qo os oo os r- GS 00 I so I I & 3> a G qj G G ^ § tjO ON CM CX XT 00 XT no © xr © cNN ‘'N CNN xr xr XT CX d 6 o d o o d © d d d 00 00 CX 00 CX ^N 00 >TN © XT xr © © 00 00 f-i rH cnn XT NO d d © d d d © d rH 7-1 1 — i f s ^ ^ x __ ^ V 00 00 xr © xr © CX © o © xr xr m xr VTN © q © © ^N 1TN ICN 1 00 © NO 1 xr 1 xr NO 1 xr 1 xr 1 O 1 © 1 © rH © r-N © © t-H 00 q CO' xr xr © T-l r-i cx d m d xr xr xr V^N W 'W w w w W © © © TS 00 fNN \T\ 00 T— 1 © F- xr ex 04 rH q (X © xr © q © © © r-1 r-1 ex d cC> CNN xr xr l/N i/N V xr © 00 00 xr VO © 00 © VO vo 0 xr '/N © F- © ex xr © rH q © © 1 © © I T *7 H | 1 Cd 1 cx 1 1 cx 04 00 xr 1 00 1 r\ F-- 00 ITN © © © q cO © q q q "9 © © © © © r-| r-J r-l H d «N cx £ > > > TERMINAL ZOEA 345 ( 3.20-3.80 ) 2.54 ( 2.36-2.80 ) 4.88 ( 4.68-5.39 ) 2.16 ( 1.96-2.88 ) 2.81 ( 2.64-3.00 ) 2.04 ( 1.96-2.24 ) 0.71 Larval Development of Emerita rathbunae — Knight 63 Figs. 1-5. Emerita rathbunae. Zoea I-V. 64 PACIFIC SCIENCE, Vol. XXI, January 1967 broad rostral spine; the lateral spines character- istic of later stages are absent. The large eyes are stalked. The first antenna (antennule) (Fig. 11) is conical and unsegmented, tapering distally to a blunt tip which bears 3 aesthetes and 2 hair- like setae. The protopodite of the second antenna (Fig. 17) is produced into a strong lateral spine and bears a slender inner spine of approximately the same length. A small spine is situated ventrally at the base of the medial spine. The mandibles (Fig. 23) are armed with strong ventral teeth, short triangular teeth, and slender spines. The ventral tooth of the left mandible is split shallowly at the tip. There is little change except growth during zoeal development. The coxal endite of the first maxilla (maxil- lule) (Fig. 24) bears 3 terminal setae with 1 small seta subterminally on the inner margin. The basal endite bears 2 strong curved spines armed with tiny spines, the small unsegmented endopodite bears a single long seta. The protopodite of the second maxilla (Fig. 26) is triangular, bearing 3 setae on the blunt anterior margin, 1 set slightly apart toward the scaphognathite, and 1 small seta subterminally on the inner margin. In later stages, the anterior tip becomes more pointed and the 3 setae more evenly spaced (Fig. 27). The scaphognathite bears 7-8 plumose setae on the anterior-outer margin; one-third of the specimens dissected, from both hatching and plankton, had 7 setae. The short coxopodite of the first maxilliped (Fig. 29) is unarmed. The basipodite bears 7 setae along the medial margin in groups of 1—1— 2— 3 progressing distally. The endopodite is 4-segmented; the first three segments are armed along the inner margin as follows: first segment with a group of 3 setae, one conspicu- ously stronger than others and armed with tiny spines; second segment with 2 setae, again one being stout and armed with spinules; third seg- ment with 2 setae spaced around the distal mar- gin of the segment. The fourth segment bears 4 terminal setae; the outer 2 setae are quite long and armed with spinules on the inner margin. There is also a short hairlike seta placed subterminally on the outer margin which fre- quently curves in between the terminal setae and is difficult to see without high magnifica- tion. The exopodite consists of 2 segments; the very short, often weakly delineated, terminal segment bears 4 long plumose natatory setae. The coxopodite of the second maxilliped (Fig. 28) is unarmed, the basipodite bears 3 setae on the inner margin in groups of 1-2 progressing distally. The endopodite consists of 4 segments. Along the medial margin, the first segment bears 3 setae distally, the second seg- ment bears 1 seta, and the third segment has 2 setae around the distal margin of the segment. The fourth segment bears 4 terminal setae and 1 small subterminal seta on the outer margin as described for the first maxilliped. The exopodite is 2-segmented, the small terminal segment bears 4 plumose natatory setae. The abdomen consists of 5 segments. The first is very weakly differentiated. The sixth segment is consolidated with the telson, as shown by the position of the uropods in subsequent stages. The telson (Fig. 34) is rounded, slightly concave, and usually about as wide as long, occa- sionally slightly longer than wide. There are 2 prominent posterior-lateral spines notched near the tip on the outer margin. Between the lateral spines there are 25—27, usually 26, spines around the posterior margin of the telson, with a series of very small denticles between the spines. The eighth spine from either side is somewhat longer and more prominent than the remaining spines; all are armed near the base with small spinules. There is little change throughout zoeal development except for an increase in number of denticles between the terminal spines. In the following stages, unless noted, there is no change in setation and form of appendages described and figured for zoea I. zoea ii (Fig. 2): There is now a pair of short lateral spines on the carapace. The first antenna (Fig. 12) terminates with 1 large aesthete and approximately 3 small hairlike setae. The second antenna (Fig. 18) bears a small subterminal spine on the medial margin of the lateral spine. The basal endite of the first maxilla (Fig. 25), with the addition of 1 spine, now bears I 3 strong curved spines armed with spinules; ; Larval Development of Emerita rathbunae — Knight 65 Figs. 6-10. Emerita rathbunae. 6, Zoea VI; 7, zoea IV, dorsal; 9, zoea IV, telson. Emerita analoga. 8, Zoea IV, dorsal; 10 , zoea IV, telson. 66 the inner two spines are articulated at the base. There is no other change and the first maxilla maintains this form throughout zoeal develop- ment. The scaphognathite of the second maxilla bears usually 8, occasionally 9, plumose setae on the anterior-outer margin. The exopodites of the first and second maxil- lipeds bear 6 plumose natatory setae. zoea m (Fig. 3): The first antenna (Fig. 13) bears 3 aesthetes and 3 small setae on the tip. One aesthete is slightly larger and set apart from the other two in the terminal grouping found throughout further zoeal development. Small spines have been added distally on both lateral and medial spines of the second antenna (Fig. 19). The scaphognathite of the second maxilla may have 9 or 10, rarely 11, plumose setae on the anterior-outer margin. The exopodites of the first and second maxil- lipeds bear 8 plumose natatory setae. A pair of uniramous, 2-segmented uropods (Fig. 31) are now present on the anterior- ventral portion of the telson. Each of the distal segments bears 2 slender curving terminal setae armed with tiny spines distally ; the inner seta is longest. zoea iv (Figs. 4, 7) : The first antenna (Fig. 14) now bears a subterminal tier of 2 aesthetes on the medial margin. The flagellum of the second antenna (Fig. 20) appears in this stage as a slight rounded prominence to a small bud. The number of plumose setae along the outer margin of the scaphognathite of the second maxilla ranged from 15 to 21; most specimens had 17-20. The exopodites of the first and second maxil- lipeds bear 10 natatory setae, and the basipodite of the first maxilliped may have 8 setae along the inner margin in groups of 1-2-2-3 pro- gressing distally. Small buds of the third maxilliped and tho- racic appendages are present beneath the cara- pace, posterior to the second maxilliped. The exopodites of the uropods (Fig. 32) now bear 4 terminal setae of varying lengths; the third seta is the longest. The endopodite PACIFIC SCIENCE, Vol. XXI, January 1967 may appear in this stage as a rudiment or a I; small bud. The telson (Fig. 9) has become somewhat j longer than wide and remains so in subsequent zoeal stages. zoea v (Fig. 5): The first antenna (Fig. 15) bears usually 2, occasionally 3, subterminal groups of aesthetes along the medial margin; only one-fourth of the specimens dissected had j 3 tiers of aesthetes. The majority of larvae had j 4 subterminal aesthetes in groups of 2-2 ; rarely ; an additional aesthete was added to form groups j of 2-3 progressing distally. Those larvae with i 3 subterminal tiers of aesthetes added them in groups of 2-2-3, rarely 1-2-3. The flagellum of the second antenna (Fig. | 21) is now slightly shorter than to slightly longer than the 2 spines of the protopodite. The scaphognathite of the second maxilla i bears 24-39 plumose setae along the outer margin; most individuals had between 30 and j 34 setae. The exopodites of the first and second maxil- | lipeds bear 12, rarely 11, natatory setae. The basipodite of the first maxilliped now bears 8 j setae along the medial margin in groups of 1 -2-2-3 progressing distally. The third maxilliped and thoracic appendages have increased in size, curving under toward the thorax. The exopodites of the uropods (Fig. 33) now bear 5 or 6 setae of unequal length; the third seta is very long. More larvae had 5 than 6 setae, a few had 5 on one side and 6 on the j other. The endopodite now varies in size from ; a small to prominent bud approximately 1/3 the length of the exopodite. zoea vi (Fig. 6) : Now the first antenna j (Fig. 16) usually has 4 subterminal tiers of ; aesthetes in groups of 2-3-4-4 or 2-2-4-4 progressing distally. Of 19 specimens 5 had only 3 tiers in groups of 2-3-4 or 2-4-4. The flagellum of the second antenna (Fig. 22) has increased greatly in length and now dwarfs the spines on the protopodite. In speci- mens close to molting to megalopa, the seg- mentation of postlarval peduncle and flagellum can be seen beneath the cuticle. The scaphognathite of the second maxilla Larval Development of Emerita rathhunae — Knight 67 27 Figs. 11-27. Emerita rathbunae. 11-16 , First antenna, zoea I-VI; 17-22, second antenna, zoea I— VI; 23, mandible, zoea I; 24-25, first maxilla, zoea I and II; 26-27, second maxilla, zoea I and VI. 68 (Fig. 27) has 43-55 plumose setae along the outer margin. The exopodites of the first and second maxil- lipeds now bear 14 plumose natatory setae. The third maxilliped and thoracic legs (Fig. 30) have increased greatly in size; the fifth leg, curved up and behind the first four, is slightly bifid at the tip. Now each of the segments 2, 3, 4, and 5 of the abdomen bears a pair of uniramous, un- segmented pleopods. The exopodites of the uropods (Fig. 35) bear 7 or 8 setae of varying lengths; twice as many larvae had 7 as had 8 setae; a few had both 7 and 8 setae. The endopodites are quite long, usually 3/4 the length of the exopodites. alternate zoeal stages: Of the larvae of Emerita rathbunae studied from preserved sam- ples, 49 had at least 14 setae on the exopodites of the first and second maxillipeds. While 33 of these larvae were in the described stage VI, with pleopods on abdominal somites, and were, in many cases, close to molting to megalopa, sixteen of the larvae seemed to have prolonged the larval cycle to seven zoeal stages. Five larvae with 14 natatory setae on the maxillipeds did not have pleopods on abdominal segments. They had only 6 setae on the exopodites of the uropods, 3 tiers of aesthetes on the first antenna, and in all other respects (measurements, de- velopment of appendages, etc.) were inter- PACIFIC SCIENCE, Vol. XXI, January 1967 mediate between the forms described as zoea V and zoea VI. The remaining 11 larvae had 16 setae on the exopodites of the maxillipeds (one had only 15), had pleopods on abdominal somites, and were slightly larger and more advanced than the form described as zoea VI. Two zoea IV and two zoea V were found which corresponded with the described stages in over-all proportions and in setation. They were slightly smaller, however, and some ap- pendages were somewhat less developed (flagel- lum of second antenna and thoracic legs), which suggests that they might be the early stages of such an extended larval cycle. The variation in development and setation of some appendages of cultured larvae within stage VI is summarized in Table 3. megalopa (Fig. 36) : The megalopa is color- less, slightly translucid and very much like the adult in form. The most noticeable differences are presence of setose pleopods on abdominal segments and relatively large eyes. The aver- age size of carapace in reared individuals was: length, 2.65 mm; width, 1.98 mm. No speci- mens from the plankton were available. The first antenna (Fig. 37) consists of a 3- segmented peduncle and a flagellum. The sec- ond and third segments of the peduncle have small ventral processes armed with setae and that of the third segment is 2-segmented. The flagellum usually consists of 10 segments armed TABLE 3 Comparison of Some Features of Cultured Larvae in Stage VI from Series with 7, 8, and 9 Zoeal Stages HATCHED LARVAE PLANKTONIC LARVA FEATURE 7 stages 8 STAGES 9 STAGES 8 STAGES First antenna: range 2 1-2 1-2 No. of subterminal tiers of aesthetes majority 2 2 1 and 2 3 Second antenna: rudiment to Development of flagellum range = spines — spines 0 to rudiment in relation to spines on protopodite majority = spines r= j spines 0 = spines First and second range 13-14 11-14 12-14 maxillipeds: majority 14 13 and 14 12 and 13 14 Natatory setae Uropods: Exopod setae range 6-7 5-7 5-7 majority 6 6 6 6 and 7 Larval Development of Emerita rathbunae — Knight 69 Figs. 28-35. Emerita rathbunae. 28, Second maxilliped, zoea I; 29, first maxilliped, zoea I; 30a, third maxilliped, b-f, thoracic legs 1-5, zoea VI ; 31-33, uropod, zoea III-V ; 34-33, telson, zoea I and VI. 70 PACIFIC SCIENCE, Vol. XXI, January 1967 Figs. 36—4 1 . Emerita rathbunae. Megalopa. 36, Dorsal; 37, first antenna; 38, second antenna; 39, man- dible; 40, first maxilla; 41, second maxilla. Larval Development of Emerita rathbunae — Knight 71 laterally and ventrally with strong setae. The distal 5 segments bear aesthetes between the ventral setae. The basal segments of the second antenna (Fig. 38) are similar to those of the adult. The flagellum consists of 23-25 segments each bearing 7 processes: 2 long plumose filtering setae, 2 strong setae armed with comblike spines, and 3 shorter unarmed setae. The mandible (Fig. 39) consists of a light gnathal lobe and a palp of 2 segments. The first segment of the palp has 3, rarely 2, stout setae on the lateral margin, and the terminal segment bears setae along the medial and ante- rior margins. The basal endite of the first maxilla (Fig. 40) is armed with short, stout teeth and numer- ous setae; 1 long plumose seta is conspicuous on the anterior-outer corner. On the coxal en- dite, a series of long setae curve sharply down toward the mouth region. The endopod is un- segmented and saclike. There is 1 long seta on the lateral angle of the protopodite below the endopodite. The scaphognathite of the second maxilla (Fig. 41) has a dense fringe of approximately 95 setae along the outer margin. The coxal endite is now bilobed; the small distal lobe bears 1 long seta. The proximal lobe and the basal endite bear many setae. The small tri- angular endoped is unarmed. The anterior portion of the protopodite of the first maxilliped (Fig. 42) is produced into a flat blade armed with rows of small setae and a series of long plumose setae on the basal por- tion. The exopodite consists of 2 segments; the bladelike terminal segment is fringed with plumose setae. The rudimentary endopod is unarmed. The exopodite of the second maxilliped (Fig. 43) is 2-segmented; the first segment bears 3-4 strong setae on the lateral margin and the small oval terminal segment is fringed with plumose setae. The endopodite consists of 4 segments with setation as figured. The meropodites of the third maxilliped (Fig. 44) are broad and opercular. The 3 slen- der terminal segments bear plumose and bristle setae; the inner surfaces are covered with dense rows of setae to form a brushlike structure. The pereiopods are like those of the adult in form, with the first three pairs directed for- ward and the fourth pair directed posteriorly. The fifth legs, slender and chelate, are curved up beneath the carapace. The abdomen now consists of 6 segments; segments 2, 3, 4, and 5 bear biramous pleopods and segment 6 carries biramous uropods. The pleopods (Fig. 45) decrease in length posteri- orly. The exopodites bear plumose setae; the first pair has 14-15 setae, the second pair bears 15-16 setae, and the third and fourth pairs have 17-18 setae. The knoblike endopodites increase in length from the first to the fourth pair and have tiny median hooks which inter- lock with those of the opposite pleopod to form a single swimming unit of the pair. The oval exopodites and endopodites of the uropods (Fig. 46) are fringed with plumose and small unarmed setae. The triangular telson (Fig. 46) has plumose setae along the lateral margins. Comparison of Species Among larvae of Emerita analoga studied from preserved plankton samples, the early stages, I-IV (the "low stage IV” described by Johnson and Lewis (1942:79) with usually 10, occasionally 9 or 11, setae on the first and second maxillipeds) , were found to be consis- tent in detail and degree of development, but later stages showed such variation in setation and development of appendages that the num- ber of molts through which any individual had progressed could not be ascertained with con- fidence. The terminal zoeas, with indications of postlarval appendages beneath the cuticle, were consistent in possession of pleopods, an extremely long flagellum on the second antenna, and 5 tiers of subterminal aesthetes on the first antenna, as described by Johnson and Lewis. From 16-19 setae were found on the exopodites of the maxillipeds. Groupings of the late stage larvae by size and relative growth of appen- dages suggested that there were at least 7 zoeal instars in the planktonic larval life of the species. No differences in morphological detail were detected between larvae of Emerita rathbunae and E. analoga in stages I and II. In stage III, the uropods of E. analoga may bear 3 setae but 7 out of 10 specimens examined had only 2 72 PACIFIC SCIENCE, Vol. XXI, January 1967 Figs. 42-46. Emerita rathbunae. Megalopa. 42, First maxilliped; 43, second maxilliped; 44, third maxil- liped; 43, pleopods (a, fourth, b, first); 46, uropods and telson. setae as found in the equivalent stage of E. rathbunae. In stage IV, the setation of the ex- opodites of the uropods becomes consistent and a useful character for differentiation of the species; larvae of E. analoga have 5 setae and those of E. rathbunae have 4 setae on this ap- pendage. In all subsequent instars, the setation of the coxal endite of the first maxilla may be used. The larvae of E. analoga develop a strong fifth seta subterminally on the proximal margin, while larvae of E. rathbunae apparently main- tain 4 setae throughout zoeal development. Differences in terms of size of carapace, ratio of length of carapace to rostral spine, and pro- portions of the telson increase between larvae of E. analoga and E. rathbunae as zoeal devel- Larval Development of Emerita rathbunae — Knight 73 opment proceeds. The larvae of E. anologa be- come progressively larger than those of E. rathbunae, and the carapace spines are consider- ably shorter in relation to the length of the carapace. The posterior margin of the telson of E. analoga larvae becomes increasingly pointed and triangular between the prominent eighth marginal spines, while that of E. rathbunae re- mains smoothly rounded. Measurements of the first four stages and of the terminal stage of larvae of E. analoga are given in Table 2 for comparison with equivalent stages of E. rath- bunae. The carapace and telson of both species in stage IV are shown in Figures 7—10. The larvae of Emerita emerita , described by Menon (1933), and E. talpoida, described by Rees (1959) and Smith (1877), and those of E. rathbunae and E. analoga appear to be very similar in structure of appendages and in form of carapace and telson. The pattern of develop- ment may be common for all species through stage III, and through stage IV (with specific variation in setation) for those species described from the plankton. Larvae of E. emerita, and E. talpoida apparently pass directly from zoea IV to terminal zoea in the plankton. Those of E. talpoida in the laboratory and of E. rathbunae and E. analoga in both laboratory and field have a variable series of intermediate instars be- tween stage IV and the terminal zoea in which there is progressive growth and setation with- out addition of appendages. The zoeal stages which appear to be common to all larvae of the genus are as follows: 1. Uropods absent a. Lateral spines on carapace absent, 4 natatory setae on maxillipeds Stage I b. Lateral spines on carapace present, 6 natatory setae on maxillipeds Stage II 2. Uropods present a. Pleopods absent, 8 natatory setae on maxillipeds Stage III b. Pleopods absent, 10 natatory setae on maxillipeds Stage IV c. Pleopods present, 12 or more natatory setae on maxillipeds Terminal Stage Distribution of Larvae The locations of zooplankton samples exam- ined and the distribution of Emerita larvae are given in Figure 47. Larvae of E. analoga were found in samples taken near Magdalena Bay, and those of E. rathbunae usually in samples taken south and east of Cape San Lucas. A group of stage I larvae were found in near- shore samples from the west coast of Baja California below Magdalena Bay. From mea- surements, they appeared to be larvae of E. rathbunae, but lack of morphological features with which to differentiate stage I larvae of the two species makes identification tentative. In addition, 16 larvae of E. rathbunae, ranging from zoea II to terminal zoea VI, were found just south of Magdalena Bay. It seems likely, inasmuch as the range of developmental stages was found in the sample, that the larvae were hatched locally and that populations of the spe- cies might be found in the sandy beaches be- tween Cape San Lucas and Magdalena Bay. Twenty larvae were found in samples taken in November and December in the Gulf of California, the majority north of La Paz, which, although slightly smaller, were almost identical with those of the coastal E. analoga in morpho- logical detail and proportions. Only late stages were obtained. These showed variation in seta- tion and development of appendages similar to that found in larvae of E. analoga. During Au- gust to December, the movements of surface water along the western coast of Baja California are predominantly offshore and westerly (Wyrtki, 1965). It therefore seems unlikely that the larvae could have been carried into the Gulf from breeding populations near Magda- lena Bay. The appearance of the analoga- like larvae in a series of samples suggests that either Emerita analoga or a closely related species may be found in the warm temperate zone of the Gulf of California, extending north from Aqua Verde Bay on the west coast and Puerto San Carlos on the east (Garth, 1955, I960). DISCUSSION This study of planktonic and cultured larvae of Emerita rathbunae has shown that the num- ber of zoeal stages in the larval period is vari- 74 PACIFIC SCIENCE, Vol. XXI, January 1967 Fig. 47. Location of zooplankton samples examined and distribution of Emerita larvae. Larval Development of Emerita rathbunae — Knight 75 able, both in nature and under conditions of laboratory culture. In the laboratory, although the larvae retained their specific proportions and pattern of development, as many as three intermediate instars could be added to the larval sequences usually observed in specimens from the plankton. Variability between individuals at comparable stages of development increased with an increase in the number of zoeal molts in the larval period. A variable number of zoeal stages has been found in the laboratory culture of two other species of Emerita. Dr. Ian Efford (personal communication) has noted 9-11 zoeal molts among reared larvae of E. analoga, and larvae of E. talpoida cultured by Rees (1959) passed through 6 or 7 zoeal stages before the molt to megalopa. Rees, using setation of the maxilli- peds as an indication of the stage of develop- ment, compared the zoeal stages observed in the laboratory with those described by Smith (1877) from the plankton and found that larvae from the plankton which molted to meg- alopa in the laboratory were apparently only in stage V. Rees noted as well that "zoea from nature possess features (appearance of thoracic limb buds, pleopods, etc.) which show them to be farther advanced in development than the corresponding laboratory stages.” This relation between cultured and ''natural” larvae was also observed in the present study. Some indication of seasonal variation in num- ber of zoeal stages was found in specimens of E. rathbunae examined from the plankton. Most (94%) of the late-stage larvae taken in August and September appeared ready to molt to mega- lopa after six zoeal stages; in the December samples only 32% would metamorphose after six stages and 68% after seven stages. Differences in setation and development of appendages between individuals in comparable intermediate stages were not noted by Rees in cultured larvae of E. talpoida and were rare in larvae of E. rathbunae from the plankton. In- dividual variation became more pronounced among reared larvae of E. rathbunae with an increase in the number of zoeal stages, and might be related to rate of development. Larvae of E. rathbunae subjected at an early age to low temperatures (and perhaps other variables) of the laboratory environment had a larval span of 81-94 days. E. talpoida completed zoeal develop- ment in 23-33 days at 30 °C. The molting fre- quency of the planktonic larva of E. rathbunae cultured through two instars at 27°-30°C was consistent with that found by Rees for E. tal- poida, and it appears likely that E. rathbunae would have a much shorter larval life at the higher temperatures in its natural environment. Perhaps culture of Emerita larvae over a range of controlled temperatures would show a rela- tion between duration and number of zoeal stages and degree of individual variability. The consistent difference in size found between cul- tured and planktonic specimens suggests that larvae of this species are restricted in over-all size by the conditions of laboratory culture. Costlow (1965) has reviewed accounts in the literature of variability within larvae of Crustacea and has discussed effects of environ- mental factors (light, diet, temperature, salinity, etc.) on frequency and variability in molting, as well as current investigations of endocrine mechanisms related to molting in larvae of brachyuran decapods. Variation in number of zoeal intermolts has been noted in the laboratory culture of several anomuran decapods by Pro- venzano (1962a, b') for two species of pagurid crabs, and by Boyd and Johnson (1963) for the galatheid, Pleuroncodes planipes, but ap- parently such variation is rare among brach- yurans. Gurney (1942) suggested that artificial rearing might give misleading results, and stated that, while stages I— III in the develop- ment of larval decapods seemed to be relatively fixed, the natural course of development after that might be disturbed with addition of stages not found in nature. He noted as well that there is no certainty that all stages observed in nature are passed through by all individuals of a species. The use of only laboratory-reared material to investigate the growth patterns of a species with such capacity for variability in larval development as that shown by E. rath- bunae would indeed have been misleading unless supplemented by a study of the larvae taken from their natural environment. ACKNOWLEDGMENTS This work was supported by the Marine Life Research Program, the Scripps Institution of 76 Oceanography’s component of the California Cooperative Oceanic Fisheries Investigations, a project sponsored by the Marine Research Com- mittee of the State of California. I would like to thank Dr. M. W. Johnson, Dr. E. W. Fager, Dr. J. A. McGowan, and Dr. W. A. Newman for their valuable criticism of the manuscript. REFERENCES Boyd, Carl M., and M. W. Johnson. 1963. Variations in the larval stages of a decapod crustacean, Pleuroncodes planipes Stimpson (Galatheidae) . Biol. Bull. 124(3) :l4l— 152. Butler, T. H. 1959. A record of the anomuran crustacean Emerita analoga (Stimpson) from British Columbia. J. Fisheries Research Board Canada 16(5) :76l. Costlow, John D., Jr. 1965. Variability in larval stages of the blue crab, Callinectes sapidus. Biol. Bull. 128(1) :58-66. Garth, John S. 1955. The case for a warm- temperate marine fauna on the west coast of North America. Essays in the Natural Sciences in Honor of Captain Allan Hancock, pp. 19- 27. Univ. Southern Calif. Press, Los Angeles. I960. Distribution and affinities of the brachyuran Crustacea. Systematic Zool. 9(3) : 105-123. Gurney, Robert. 1942. Larvae of Decapod Crustacea. Ray Society, London. 306 pp. Haig, Janet. 1955. The Crustacea Anomura of Chile. Rept. Lund Univ. Chile Expedition 1948-49, no. 20. 68 pp. Johnson, Martin W., and W. M. Lewis. 1942. Pelagic larval stages of the sand crabs Emerita analoga (Stimpson), Blepharipoda occidentalis Randall and Lepidopa myops Stimpson. Biol. Bull. 83(1) :67-87. Menon, M. Krishna. 1933. The life-histories of decapod Crustacea from Madras. Bull. Madras Govt. Mus., New Series, Nat. Hist. Section 3(3): 1-45. Provenzano, Anthony J., Jr. 1962^. The larval development of the tropical land her- mit Coenobita clypeatus (Herbst) in the laboratory. Crustaceana 4(3) :207-228. PACIFIC SCIENCE, Vol. XXI, January 1967 1962 b. The larval development of Cal- ! cinus tibicen (Herbst) (Crustacea, Anomura) in the laboratory. Biol. Bull. 123(1) :179- ! 202. Rees, George H. 1959. Larval development of the sand crab Emerita talpoida (Say) in the laboratory. Biol. Bull. 117(2) :356-370. Sankolli, K. N. 1965. Laboratory study on the life history of mole crab, Emerita hol- thuisi Sankolli. Abstracts of Papers, Sympo- sium on Crustacea, Marine Biol. Assoc. In- dia. 85 pp. Schmitt, Waldo L. 1935. Crustacea Macrura and Anomura of Porto Rico and the Virgin Islands. Scientific Survey of Porto Rico and the Virgin Islands, New York Acad. Sci., Vol. 15, pt. 2, pp. 124-227. Smith, Sidney I. 1877. The early stages of Hippa talpoida , with a note on the structure of the mandible and maxillae in Hippa and Remipes. Trans. Connecticut Acad. Sci. 3:311-342. Snyder, H. George, and A. Fleminger. 1965. A catalogue of zooplankton samples in the marine invertebrate collections of Scripps Institution of Oceanography. Univ. California, Scripps Institution of Oceanog- raphy, SIO Ref. 65-14. Steinbeck, John, and E. F. Ricketts. 1941. The Sea of Cortez. Viking Press, New York. 598 pp. University of California, Scripps Institu- tion of Oceanography. 1952^. Prelimi- nary report of temperature, salinity and cur- rent measurements, cruise 32, 25 Novembei to 9 December, 1951. SIO Ref. 52-1. 1952A Preliminary report of tempera- ture, salinity and current measurements, cruise 33, 8-28 January, 1952. SIO Ref. 52-7. 1961. Physical and chemical data re- port, ccofi cruise 5612, 24 November to 21 December 1956. SIO Ref. 61-22. 1962. Physical and chemical data re- port, ccofi cruise 6108, 13-28 August 1961. SIO Ref. 62-16. Wyrtki, Klaus. 1965. Surface currents of the Eastern Tropical Pacific Ocean. Inter-Ameri- can Tropical Tuna Commission, Bull. 9(5): 271-304. Chromatophore Responses in Relation to the Photoperiod and Background Color in the Hawaiian Ghost Crab, Ocypode cerato phthalma (Pallas)1 Georgiandra Little2 The functional activities of chromato- phores are classified as morphological when there is a change in the amount of pigment present over a period of time, and as physiolog- ical when there is a relatively rapid change due to changes in the degree of dispersion of the pigment. A primary chromatophore response is a response to a nonvisual stimulus, while a secondary response is a response to a visual stimulus (Fingerman, 1963:8). The most common method of describing the degree of dispersion of the chromatophore pig- ments is that of Hogben and Slome (1931:12, fig. 1) using a one to five scale, where one corresponds to maximum concentration and five to maximum dispersion of the pigments, and the intermediate stages are described as two, three, and four. The present study deals with the ability of the Hawaiian ghost crab, Ocypode ceratoph- thalma (Pallas), to maintain a rhythmic physio- logical chromatophore response with different periods of light and darkness and on different backgrounds. The author wishes to thank Dr. Ernst Reese at the University of Hawaii and Dr. John D. Costlow, Jr. and Dr. Edward C. Horn at Duke University for their kind assistance and many helpful suggestions. MATERIALS AND METHODS Male and female ghost crabs were collected from three different areas on the island of Oahu, where the beaches are composed of very fine 1 This work was supported in part by an award granted to the Department of Zoology, Duke Uni- versity by the National Science Foundation Under- graduate Science Education Program (G-21879). Manuscript received December 1, 1965. 2 Department of Zoology, Duke University, Dur- ham, North Carolina. sand presenting a white background. Large crabs, with a carapace length of more than 22 mm, and medium crabs, with a carapace length of less than 22 mm, were collected. A length of 22 mm was chosen as the dividing point be- tween large and medium crabs because ghost crabs begin to develop eye stiles, a characteris- tic of mature crabs, when they reach a carapace length of more than about 20 mm (Crane, 1941:303). Several attempts were made to use small crabs with a carapace length of less than 12 mm, but these tiny crabs almost always died within 18 hr, making any prolonged experimen- tation impossible. Since no apparent difference was observed in the chromatophore responses of the two crab sizes, the results have been com- bined. Crabs were caught either at night or shortly after dawn by chasing them with nets or by digging them out of their burrows. They were placed immediately in individual 8X4x4 inch plywood boxes with y2 inch mesh wire tops, filled about % inch deep with sand. These crabs were returned to the laboratory within an hour after collection, and a petri dish of sea water was put in each box. Within 8 hr after return to the laboratory the crabs were placed on one of the following regimes. White background, normal photoperiod White background, reversed photoperiod White background, constant darkness White background, constant illumination Black background, normal photoperiod Black background, reversed photoperiod Black background, constant darkness Black background, constant illumination During an experiment the animals’ black chromatophores on the proximal, anterior sur- face of the third walking leg were indexed, using the Hogben and Slome (1931) indexing 77 78 PACIFIC SCIENCE, VoL XXI, January 1967 scale at 6-hr intervals with the aid of a Kyowa dissecting microscope (X 45) and a Spencer illuminator. The crabs were kept in their in- dividual wooden boxes with a petri dish of sea water to which fresh sea water was added every 6 hr if needed. They were fed daily, either a small bit of raw meat or bread soaked in milk. The normal photoperiod was daylight and nighttime with the animals kept outside the laboratory. Two indexing time patterns were used. The 0600, 1200, 1800, 2400 sequence provided three indexing periods in daylight and one in the dark. The 0500, 1100, 1700, 2300 sequence provided two daylight and two dark indexing periods. The animals in constant dark- ness were kept in a dark closet and were exposed to the light only for the few seconds it required to index their chromatophores every 6 hr. A 15-watt Westinghouse cool white fluorescent light in the closet was used to maintain the animals in constant illumination with the ani- mals kept at a distance of 3.5 ft from the light source. The artificial light and the dark closet were also used to reverse the normal photo- period by keeping the animals in the dark for 12 hr during the day and turning on the light for 12 hr at night. White sand from Oahu’s beaches provided the white background used in the experiments. Black sand from the island of Hawaii, origi- nating from black lava, was used in the experi- ments where a black background was needed. As a final experiment an attempt was made to see if the observed chromatophore responses were the result of visual stimuli. Medium crabs were collected and maintained on a white back- ground. In half of these crabs the eyestalks were completely covered with dark red Revlon nail polish. The eyestalks of the other half of the crabs were covered with clear Revlon polish. The animals were kept outside the laboratory in the normal photoperiod and their chroma- tophores were indexed every 6 hr at 0600, 1200, 1800, 2400. RESULTS The data are presented in Figures 1-6. Time of day is given on the abscissa, and the chroma- tophore rating scale of Hogben and Slome (1931) is given on the ordinate. Unless other- wise indicated, 10 crabs were used for each experiment. The data have not been treated [ statistically. In the normal photoperiod, using the index- ing times of 0600, 1200, 1800, and 2400, 20 crabs maintained on a white background dis- played a daily rhythmic chromatophore change with maximum pigment dispersion at 0600 and minimum dispersion at 2400 (Fig. 1). The 20 crabs maintained on a black background under the same light conditions displayed a daily rhythm with a peak at 1200 and a low point at 2400 (Fig. 1), though they displayed much less marked response. Crabs on a black back- ground were generally darker at any hour. Crabs were maintained under the same light conditions with the chromatophores indexed at 0500, 1100, 1700, and 2300 in order to have two indexing periods in the dark (2300 and 0500) and two periods in the light (1100 and 1700). Crabs on white sand maintained a rhythmic change where a peak was reached during the day and a low point at 2300, while crabs on black sand displayed a daily rhythmic change with a similar pattern of peaks and low points (Fig. 2), but with less total variation. These crabs were generally darker at all times. As shown in Figure 3, crabs maintained in the reversed photoperiod showed a reversal of their daily rhythmic changes. The crabs were in total darkness for 12 hr, including the 1100 and 1700 indexing times, and in constant arti- ficial light for 12 hr, including the 2300 and 0500 indexing times. On both black and white backgrounds the crabs displayed more pigment dispersion in the simulated daytime than in the artificial night. In comparing the reversed (Fig. 3) and normal (Fig. 2) photoperiods, the 0500, 1100, 1700, 2300 normal photoperiod results have been used so that both experiments have two indexing times in the light and two in the dark. Comparison of the chromatophore indices of crabs during the two photoperiods showed that the peaks of the normal photoperiod oc- curred simultaneously with the low points of the reversed photoperiod and vice versa, indi- cating that the chromatophore rhythms were reversible. Crabs maintained on white sand dis- played this reversal more distinctly than crabs maintained on black sand. Crabs on a black background displayed a greater degree of pig- ment dispersion at all times. Chromatophore Responses — Little 79 5 X UJ D LxJ tr O x CL O < o cc X u 4 3 2 1 l I i I J i l i i i I i i i 0600 1800 0600 1800 0600 1800 0600 TIME OF DAY Fig. 1. The average daily indices of the darkly pigmented chromatophores in Ocypode ceratophthalma. The crabs were maintained in the normal photoperiod for 3 days on a white or a black background. The chromatophores were indexed every 6 hr. 0500 1700 0500 1700 0500 1700 0500 TIME OF DAY Fig. 2. The average daily indices of the darkly pigmented chromatophores in Ocypode ceratophathalma. The crabs were maintained in the normal photoperiod for 3 days on a white or a black background. 80 PACIFIC SCIENCE, Vol. XXI, January 1967 X LU a z LU cl O x CL o I— < 0 CL 1 u TIME OF DAY Fig. 3. The average daily indices of the darkly pigmented chromatophores of Ocypode ceratophthalma when the crabs were maintained in the artificially reversed photoperiod for 4 days on a black or white back- ground. X LJ o LU CL o X CL o o CL X U 5 4 3 2 sand sand 0600 1800 0600 1800 0600 1800 0600 TIME OF DAY Fig. 4. The average daily indices of the darkly pigmented chromatophores in Ocypode ceratophthalma when the crabs were maintained in constant darkness for 3 days on a black or a white background. Chromatophore Responses — Little 81 Figure 4 shows the results of maintaining crabs in constant darkness. In crabs maintained on white sand a daily chromatophore change rhythm similar to the one displayed during the normal photoperiod was observed, though it was at a lower level on the Hogben and Slome (1931) scale (compare with Fig. 1). The chromatophore rhythm of crabs maintained in constant darkness on black sand became irregu- lar (Fig. 4). When crabs were maintained on a white back- ground under constant illumination, a small rhythmic chromatophore change was observed the first day, but then it began to decay. In crabs maintained under similar light conditions on a black background the rhythm died out (Fig. 5). As in preceding experiments, crabs on the dark background displayed a consistently greater degree of pigment dispersion. Figure 6 shows the results of covering the eyestalks of medium-sized crabs with either dark red nail polish or clear nail polish and maintain- ing them on a white background in the normal photoperiod. When the chromatophores were indexed at night, the two observed indices were very close, but when the chromatophores were indexed in the light, the crabs with the clear nail polish on their eyestalks showed a greater degree of pigment dispersion (Fig. 6), at least during the first two days. Toward the end of the experiment the difference in response between the two experimental groups diminished appre- ciably. This phenomenon was attributed to chipping of the red nail polish. DISCUSSION Other studies have been conducted on the existence of persistent rhythmic physiological chromatophore changes. An endogenous rhythm in the fiddler crab Uca has been reported by Brown and Sandeen (1948:370). The persis- tence of this rhythm in total darkness has been tested (Brown, Fingerman, Sandeen, and Webb, 1953:36), and Webb (1950:336) found that the rhythm could be altered by artificially chang- ing the normal time of night and day. The results obtained in these experiments suggest that the Hawaiian ghost crab exhibits a daily rhythm of chromatophore changes under normal conditions, with maximum concentra- tion of the dark pigment at night and maximum dispersion of this pigment during the day. In most chromatic decapods the dark pigment in X LU Q Z LU oc 0 1 Q_ o h- < 2 0 DC 1 O TIME OF DAY Fig. 5. The average daily indices of the darkly pigmented chromatophores in Ocypode ceratophthalma when the crabs were maintained in constant fluorescent illumination for 4 days on a black or a white back- ground. 82 PACIFIC SCIENCE, Vol. XXI, January 1967 X LlJ Q 5 4 LjJ cr 0 1 CL O h- < X o cr x U 1 J 1 L 1 I 1 i 1 1 | | 1 i | I i. . I 0600 1800 0600 1800 0600 1800 0 600 1800 0600 TIME OF DAY Fig. 6. The average daily indices of the darkly pigmented chromatophores in medium-sized Ocypode cera- tophthalma when the crabs were maintained in the normal photoperiod on white sand for 4 days with eye- stalks covered with clear or red nail polish. the chromatophores concentrates at night (Parker, 1948:51). On white sand this concen- tration of pigment at night enabled the crabs to blend with the white background. On black sand there was less blanching at night. This definite change in the normal pattern of chroma- tophore responses would indicate that the re- sponses are flexible and are influenced not only by the light intensity but by the color of the background as well. Smith (1938:252) observed in Ligia that a sudden change in the color of the background was frequently accompanied by a chromatophore response in which the animal gradually adapted to the new background color. Studies on Uca (Brown and Sandeen, 1948:366) showed a greater dispersion of dark pigment in animals on a dark background than on a light back- ground when light intensity was the Same. Constant laboratory conditions were used for the reversed photoperiod experiments. The chromatophore rhythm which was established corresponded to the artificial night and day, with the animals being darkest during the arti- ficial day and lightest during the artificial night, indicating that the rhythmic chromatophore changes can be reversed. Brown (1961:510) reported that a daily rhythm of chromatophore responses will persist under conditions of constant darkness in many species. In the experiment reported here the re- sults showed that a diurnal rhythm was main- tained, although at a lower level on the Hogben and Slome scale when the crabs were main- tained on white sand. The normal environment of these crabs is on white sand and in constant darkness, for they are active at night and spend the day underground. The conditions of the experiment conducted in total darkness were, therefore, quite similar to those of their natural surroundings. During the experimental period the crabs were very active, behaving as they normally do at night. Although the chroma- tophore rhythm was not destroyed under the conditions of constant darkness, it was evident toward the end of the last day that the pattern had started to decay. The crabs maintained in total darkness on black sand, however, showed very little change in pigment concentration with respect to time. Because the black sand is much coarser than the white, one must include the possibility that substrate characteristics other than reflectivity might influence chromatophore responses. Chromatophore Responses — Little 83 The observed daily chromatophore rhythm did not persist in the experiments in which the crabs were maintained in constant illumination. The light conditions of these experiments were completely foreign to the animals, and the crabs behaved sluggishly. There were no cues from the environment to aid in maintaining the rhythm, and it did not persist. There was no evidence of a persisting diurnal rhythm. The amount of illumination was the same for both groups of crabs, yet those on black sand showed a consistently greater degree of pigment disper- sion, showing that there is a response to the color of the background. In the final experiment an attempt was made to determine whether the observed daily re- sponses were responses to a visual stimulus. Cowles (1905:23-24) reported that after paint- ing the eyestalks of Ocypode with lampblack, no further color changes were observed. This observation, however, was not quantitative, for he observed only gross appearance and did not describe the condition of the chromatophores. In the present study the eyestalks of 10 medium crabs were covered with dark red nail polish. While it was uncertain whether this treatment blocked all the light transmission, the response suggested that the red coating effectively re- duced visual reception. Normally the ghost crab demonstrates a shadow reflex, depressing its eyestalks when moved from the light into the shade or vice versa. The crabs whose eyestalks were painted with clear polish did demonstrate the reflex; the crabs whose eyestalks were painted red did not. The degree of pigment dispersion in the red-painted crabs was much less. Clear nail polish may have cut out some of the light, which might account for the de- crease in amplitude of the daily chromatophore rhythm as compared with crabs with normal eyestalks. The results, however, suggest that the visual reception of light is an important factor in maintaining a daily rhythm of chromatophore changes. SUMMARY 1. The Hawaiian ghost crab, when main- tained on white sand, demonstrates a daily rhythm of chromatophore changes with maxi- mum dispersion of dark pigment during the day and maximum concentration at night. On a black background the same daily rhythm of chromatophore changes is observed, but there is generally less concentration of pigment at all times. 2. In an artificially reversed photoperiod the crabs on both black and white backgrounds dis- play a reversal of the daily rhythm of chromat- ophore changes. 3. In constant darkness crabs on white sand still display the daily rhythm but at a lower over-all level of pigment dispersion; on black sand the rhythm becomes irregular in constant darkness. 4. Under conditions of constant illumination crabs maintained on both backgrounds show little if any rhythm of chromatophore dispersion and concentration. 5. The observed chromatophore responses are primarily responses to visual stimuli, al- though in the absence of light evidence is given for an endogenous rhythm and for alteration of rhythm by substrate. REFERENCES Brown, F. A., Jr. 1961. Chromatophores and color changes. In: C. L. Prosser and F. A. Brown, Jr., eds., Comparative Animal Physi- ology. W. B. Saunders Co., Philadelphia. Chap. 19. M. Fingerman, M. Sandeen, and H. M. Webb. 1953. Persistent diurnal and tidal rhythms of color changes in the fiddler crab, Uca pugnax. J. Exptl. Zool. 123:29-60, 10 figs. and M. Sandeen. 1948. Responses of the chromatophores of the fiddler crab Uca to light and temperature. Physiol. Zool. 21:361-371, 5 figs. Cowles, R. P. 1905. Habits, reactions and associations in Ocypode arenaria. Papers from the Tortugas Laboratory, Carnegie Inst. Washington 2:3-39, 4 pis., 10 figs. Crane, J. 1941. On the growth and ecology of the brachyuran crabs of the genus Ocypode. Zoologica 26:297-310, 2 pis., 7 figs. Fingerman, M. 1963. The Control of Chro- matophores. The Macmillan Co., New York. Hogben, L. T., and D. Slome. 1931. Pigmen- tary effector system, VI. Proc. Roy. Soc. London B, 108:10-53, 2 pis., 10 figs. 84 PACIFIC SCIENCE, Vol. XXI, January 1967 Parker, G. H. 1948. Animal Color Changes and Their Neurohumors. Cambridge Univer- sity Press, Cambridge. Smith, H. 1938. Reception mechanisms of background responses in chromatic behavior of Crustacea. Proc. Roy. Soc. London B, 125:249-263, 1 pi., 9 figs. Webb, H. M. 1950. Diurnal variations of re- sponse to light in the fiddler crab, Uca. Physiol. Zool. 23:316-337, 21 figs. Overt Responses of Polychoerus carmelensis (Turbellaria: Acoela) to Abrupt Changes in Ambient Water Temperature Robert G. Schwab1 Knowledge of an animal’s response to a change in the ambient environment contributes to an understanding of its behavior, activity pattern, and methods of survival. Responses to environ- mental stimuli by triploblastic animals having a comparatively low order of tissue/organ devel- opment as in the Acoela has special evolutionary significance because, according to the hypothesis of Hadzi, they are the stem group of the Eumetazoa and were derived from ciliates (de Beer, 1954; Hanson, 1958). Of special interest are the flatworms which inhabit tide pools of the midtide horizon, where ambient temperatures may fluctuate because of shallow, relatively nonturbulent water. Such is the habi- tat of the Acoela flatworm Polychoerus car- melensis in the vicinity of Monterey, California (Ricketts and Calvin, 1952). During low tide this species is often abundantly present on algae- covered rocks. At high tide Polychoerus takes shelter under rocks and gravel, apparently in response to water turbulence caused by the in- coming tide. Because of potentially pronounced environmental changes within its ecosystem, Polychoerus was selected for study of the overt responses by an exothermic marine animal to changes in the ambient water temperature. Dr. Donald P. Abbott, Assistant Director of the Hopkins Marine Station, contributed several much-appreciated suggestions during this study. METHODS Specimens of P. carmelensis collected at Point Pinos (vicinity of the Hopkins Marine Station of Stanford University, Pacific Grove, California) were transported to the University of California Animal Physiology Laboratory at 1 Department of Animal Physiology, University of California, Davis, California. Manuscript received December 3, 1965. Davis, California. One group of animals was maintained for 24 hr and another for 48 hr in a darkened chamber at a temperature of 13°- 14°C. This temperature approximates that of Monterey Bay in late spring. In the following tests an individual flatworm was removed from the chamber, placed on a horizontal plastic grid, and quickly submerged to a depth of 1 cm in a controlled-temperature sea water bath. During the experiments the water temperature was maintained within ±0.5°C of the desired temperature at the upper surface of the sub- merged plastic grid. The amount of illumination from fluorescent room lights at water level was constant at 60 ft-c throughout all tests. This was sufficient to induce a photokinetic response from the dark-conditioned animals. As an animal moved across the plastic grid, its horizontal movements during a 30-second period were transcribed by the author onto a record sheet grid. Such a record was obtained for each individual tested. All animals were allowed to travel a distance of about 1 cm be- fore the record tracings were initiated. The total distance traveled, often in a highly erratic course, was measured from the record sheet grid; the rate of locomotion was computed in mm/min. After being dark-conditioned for 24 hr, 10 individuals were tested at each of the following temperatures: 3°, 5°, 8°, 11°, 14°, 17°, 21°, 25°, 29°, 33°, 35°, and 38°C. No further tests were made on any individual, once its lo- comotion rate at a specific temperature was de- termined. On the following day, each test was repeated using 10 animals dark-conditioned for 48 hr. There was no apparent difference in the average rate of locomotion related to duration of conditioning. Therefore, the locomotion rate data for all 20 individuals tested at each specific temperature were consolidated. 85 86 PACIFIC SCIENCE, VoL XXI, January 1967 RESPONSES TO AMBIENT WATER TEMPERATURE Reactions to Cold Temperature P. carmelensis placed in sea water at a tem- perature of 3°C immediately contracted into a U-shaped posture with normally ventral portion of the animal forming the inside surface of the U. Few animals exposed to this temperature had noticeable muscular or ciliary motion while in this posture. Thus, the well-developed auto- maticity normally associated with ciliary motion (Prosser and Brown, 1961) apparently did not occur in P. carmelensis exposed to sea water at 3°C After several minutes’ exposure, disinte- gration of the epidermal cells occurred, and shortly thereafter a gentle motion of the water caused by stirring with a probe resulted in dis- organization of the body structure. All polychoerus exposed to an ambient water temperature of 5°C contracted into the U- shaped posture and were motionless for several minutes. Thereafter most were capable of loco- motion (it was necessary to test 24 individuals to obtain locomotion rates for 20 animals). These animals were motile long enough to mea- sure locomotion rates ranging from 5.0 to 32.5 mm/min with an average of 17.4 mm/min. This was the lowest average obtained in these tests. At the same temperature, 16 individuals moved about while in the U-shaped posture. Apparently this movement was accomplished entirely by motion of cilia on the dorsal surface, since no muscular contractions were noticed as the animals glided over the surface of the plas- tic grid. These animals moved about for only a minute or two and thereafter tissue disintegra- tion took place as described above. The highest rates of locomotion at 5°C were obtained from 4 animals that, after a short period in the U- shaped posture, moved in the typical flatworm posture. Within a few minutes movement ceased, whereupon they again contracted into the U-shaped posture and died. Reactions to Changes in Ambient Water Temperature The locomotion rate of P. carmelensis was clearly influenced by the temperature of am- bient sea water under the conditions of these experiments. However, there were pronounced changes in the manner in which locomotion occurred. Several of the 20 animals exposed to water at 8°C began to move while in the U- shaped posture. Locomotion was accomplished by action of the dorsal cilia, the only portion i of the body in contact with the plastic grid. These animals soon reoriented to the typical flatworm posture and the rate of locomotion at 8°C was measured from this posture only. At this temperature Polychoerus traveled at an av- erage rate of 44.9 mm/min. Accelerated loco- motion rates associated with increases in am- [ bient water temperatures were measured at 11°, 14°, and 17°C with average values of 64.8, 83.0, and 90.4 mm/min respectively. Thus, the average rate of locomotion for P. carmelensis acclimated at 13°-l4°C increased from 17.4 to 90.4 mm/min in response to a 12 -degree rise i (from 5° to 17°C) in temperature. This in- crease in speed of locomotion took place at a nearly uniform rate of 6.1 mm/min/°C in- crease in water temperature (see Table 1). P. carmelensis specimens respond to tempera- tures higher than 17°C by decreasing their rate of locomotion. A reduction in average loco- motion rate was measured at 21°C (78.5 mm/min), 25°C (66.2 mm/min), 29°C (50.2 mm/min), and 33°C (32.8 mm/min). This results in a steady decrease in the speed of locomotion at an approximate rate of 4.4 ! mm/min/ °C rise in temperature between 17° and 33°C. Reactions to Warm T emperatures At an ambient water temperature of 29° C Polychoerus usually contracted into a curled I position with the posterior portion of the body drawn up under the more anterior portion. Lo- comotion in this posture was primarily accom- plished by action of the anterior portion of the body, since much of the posterior portion was not in contact with the plastic grid. The animals had an average locomotion rate of 50.2 mm/min while in this posture and, although several ani- mals died after 5-10 minutes’ exposure, it is reasonable to assume that they would have found a more suitable temperature within this length of time in their natural habitat. Most of the individuals exposed to an am- bient water temperature of 33 °C immediately contracted into the curled posture, mentioned above and were capable of locomotion for only Polychoerus carmelensis and Temperature Changes — Schwab 87 TABLE 1 Reactions of P. carmelensis To Changes in Ambient Water Temperature* TEMPERATURE (°C) LOCOMOTION RATE Average (mm/min) Range STANDARD DEVIATION STANDARD ERROR OF THE MEAN 3 ❖ * 5 17.4 5.0- 32.5 6.83 1.51 8 44.9 25.0- 62.5 9-81 2.19 11 64.8 40.0- 90.0 14.21 3.18 14 83.0 62.5-115.0 14.70 3.25 17 90.4 57.5-122.5 17.06 3.82 21 78.5 45.0- 95.0 13.89 3.10 25 66.2 45.0- 87.5 11.40 2.53 29 50.2 35.0- 62.5 7.41 1.65 33 o er sfe 32.8 17.5- 65.0 11.49 2.57 35** 38** * A total of 20 individuals was tested at each of the temperatures. ** See text. 1-3 minutes. Thereafter disintegration of the tissues took place. Some of the flatworms tested at this temperature immediately formed the U- shaped posture. After about a minute most indi- viduals reoriented to the curled posture and moved about for a minute or two. After this short period of movement they again formed the U-shaped posture and all movement ceased. Animals removed after about 3 minutes’ expo- sure to 33 °C water temperature did not recover when placed in 14°C sea water. An exposure of 5-10 minutes at this temperature results in an apparently complete disorganization of the body structure. Exposure to sea water at a temperature of 35 °C resulted in a very brief but rapid locomo- tion by several of the 20 individuals tested. Generally this occurred while the animal was in the curled posture. However, most individ- uals remained in the U-shaped posture assumed immediately upon contact with the 35 °C water and had no measurable amount of locomotion. All animals showed signs of tissue disintegra- tion within 60 seconds after exposure. Polychoerus exposed to sea water at 38 °C ap- peared to die immediately. Several individuals were dipped into water at this temperature and then quickly returned to 14° C sea water but there were no recoveries. DISCUSSION The locomotion rate of P. carmelensis was clearly related to the water temperature under the conditions of these experiments. The nearly uniform increase of locomotion rate at 6.1 mm/min/°C rise in temperature between 5° and 17°C suggests that changes in tide pool temperatures may have a pronounced effect on the activity and behavior of this species. The mechanism by which temperature induces in- creased locomotion activity is not known. How- ever, it is likely that this accelerated locomotion is fundamentally similar to the increases of chemical and physical reactions normally asso- ciated with an increase in temperature. Many biological processes, including rate of develop- ment, behavioral reactions, speed of locomotion, and metabolism show increases associated with higher temperatures (Prosser and Brown, 1961). It is significant that the highest rate of loco- motion occurred at a temperature (17°C) near that measured in these tide pools during late spring. This suggests that the maximum loco- motion rate of Polychoerus may be a function of the most suitable ambient environmental temperature with respect to possible acclimatiza- tion of the animal. The decrease in locomotion rate measured at ambient water temperatures above 17°C (4.4 mm/min/°C rise in temperature between 17° and 33 °C) is considerably less than that mea- sured for locomotion increases (6.1 mm/min/°C rise in temperature between 5° and 17°C). This suggests a temperature-related differential 88 response rate as well as a differential behavior response. The temperature threshold at which the type of response, increased or decreased locomotion, and the rate at which the response takes place is approximately 17°C for P. car - melensis accustomed to an ambient water tem- perature of 13°-14°C. A possible explanation for this is that accelerated locomotion results from a direct influence by the ambient tempera- ture on body processes, and decreased locomo- tion results as a secondary effect of temperature- related factors, such as the reduced availability of oxygen as the ambient water temperature in- creases. Polychoerus obtains oxygen from the aquatic environment by diffusion through epi- dermal tissues. The physical characteristics of this species are such that sufficient oxygen for metabolic processes should be available from the environment at the higher temperature levels tested if the entire surface of the animal is effectively exposed to the environment. How- ever, it is possible that little oxygen diffuses through the ventral epidermal tissues because of the close proximity of the animal to the surface on which it crawls. This would reduce by nearly one-half the effective diffusion surface and is a possible explanation for the curled position re- sulting from elevated temperatures in that this position exposes about one-half of the ventral epidermal tissue to the oxygen-bearing environ- ment. Locomotion of P. carmelensis at temperatures from 8° to 25 °C inclusive takes place in the typical flatworm-type posture. The highest in- dividual rate of locomotion obtained during these tests (2.08) was noted from a worm tested at 17°C. The lowest average locomotion rate in this temperature range was 0.75 at 8°C, and the highest average of 1.50 occurred at 17°C. These locomotion rates are expressed in mm/sec and were measured at a light intensity of 60 ft-c. They compare closely with the loco- motion rates reported by Armitage (1961) of 0.86 mm/sec and 1.34 mm/sec measured at illumination levels of 6 and 37 ft-c, respec- tively. However, he states that the behavior of Polychoerus was highly erratic during his loco- motion rate tests, in that some individuals spent considerable time turning the head from side to side and others ceased crawling before reaching the end of a 5 -cm course. A possible explana- PACIFIC SCIENCE, Vol. XXI, January 1967 tion of this erratic activity, based on the tem- perature-locomotion rate relationship obtained in my experiments, is that the water in the petri dish used by Armitage in his tests became warm during the course of the observations be- cause of warm room temperature and/or heat from the light source used to stimulate locomo- tion. In the present tests, none of the 120 worms tested between 8° and 25 °C showed such behavioral responses. However, several of those tested at 29 °C and essentially all at 33°C reacted in the erratic manner described by Armi- tage. Survival Value of Temperature-Locomotion Relationship According to Armitage (1961), P. carmelen- sis does not possess a persistent diurnal rhythm. Therefore, this species must depend on an en- vironmental "cue” or a combination of such environmental stimuli to regulate its daily ac- tivity pattern. Observations by Armitage indi- cate that light intensity and water turbulence play a pronounced role in the regulation of ac- tivity and behavior of this animal. Because of j the pronounced influence on the velocity of movement resulting from slight changes in the i ambient water temperature, it seems reasonable that temperature and temperature-related factors may also function as stimuli regulating activity and behavior. Armitage postulates that on bright days the | absence of Polychoerus from the upper surface of rocks and gravel during low tide and rela- tively calm water is caused by an increased i negative phototropic response to high light in- j tensity. However, on April 29, 1965, a cool but very bright day with morning sea water at ! about 15°C, I observed Polychoerus active on the upper surfaces of rocks and gravel through- out the period of low tide. This observation, made at the same location but earlier in the season than that by Armitage, documents the fact that high light intensity does not always cause a negative phototropic response, and sug- gests that there is more than a single factor regulating this behavior. Ambient water tem- peratures above 17°C cause a reduction in the rate of locomotion for animals conditioned to 13°-l4°C. It is likely that this reduced rate of locomotion in response to such temperatures or Polychoerus carmelensis and Temperature Changes — Schwab 89 temperature-related factors corresponds to a less suitable ambient environment, and that such conditions in their natural environment may cause P. carmelensis to vacate the upper sur- faces of rocks and gravel. In my experiments Polychoerus specimens were totally incapacitated soon after exposure to ambient water temperatures above 29 °C. It is reasonable to assume that environmental fac- tors would trigger a behavioral escape mechan- ism should such temperatures occur in their natural habitat. The survival value of such an environmental stimulus or combination of stim- uli is dependent upon the sensitivity and re- sponse of the animal to this factor or factors. Armitage reports a 55% increase in rate of crawling when the light intensity was increased 640%, and that in his tests P. carmelensis was negatively phototropic. Although the magnitude of the light intensity change (in ft-c or in per- cent increase) and the sensitivity of the animals in terms of response to this factor (rate of crawling in mm/sec or percent increase) cannot be directly compared to similar calculations with respect to temperature change and asso- ciated activity response, it seems certain that Polychoerus is at least as sensitive to ambient water temperature as it is to light intensity. Costello and Costello (1938) indicate that P. carmelensis may be positively phototropic in that individuals showed a tendency to group on the moderately lighted side of an aquarium. Therefore, there is evidence of both positive and negative phototropic response for this spe- cies. Armitage suggested that this species may have a differential response to low and high light intensities. Whether such a light sensi- tivity threshold exists, or whether the behavioral evidence supporting this possibility results from a water temperature-light intensity relationship, is at present unknown. The physical properties of water are such that the heat energy associated with even relatively high light intensities may have little immediate effect on the temperature of the tide pool. Con- versely, tide pool temperatures may in time be- come elevated on overcast days with relatively low light intensities. Thus, it is possible that the absence of this species from tide pools on bright days as observed by Armitage may be at least in part the result of an elevated ambient water temperature associated with high insola- tion, rather than the result of high illumination levels as a discrete factor. Regardless of the nature of causative stimuli, the rate of locomo- tion between 8° and 29°C is sufficient for Polychoerus to seek a more desirable situation under rocks and gravel should the tide pool environment warrant such behavior. Environment-Induced Posture Responses The U-shaped posture resulting from low, high, and often from moderately high ambient water temperatures appears to be the same posture described by Armitage for Polychoerus exposed to osmotically unsuitable salinity con- centrations. Apparently this species has only the U-shaped posture response to both hypo- and hypertonic sea water, since Armitage does not mention other postures such as the curled pos- ture observed in my experiments at certain elevated water temperatures. During my experiments the U-shaped pos- ture often occurred as a response to a gentle mechanical stimulus from contact with a soft- bristled brush used to transfer the animals from one container to another. The fact that this posture occurs as a result of temperature change, salinity change, and mechanical stimulus sug- gests that this posture is a characteristic behav- ior response to many undesirable environmental conditions. Further experimentation may show that the curled posture also occurs as a response to a variety of stimuli. However, it is certain that this response is clearly associated with ambient water temperatures near the upper lethal level. It is not known if the curled-posture response occurs as a direct result of temperature percep- tion by a discrete thermoreceptive mechanism or as a characteristic response to temperature- related environmental factors. It may be that the curled posture common at 29°C functions to increase the amount of surface area in direct contact with the aquatic environment, thus facili- tating gaseous exchange, while enabling the animal to take shelter. The U-shaped posture exposes even more surface area but has no ob- vious survival value at temperatures above 29 °C, inasmuch as the locomotion rate is low and the animal soon dies. However, this con- clusion is hypothetical, since possible advan- 90 tages of either posture to animals exposed to high water temperatures are not known. CONCLUSIONS 1. The ambient water temperature clearly influences both the rate and the sign of Poly- choerus’ locomotion. A rate-directional response threshold was measured at 17°C for animals conditioned to 14°C in that, as temperatures increased from 5° to 17°, the speed of locomo- tion increased from an average of 17.4 to 90.4 mm/min, at the rate of 6.1 mm/min/°C. Fur- ther temperature increases from 17° to 33° C caused a steady decrease in locomotion speed from an average of 90.4 to 32.8 mm/min, and the rate of reduction per degree of temperature increase, — 4.4 mm/min, was relatively low. 2. The highest average speed of locomotion (90.4 mm/min) and the greatest individual rate (122.5 mm/min) were measured at 17°C, which is near the temperature of sea water at the location and season at which these experi- ments were conducted. This suggests the pos- sibility that the maximum locomotion rate is a function of the ambient water temperature with respect to possible seasonal acclimatization by the animal. 3. It is postulated that the increased rate of locomotion as temperatures changed from 5° to 17°C corresponds to a general temperature- related acceleration of the body processes. It is not known why temperature increases above 17°C cause a reduction of locomotion speed. Possibly this is a function of decreased amounts of available oxygen due to elevated water 'tem- peratures and/or crossing a critical temperature threshold for enzymatic action. 4. Polychoerus has a differential movement posture with respect to high and low water temperatures. At 5°C this species contracts into a U-shaped position, and movement at an aver- age rate of 17.4 mm/min results from motion of the cilia on the dorsal surface of the animal. In ambient sea water temperatures of 29° C and PACIFIC SCIENCE, Vol. XXI, January 1967 above, movement generally occurs while the animal is in the usual flatworm position, but with the posterior portion of the body drawn | up under the more anterior portion. In this po- sition the average rate of movement was 50.2 [ mm/min at 29°C and 32.8 mm/min at 33°C ! 5. Sea water temperatures slightly below 5°C and above 29 °C are not suitable for the survival of P. carmelensis conditioned at a temperature of about 14°C. However, the rate of locomotion at these temperatures appears | sufficient to allow this species to avoid such conditions should they occur in the tide pool. Measurements of temperatures prevailing be- neath the rocks and gravel in these tide pools is needed. 6. It is probable that the temperature of |! the ambient sea water, as well as the intensity of illumination and the turbulence of water, function as an environmental stimulus regulat- ! ing the activity and behavior of Polychoerus . REFERENCES !; Armitage, K. B. 1961. Studies of the biology j of Polychoerus carmelensis (Turbellaria: j: Acoela). Pacific Sci. 15:203-210. Costello, H. M., and D. P. Costello. 1938. A new species of Polychoerus from the Pa- cific Coast. Ann. Mag. Nat. Hist., Ser. 11, 1:148-155. De Beer, G. R. 1954. The evolution of the metazoa. In: Julian Huxley et ah, eds., Evo- ; lution as a Process. George Allen and Unwin, London. 367 pp. Hanson, E. D. 1958. On the origin of the Eumetazoa. Systematic Zook 7:16-47. Prosser, C. L», and F. A. Brown, Jr. 1961. :: Comparative Animal Physiology. 2nd ed. W, B. Saunders Co., Philadelphia. 688 pp. Ricketts, E. F., and J. Calvin. 1952. Between I Pacific Tides. 3rd ed. Rev. by J. W. Hedg- peth. Stanford University Press, California. 502 pp. The Osteology of the Congrid Eel Gorgasia punctata and the Relationships of the Heterocongrinae1 Richard H. Rosenblatt2 ABSTRACT : The osteology of Gorgasia punctata is described, figured, and com- pared with that of other congrids. Gorgasia is clearly referable to the subfamily Heterocongrinae. The heterocongrines agree with the Congridae in several important features, and do not differ in fundamental respects. Therefore, the group is recog- nized as a subfamily of the Congridae. Gorgasia is the most primitive hetero- congrine, and agrees with the anagoine congrids in having a lateral ethmoid process. Because of this and other similarities it is suggested that the Anagoinae and Heterocongrinae arose from a common stem. The genus Xarifania was erected on the erroneous basis of lack of caudal rays. The congrid eel Gorgasia punctata was placed by its describers in the little known apodal family Derichthyidae (Meek and Hildebrand, 1923). Bohlke (1951) was the first to point out that the affinities of Gorgasia were with Heteroconger. Gosline (1952) provisionally placed Heteroconger and Gorgasia in the Con- gridae. Bohlke (1957) described the osteology of the related Nystactichthys halis and placed the eels allied to Heteroconger in the Congridae, but considered them to constitute the distinct subfamily Heterocongrinae. Bohlke considered Gorgasia to be the most primitive genus of the Heterocongrinae on the basis of its more complete complement of head pores, its uncoalesced upper labial flanges (called by him the "free edge of lip”), a well- developed pectoral fin, and its unspecialized maxillary dentition. He considered Gorgasia to be specialized, however, in that the caudal rays are much reduced and covered by thick skin. Internal characters were not considered, since the only complete specimen then available was the holotype. Recent collections made by personnel of the Scripps Institution of Oceanography have amassed rich material of several species of 1 Contribution from the Scripps Institution of Oceanography, University of California, San Diego, and the Institute of Fisheries, University of British Columbia. Manuscript received December 22, 1965. 2 Scripps Institution of Oceanography, University of California, San Diego, La Jolla, California. It is synonymized with Taenioconger. eastern Pacific heterocongrines, including Gor- gasia punctata. Because of previous uncertainties regarding the exact position of Gorgasia in eel classification, a study of the osteology of this species seemed worthwhile. ACKNOWLEDGMENTS The figures were drawn under my supervision by E. David Lane. Part of the cost of this investigation was defrayed by a grant to the Institute of Fisheries, University of British Columbia, from the H. R. McMillan Expedition- ary Fund. The specimen of T aenio conger has si used in this study was collected as a part of the U. S. Biological Program, Indian Ocean Ex- pedition. MATERIALS AND METHODS The two adults of Gorgasia punctata were taken from a series of 69 (SIO62-720-26A, Bahia Magdalena, Baja California, Mexico). These were bone-stained with alizarin and cleared in glycerine. The neurocranium was dis- sected out of one specimen, and the drawings were made from the dried preparation. In addition I have utilized single-stained and cleared specimens of Taenioconger digueti Chabanaud (SI065-278, Gulf of California), T. herrei Wade (SI061-261, Gulf of Califor- nia), T. hassi 3 (Klausewitz and Eibl-Eibesfeldt) 3 Taenioconger hassi was originally described in the genus Xarifania, of which it is the type species 91 92 PACIFIC SCIENCE, Vol. XXI, January 1967 (unnumbered, D’Arros Island, Amirantes Is- lands), T. n.sp. (81062-42, Bahia Banderas, Mexico), and Ariosoma gilberti (Ogilby) (SI062-77, Sinaloa, Mexico). OSTEOLOGY neurocranium (Fig. l): The skull is truncated posteriorly, except where the exoc- cipital flanges break the outline. The pre- maxillaries, ethmoid, and vomer are fused, with no suggestion of articulations, such as were reported by Bohlke (1957) for Nystactichthys halts. The anterior, triangular tooth-patch may represent the premaxillary dentition. The dorsal or ethmoid portion of the complex is very thin, and is reduced medially to a septum, so that it is shaped much like an I-beam in cross section. Laterally the ethmoid portion is expanded, and gives rise to two heavy, forward-curving pro- cesses. These are very similar to the "lateral ethmoid processes" reported by Asano (1962) for Anago and Alloconger. Below this are two small projections from the lateral face of the vomer. These vomerine processes are difficult to distinguish from the base of the lateral ethmoid process. They are more evident in Taenioconger and, judging from Bohlke’s figures 3B and 3C, are developed in Nystactich- thys as well. Posteriorly on the under side of the cranium, parts of the prootic and basioccipital are expanded to form a prominent auditory bulla, which contains a large otolith (presum- ably the sagitta). The foramen magnum is surrounded by exoccipital flanges, which grasp the first vertebra. The supraoccipital is well developed, but does not completely separate the epiotics, which are in contact posteriorly. The parietals are sutured in the specimen figured, (Klausewitz and Eibl-Eibesfeldt, 1959). The sole distinction of the genus Xarifania was the supposed lack of caudal rays. The tail-tip of X. has si is fleshy but flexible and clearly contains well-developed caudal rays. These are visible under direct light and are ob- vious when transmitted light is used. In addition to the Amirantes specimen, I have examined a paratype of X. has si (ansp 94706) through the courtesy of J. Bohlke. I can find no other important differences, either in external morphology or osteology, between X. has si and the species of Taenioconger examined. The nominal genus Xarifania is considered, therefore, to be a synonym of Taenioconger. Fig. 1. Head skeleton of Gorgasia punctata. A, Lateral view, including pectoral girdle; B, neurocran- ium, lateral view; C, neurocranium, top view; D, neurocranium, bottom view. A A, Articular angular; BO, basioccipital; BR, branchiostegal ray; CL, cleith- rum; CO, circumorbital ; D, dentary; EO, exoccipital; EP, epiotic; FR, frontal; HMD, hyomandibular ; HOC, hypocoracoid; HRC, hypercoracoid; IOP, in- teropercle; LEP, lateral ethmoid process; MX, maxil- lary; N, nasal; OBS, orbitosphenoid ; OP, opercle; PAR, parietal; PAS, parasphenoid ; PEV, premaxil- lary ethmovomerine block; POP, preopercle; PP, palatopterygoid ; PR, pectoral ray; PRO, prootic; PT, pterygiophore; PTO, pterotic; PTS, pterosphen- oid; QU, quadrate; SCL, supracleithrum;. SO, supra- occipital; SOP, subopercle; SPO, sphenotic. but in another they are fused for the anterior one-quarter of their lengths. The frontals are completely fused, with no sign of a suture or median ridge. There are well-developed canals along the lateral margins of the pterotics and frontals, with two large foramina anteriorly, but there is no transverse canal across the frontals. SUSPENSORIUM AND JAWS (Fig. 1): As Osteology of Gorgasia punctata — Rosenblatt 93 might be expected from the short oblique mouth, the suspensorium is strongly inclined forward. The hyomandibular and quadrate are massive. The palatopterygoid is developed as a broad lamina, which is attached by a ligament to the vomer. The maxillary contacts the neuro- cranium at the tip of the snout. The posterior end of the maxillary is expanded, but the remainder is a narrow lamina. There is no pedicel anteriorly. opercular series (Fig. 1) : The well-devel- oped opercular bones are strongly ossified. The preopercle is triangular, like that of Taenio- conger, but unlike that of Nystactichthys as illustrated by Bohlke (1957). The blocky and subtriangular interopercle has a pronounced anterior extension. The crescentic subopercle curves upward under the lower angle of the opercle. The dorsal margin of the broadly crescentic opercle is deeply concave; its upper- rear corner is far above the upper end of the hyomandibular. hyoid arch (Fig. 2): This arch consists of the unpaired glossohyal and urohyal, and paired upper hypohyal, ceratohyal, and epihyal. The interhyal is absent. All the branchiostegal rays are inserted on the lateral surfaces of the arch, one on the ceratohyal and seven on the epihyal. In Nystactichthys and Taenioconger, in contrast, two branchiostegals are inserted on the cerato- hyal. The urohyal is needle-like, with an ex- panded and flattened anterior end. The dorsal surface of the glossohyal is grooved. EH, epihyal; GH, glossohyal; HH, hypohyal; UH, urohyal. shoulder girdle (Fig. 1): The cleithrum and supraclei thrum are well developed. The ex- panded head of the supracleithrum is bifurcate in the specimen illustrated, but not in another. The well-ossified hypercoracoid and hypocora- coid are connected by cartilage. The four hour- glass-shaped actinosts are small, but well ossi- fied. According to Bohlke Nystactichthys has no actinosts, but the species of Taenioconger that I have examined are like Gorgasia in this respect. VERTEBRAE AND ASSOCIATED BONES (Fig. 3) : In one specimen the vertebrae number 144, of which 45 precede the anus. Figure 3A repre- sents a cross section at the level of the 17th vertebra; Figure 3B illustrates the 17th to 19th vertebrae in lateral view. The vertebrae anterior to the dorsal origin bear well-developed, crest- like neural spines. The remainder of the ab- dominal vertebrae have large neural arches, but EN NA IM .1 mm. A Fig. 3. Vertebrae and associated bones of Gor- gasia punctata: A, Front view of 17th vertebra; B, side view of 17th to 19th vertebrae. CE, Centrum; EN, epicentral; IM, intermuscular; NA, neural arch; P, parapophysis ; PL, pleural rib. 94 no neural spines. The first 4 vertebrae bear strongly-developed, winglike transverse pro- cesses.These curve out and back, and each has on its posterior margin a long, thin, backward- directed process that seems to represent a fused epicentral (these structures lie on the same plane as the epicentrals associated with the more pos- terior vertebrae; furthermore, epicentrals are otherwise lacking on the first 4 vertebrae). The 5th through 9th vertebrae bear epicentrals, and weak transverse processes without backward prolongations. Bohlke mentioned no such pecu- liar condition in Nystactichthys halts , nor can I find transverse processes on the anterior verte- brae in Taenioconger digueti or T. hem. In T. hassi, however, weak transverse processes are developed on the first few vertebrae. In Gor- gasia the transverse processes are more weakly developed posterior to the 4th vertebra, and are not noticeable posterior to the 10th vertebra. The abdominal vertebrae bear strong para- pophyses to which, posterior to the 6th vertebra, are articulated strong pleural ribs. There is a strong median vertical ridge on each parapo- physis. The first haemal spine appears 15 ver- tebrae behind the anal origin, and the pleural ribs are present to this point. The caudal vertebrae bear transverse pro- cesses, commencing 6 vertebrae behind the anal origin. The transverse processes regress toward the tail-tip and are no longer apparent on the 10th vertebra before the caudal. The neural arches of the caudal vertebrae are smooth until about the 60th postanal vertebra, which bears the first neural spine, in the form of a small projection. The neural spines persist as low conical projections until 17 vertebrae from the tail-tip, behind which they become increasingly higher and more bladelike until they assume the shape shown in Figure 3. The haemal spines are also small and inconspicuous anterior to the 17th vertebra from the tail-tip. Thereafter, like the neural spines, they become increasingly higher and more bladelike. Shortly before the tail-tip the haemal spines become divided, so that the haemal arches are again open, as on the precaudal vertebrae (Fig. 4). Epicentrals are associated with all vertebrae except the last 10. Epipleurals appear 6 verte- brae behind the anus and persist until 15 verte- brae before the tail-tip. PACIFIC SCIENCE, Vol. XXI, January 1967 Fig. 4. Last 3 vertebrae of Gorgasia punctata, lateral view. AR, Anal ray; CE, centrum; BAS, basal element of pterygiophore ; CR, caudal ray; DR, dorsal ray; HY, hypural; NA, neural arch; RAD, radial element of pterygiophore. Dorsal and ventral intramuscular bones are well developed, and associated with all but the first 12 and last 4 vertebrae. Most of the intra- musculars are simple ; one of those illustrated in Figure 3 happens to be bifurcate. CAUDAL AND ASSOCIATED STRUCTURES (Fig. 4) : The tail-tip of Gorgasia is hard and pointed, with the fin rays concealed. However, the caudal skeleton is well developed and com- plex. As mentioned above, the neural and haemal arches and associated spines become expanded posteriorly, and the haemal and, to a lesser extent, the neural arches become open. According to the terminology of Nybelin (1963), there is but one ural centrum. Fused to it is a large hypural plate, probably consisting of several fused hypurals. The structure labeled HY? in Figure 4 is somewhat problematical. It has a basal-less fin ray associated with it and thus might be termed a hypural, but from its posi- tion it is difficult to determine whether it is itself associated with the last preural centrum or the ural centrum. Likewise, the nature of the dorsal element labeled NA is somewhat ambig- uous. It might be termed an epural, but since it is fused to the centrum, and divided anteriorly like the preceding neural arch, I term it a neural arch, despite the circumstance that it bears two "prin- Osteology of Gorgasia punctata — Rosenblatt opal” caudal rays. A strong process curves forward and downward from the neural arch element. It probably serves as a muscle attach- ment and may be associated with tail-first dig- ging in sand. An essentially similar caudal skeleton has been figured by Bohlke for Nystac- tichthys halts, and I have found the caudal skeleton of 7 aenio conger digueti, T. herrei, T. hassi, and T. n.sp. to be basically the same. CIRCUMORBITALS AND LATERAL LINE CA- NALS: The circumorbital ring is complete, con- sisting of at least five weakly-ossified and roof- less bones. Apparently the small "supraorbital” illustrated by Asano (1962) for several Japa- nese congrids is absent. The temporal canal is encased in bone in G. punctata only, among the species examined. The lateral-line canal along the body is contained in a series of ossi- cles (lateral-line scales?). Anteriorly these are developed as unconnected but closely opposed tubes, which posteriorly gradually become less strongly ossified, so that along the midbody there is an open trough consisting of a series of ossified half-rings. In the species of Taenio- conger examined, the lateral-line ossicles are developed as short, widely-spaced, troughlike ossifications. RELATIONSHIPS OF THE HETEROCONGRINAE The heterocongrines resemble the Ophichthi- dae in several respects. In both groups the ribs are laminar, and the neural spines reduced (vestigial in the ophichthids) , as are the cir- cumorbitals. In Gorgasia and in some species of Taenio conger the caudal fin is short and the tail-tip fleshy. In all, the body is elongate and circular in cross section. The pectoral fin, as in many ophichthids, is reduced (varying from small in Gorgasia to minute in Nystactichthys and Taenioconger to absent in Heteroconger) . In addition to these structural characters, both groups share the sand-dwelling habit. However, the characters given by Gosline (1951) to separate the Congridae and the Ophichthidae (except that it is now known that many con- grids have an auditory bulla) serve to distin- guish Gorgasia and its allies from the ophich- thids as well. In addition, it may be noted that the Congridae have the parapophyses divided by a vertical ridge and have a vomerine process, 95 to which the palatopterygoid is ligamentously attached — features that appear to be lacking in the Ophichthidae. The superficial similarities between Gorgasia and the ophichthids are certainly parallel adap- tations to a similar mode of life, and the minor osteological similarities may be adaptations as well. Although recognizing the close relationship between the two groups, Klausewitz and Eibl- Eibesfeldt (1959) maintained the family Het- erocongridae as distinct from the Congridae. Their action was based on behavioral differ- ences and on bone reduction and "Fensterbil- dung” (fenestration?) in the skeleton. How- ever, their illustration of the head skeleton of Xarifania h. hassi shows a well-developed skull and well-integrated jaws, suspensorium, and opercular series. On the basis of the present investigation there are no grounds (except that the circumorbital series is less well developed) for the contention that the skeleton has under- gone important reduction in comparison with that of the Congridae. There is, in fact, nothing in the osteology of the heterocongrines I have examined that would preclude the inclusion of the group in the Con- gridae. The heterocongrines possess the basic congrid characters of ankylosed frontals, for- ward-inclined suspensorium, few and nonover- lapping branchiostegals, maxillary-ethmoid ar- ticulation near tip of snout, caudal vertebrae with transverse processes, skull truncate poste- riorly, parapophyses divided by a vertical ridge, and a lateral process on the vomer. The chief osteological differences are: neural spines absent on most abdominal vertebrae (and on most caudal vertebrae as well in Gorgasia) ; neural and haemal arches becoming high and bladelike near the tail-tip ; urostylar vertebra better developed, and supporting structure of caudal more complex; epineurals and epipleu- rals lost 10-15 vertebrae before tail-tip; cir- cumorbital series reduced and less ossified. Stronger divergences from the basic congrid type are found in nonosteological characters. In most of the Congridae the muzzle is elon- gate, and the olfactory organ is well developed, with numerous lamellae. In the heterocongrines the muzzle is short, the eye is relatively large, and the olfactory rosette is much smaller, with 96 few lamellae (ca. 20). This distinction is no doubt correlated with a change in food habits (Klausewitz and Eibl-Eibesfeldt, 1959). Also the habit of living colonially in sand tubes is unknown in other congrids. These differences and similarities seem to bear out Bdhlke’s (1957) contention that the Heterocongrinae should be regarded as a well- defined subfamily within the Congridae, That there are profound differences in behavior and ecology is undoubted, but these have not in- volved any fundamental changes in the basic congrid body plan. Until recently, little information has been available on the osteology of the family Con- gridae. However, Asano (1962) has presented detailed information on the anatomy of 10 genera and 14 species of Japanese congrids. On the basis of his study, Asano recognized two subfamilies, the Anagoinae and the Con- grinae (the Heterocongrinae were not consid- ered). The Anagoinae were said to differ from the Congrinae in that there is a forward and laterally directed process on the ethmoid, the supraoccipital is absent, there are only four suborbitals, the abdominal and caudal vertebrae are about equal in number, the gas bladder is attached to the parapophyses, the tail-tip is hard, the caudal rays are short, the fin rays are un- segmented, and the lateral-line scales are well developed. Asano assigned two genera, Anago and Allo- conger , to the Anagoinae. I can confirm that Arias oma belongs here, as does the recently described Paraconger Kanazawa 1961. The heterocongrines share characters with both the Anagoinae and the Congrinae. They agree with the congrines in that the supraoc- cipital is present, there are many more caudal than abdominal vertebrae, and the gas bladder is free from the parapophyses. They agree with the anagoines in that the fin rays are unseg- mented, the caudal is reduced, and the lateral- line scales are well ossified (corresponding to Asano’s " Anago type”). I have been unable to determine with certainty the number of sub- orbitals in the heterocongrines. Gorgasia alone agrees with the Anagoinae in having a lateral ethmoid process. In this connection it is important to establish the evo- lutionary position of Gorgasia. Bohlke (1957) PACIFIC SCIENCE, Vol. XXI, January 1967 gave reasons for considering Gorgasia to be in most respects the most primitive of the Hetero- ! congrinae. His conclusions are borne out in this study, except for the discovery in Gorgasia of ! peculiar, expanded transverse processes on the anterior vertebrae, and the loss of an anterior maxillary pedicel. These specializations proba- bly preclude Gorgasia as an ancestor, but they do not militate against the hypothesis that j Gorgasia is more generalized over-all, and prob- \ ably was an earlier offshoot of the heterocon- [ grine line. The retention in Gorgasia of a lateral ethmoid process indicates relationship with the anagoine line. It seems unlikely that the agreement rep- resents convergence. Eels have evolved a num- :: ber of structures bracing the maxillary, cor- j related with elongation of the gape and with the use of the jaws in biting and crashing (Gos- I line, 1951; Asano, 1962). However, the trend 1 in heterocongrine evolution has been in the i! other direction, toward shortening of the gape ft and development of a jaw structure and denti- tion suitable for snapping at planktonic prey. It may be that the retention of the lateral eth- moid process in Gorgasia has allowed the loss i; of the maxillary pedicel. It seems plausible to hypothesize that the Heterocongrinae and Anagoinae arose from a common ancestor which had a lateral ethmoid 1 process, a supraoccipital, unsegmented fin rays, ;j and well- developed lateral-line scales. It seems i! likely that the sand-burrowing habit (known for Anago') had already been developed. The two groups have diverged sharply, however. The development of the plankton-feeding habit in the' heterocongrines has been accompanied by important changes in the head. The mouth has become short and oblique, and the denti- j tion specialized. The development of a short oblique mouth as an adaptation to snapping at 1 plankton or small prey has taken place in a number of fishes. Compare, for example, the serranid genus Epinephelus , which feeds on relatively large prey, with the plankton-feeding Paranthias. A similar phenomenon can be seen if the bottom-feeding embiotocid genus Micro- |j metrus is compared with the closely related genus Brachyistius , which feeds in midwater !i (Hubbs and Hubbs, 1954). Walter A. Starck II has pointed out to me that the shortening of the J Osteology of Gorgasia punctata — Rosenblatt 97 muzzle in these fishes results in the placement of the eye close to the tip of the snout, and thus allows for close-up binocular vision. Thus, vision has become more important in prey find- ing in heterocongrines, and the eye is much enlarged and the olfactory organ much re- duced. The lateral-line system on the head has likewise become reduced, again probably cor- related with the increased dependence on vision. The great elongation of the slender body would seem to be an adaptation to getting the head well off the bottom, and yet maintaining con- tact with the sand tube which is used for cover. (The normal posture of a heterocongrine is vertical, with the anterior one-half to two-thirds of the body out of the sand tube.) On the other hand, the anagoines, except in the loss of the supraoccipital, have diverged much less from the basic congrid type, either in structure or in behavior. REFERENCES Asano, H. 1962. Studies on the congrid eels of Japan. Bull. Misaki Mar. Biol. Inst. 1:1- 143. Bohlke, J. 1951. A new eel of the genus Tae- nioconger from the Philippines. Copeia 1951 (1) : 32 — 35. 1957. On the occurrence of garden eels in the western Atlantic, with a synopsis of the Heterocongrinae. Proc. Acad. Nat. Sci. Philadelphia 109:59-79. Gosline, W. A. 1951. The osteology and clas- sification of the ophichthid eels of the Ha- waiian Islands. Pacific Sci. 5 (4) : 298-320. 1952. Notes on the systematic status of four eel families. J. Washington Acad. Sci. 42(4) : 130— 135. Hubbs, C. L., and L. C. Hubbs. 1954. Data on the life history, variation, ecology and rela- tionships of the Kelpperch, Brachyistius fre- natus, an embiotocid fish of the Californias. Calif. Fish and Game 40(2) : 183-198. Kanazawa, R. 1961. Bar aeon ger, a new genus with three new species of eels (family Con- gridae). Proc. U. S. Natl. Mus. 113(3450): 1-14. Klausewitz, W., and I. Eibl-Eibesfeldt. 1959. Neue Rohrenaale von den Maldiven und Nikobaren (Pisces, Apodes, Heterocon- gridae). Senck. Biol. 40(3/4) : 135-153. Meek, S. E., and S. F. Hildebrand. 1923. The marine fishes of Panama, Pt. I. Field Mus. Nat. Hist. Zool. Ser. 15. 330 pp. Nybelin, O. 1963. Zur Morphologie und Ter- minologie des Schwanzskelettes der Actinop- terygier. Ark. Zool. (2) 15 (35) :485-5l6. The Flora of Romonum Island, Truk Lagoon, Caroline Islands Benjamin C Stone1 Romonum Island (7° 25' N, 151° 40' W) is one of the smaller central islands in Truk, a large island complex comprising several peaks of volcanic origin within a large atoll-like reef. Hence sometimes Truk is called an "almost- atoll,” because it is in a transitional stage be- tween younger islands, such as Ponape or Kusaie, and older atolls, such as those of the Marshall Islands group. Romonum itself is rela- tively small and low, nearly a mile in length and half a mile in width, with a rounded hill at the eastern end rising to a height of 167 ft, with fairly steep sides on the east and northeast, and flat or gently sloping land to the west and south. Two extensive swamps occur, one toward the western end and another larger one toward the eastern end, both on the south side of the island. A sandy beach occurs along the south- western tip and at several other localities on the western and southern coasts, while ramparts of black basalt boulders occur at several localities around the perimeter, especially on the east end. The island is situated slightly northeast about 4 miles from Tol Island (Truk’s largest and highest island), and about 2.9 miles due north of Fala-beguets I. (using the name shown on the 1944 edition Hydrographic Office map), and about 2.4 miles slightly northwest of Udot I. Moen Island, location of the U.S. Trust Territory Truk District Headquarters, is nearly 12 miles to the northeast. As is true of virtually all of the islands within the encircling reef (excluding the coral- line reef islets), Romonum is of volcanic origin. Except for the well-developed sandy beach, the island is composed of black basalt; no high raised limestones are found here or anywhere in Truk (although a few terraces scarcely a meter high do occur). The geological history of Truk is complex: the islands are much sunken or eroded; there are drowned valleys, wave-cut terraces (at about 40 m alt. and again 1 Department of Botany, University of Malaya, Kuala Lumpur. Manuscript received January 3, 1966. at 100 m alt.), and other evidences of both subsidence and emersion. However, little of this is in sight in Romonum. For a fuller geo- logical account, publications by Tayama (1940), Hess (1946), Bridge (1948), and, for a brief description, Gressitt (1954) may be consulted. In January 1965, I was enabled to visit both Truk and Ponape (as well as Saipan and Rota) through the generosity of the Trust Territory Government. At that time Prof. Ward Goode- nough of the Department of Anthropology, University of Pennsylvania, was engaged in a lengthy restudy of the people of Romonum Island, and he invited me to stay for a time there. This invitation led to a sojourn of several days, from January 28 to 31. During this time a collection of plants was made, and most parts of the little island were visited, with the help of Oliver Goodenough as guide. Dr. Goode- nough has allowed me to make use of his map, to which I have added some indications of the vegetation (Fig. 1). He has also provided his critical ear, a knowledge of Trukese dialects, and the orthography for most of the plant names given herein. Most names were verified by Dr. Goodenough; other names are in the form shown in P. J. R. Hill’s mimeographed list of Trukese plant names, or are approxi- mations in my own spelling. ACKNOWLEDGMENTS I am grateful to the Department of Agricul- ture, Trust Territory Government, Saipan, for the opportunity to visit Truk and several other islands in December 1964 and January-Febru- ary 1965 ; and particularly to Mr. Manuel Sproat, Director of Agriculture, for his con- tinued encouragement, assistance, and hospi- tality. I also must thank several District Agri- cultural Officers, both in Truk and Ponape, i especially Mr. Leonard Aguigui in Truk, and Mr. Ed. Pavao, Mr. J. D. Zaiger, and Mr. Kesner Hadley in Ponape, for their help. Peter J. R. Hill, Educational Administrator for Truk, 98 Flora of Romonum Island — Stone 99 WI o Fig. 1. Map of Romonum Island (courtesy of W. H. Goodenough). Characteristic plant species are indicated. Distances, outline, and details should be considered as approximations. 100 PACIFIC SCIENCE, Vol. XXI, January 1967 Fig. 2. Prof. W. H. Goodenough, during his lengthy visit on Romonum. was helpful in many ways. Prof. W. H. Goode- nough and Mrs. Goodenough provided house and sustenance on Romonum, and Oliver Goodenough acted as guide. Finally, I must thank the College of Guam and especially Dr. A. C. Yamashita for the opportunity and sup- port provided for the work undertaken on this trip ; and thanks are also due to the many other friends who were of assistance. THE FLORA OF ROMONUM ISLAND The following key and species list is un- doubtedly not complete, but it includes the most common species of vascular plants on Romonum. There are probably additional spe- cies on the northeastern part of the island, which I did not cover thoroughly ; and no doubt some weeds and cultivated plants were missed, or will be introduced in the future; still, the species cited here represent, I believe, the bulk of the island’s vegetation. A key to species is provided, but of course it can only account for the species listed, and additional discoveries will have to be added. A number of plants are absent or represented very sparsely; Messer- schmidia , for example, common on atolls and limestone areas of other islands, was not found ; Pemphis was seen only once, and as an isolated individual; Ipomoea pes-caprae was not found, but may well occur on the north beach (its com- mon associates, such as Canavalia maritime i, W edelia bi flora, and Scaevola were found) ; Polys cias grandifolia, Soulamea amara, Suriana, Pisonia grandis, and other representative atoll species, found elsewhere in Truk, did not ap- pear. In this respect Romonum resembles Yanagi Islet, only a few miles away in Truk Lagoon, the vegetation of which was described a few years ago (Hill and Stone, 1961), and which interestingly is also deficient in species typically associated with coral atolls. KEY TO SPECIES Only the vascular plants are accounted for here ; for bryophytes the reader is referred to Miller, Whittier, and Bonner (1963) ; for marine algae to Okamura (1915) and Taylor (1950) ; for other groups there is no comprehensive treatment for Truk, but papers on lichens and fungi have been written by Imazeki (1941), Kobayasi (1939), Jatta (1903), and Sydow and Sydow (1921). I. Flowerless plants bearing spores in sporangia, these usually borne on the backs or edges of fronds (Ferns) 1. Fronds simple, to several feet long, forming a large nest or rosette; epiphytes; sori (groups of sporangia) oblique, linear Asplenium nidus 1. Fronds lobed or divided 2. Sporangia covering the entire lower surface of the leaflets; fronds to several feet long, leathery, somewhat dimor- phic; 1-pinnate; swamp ferns Acrostichum aureum 2. Sporangia in groups (sori) 3. Fronds deeply parted, the lobes entire; sori sunken in pits, these evident on the upper surface as bumps; ter- restrial or epiphytic ferns with creeping rhizomes Phymatodes scolopendria 3. Fronds pinnate or bi- or tripinnate 4. Fronds 1-pinnate . 5. Sori kidney-shaped; lobes of pinnae crenate Nepbrolepis Flora of Romonum Island — Stone 101 5. Sori round; lobes of pinnae acute Cyclosorus 4. Fronds bi- or tripinnate, deltoid in outline Davallia II. Flowering plants 1. Monocotyledons (seed with 1 cotyledon; leaves usually with parallel venation; root system fibrous or without a tap- root; woody species with numerous discrete fibres traversing a softer tissue; flower parts frequently in 3’s or multiples of 3) 2. Trees or large woody shrubs 3. Palms; trunks large, unbranched, erect, or in Nypa submerged, + horizontal, with terminal crown of large pinnate leaves 4. Stems horizontal, submerged in swamps Nypa 4. Stems erect 5. Fruits ovoid, smooth, large (to 12 inches diameter or more), edible (coconuts) Cocos nucifera 5. Fruits smaller, subglobose, covered closely by glossy brown overlapping scales, inedible .... Metroxylon 3. Trunks branching, leaves not divided 6. Leaves elongate, toothed; fruit a large head; trunks usually with basal proproots Pandanus 6. Leaves not toothed; fruit a small berry; trunks without proproots Cordyline 2. Herbs, vines, or creepers; including grasslike plants (these sometimes with tall, moderately rigid, canelike stems) 7. Plants strictly aquatic, submerged or floating 8. In salt water only; marine plants rooted in sand or sandy debris in quiet lagoon waters 9. Leaves up to 3 ft long or more, mostly 12-17 mm wide; rooted portion with black persistent fibres Enhalus 9. Leaves seldom as much as 1 ft long, 4-10 mm wide; rooted portion lacking black fibres .... Thalassia 8. In fresh water only, floating or loosely rooting in mud Blyxa octandra 7. Plants terrestrial, rooted in soil 10. Grasses and grasslike plants (grasses, sedges, reeds), i.e., usually with sublinear leaves with parallel veins, but leaves sometimes reduced or absent; small green or brown flowers; tufts of fibrous roots 11. Stems triangular in cross section; inflorescence surrounded by leaflike bracts 12. Inflorescence a buttonlike head Cyperus kyllingia 12. Inflorescence with evident branching 13. Coarse plants to 5 ft tall Cyperus odor at us 13. Slender herbs to 2-3 ft Cyperus javanicus 11. Stems round or nearly so in cross section; inflorescence without or with rather inconspicuous bracts 14. Stems solid; rosette plants 15. Leafy plants with branched inflorescences 16. Seeds brown Eimbristyl'ts 16. Seeds white Scleria 15. Leafless plants writh apical inflorescence on scapes Eleocharis 14. Stems hollow, nodose; rosette or creeping plants 17. Tall (to 9-12 ft) with canelike stems, large plumose inflorescences of minute slender spikelets 18. Inflorescence green or brown; swamp reeds Trichoon karka 18. Inflorescence white or silvery (sugarcane of gardens) Saccharum officinarum 17. Not tall reeds 19- Fruit a spiny burr Cenchrus echinatus 19. Fruit not a spiny burr 20. Inflorescence narrow cylindric, breaking at joints when mature Lepturus repens 20. Inflorescence not jointed and disarticulating 21. Spikes digitate, 2 or more borne umbellately 22. Spikelets awned 23. Spikes 3-5; spikelets of several florets each, in several rows on lower side of rachis Dactyloctenium 23. Spikes usually 2; spikelets paired, 1 stalked, 1 sessile Ischaemum 22. Spikelets not awned 24. Spikelets each with 1 floret 25. Spikes 2, conjugate; fruit indurate, broader or broad as long . . Paspalum 25. Spikes usually 3 or more; fruit cartilaginous; longer than broad Digitaria 24. Spikelets each with several florets Eleusine indica 21. Spikes not digitate 26. Spikelets all on one side of rachis 27. Spikes several, distant; grass of shady forest Oplismenus 27. Spikes 1 or 2, close, hidden; beach grass Thuarea 26. Spikelets on both sides of rachis 28. Leaves elliptic, 1/^-1 inch wide; panicle large, pale, complex Centotheca lappacea 28. Leaves narrow, less than *4 inch wide 29. Spikelets minute, in a diffuse compound panicle; no awns .... Eragrostis 29. Spikelets in a stiff reddish few-branched panicle; awns present Chrysopogon 10. Not grasses or grasslike plants; herbs, often very large or giant, as in bananas; or vines 30. Leaves with reticulate (network) venation 102 PACIFIC SCIENCE, Vol. XXI, January 1967 31. Vines Dio scored 31. Not vines 32. leaves palmately then pinnately divided; flowers on tall leafless scapes, in clusters bearing also long threadlike pendent filaments; tuberous Tacca 32. Leaves not divided; leaves heart-shaped, with rounded or pointed (then arrow-shaped) lobes 33. Large herbs with pointed (arrow-shaped) lobes Cyrtosperma 33. Small or large herbs with rounded lobes 34. Smaller plants, rarely over 3 or 4 ft tall, leaves pale or glaucous Colocasia 3 4. Larger plants with dark leaves Alocasia 30. Leaves with parallel veins 36. Terrestrial plants 37. Giant herbs with oblong leaves 3-6 ft long, often 1 ft broad, later splitting into segments to the midrib; parallel veins perpendicular to midrib (bananas) Musa 37. Smaller herbs with parallel longitudinal veins 38. Tall herbs with leaves arranged alternately along the erect stems Alpin'ta 38. Leaves basal from a short usually underground corm or bulb 39. Leaves stiff and pointed, spiny; compound fruit with a crown of leaflike bracts . . Ananas 39. Leaves otherwise; fruit not as above 40. Large herbs (leaves to several ft long); flowers white 41. Flowers with a corona Hymenocallis 41. Flowers without corona Crinum 40. Small herbs (leaves seldom to 1 ft long) with pink flowers Zephyranthes rosea 36. Epiphytes with short, somewhat leathery leaves; alternating on the stem Dendrobium sp. 1. Dicotyledons (seed with 2 cotyledons; leaves usually with reticulate venation; taproot often present; woody species often with "solid,” cambium-formed, annularly incremented wood; flowers often in 4’s or 5’s or multiples thereof) 42. Leafless, often orange-stemmed parasitic vines Cassytha 42. Not as above 43. Leaves compound, divided into fully distinct leaflets 44. Leaves with 3 leaflets (in Derris trifoliata leaves of both 3 and 5 leaflets found) 45. Trees or shrubs 46. Leaflets with slightly toothed edges; fruit a round berry Allop by lus 46. Leaflets entire, smooth edged; fruit a short flat segmented pod Desmodtum 45. Vines or herbs 47. Flowers yellow Vigna marina 47. Flowers pink or white 48. Flowers rosy pink, the banner petal with a white splotch; pod somewhat inflated; leaflets al- ways 3, broadly ovate Canavalia maritima 48. Flowers pale pink to white; pod very flat; leaflets sometimes 3, sometimes 5 . . Derris trifoliata 44. Leaves with more than 3 leaflets 49. Erect shrubs, or herbs 50. Leaves 1-pinnate 51. Flowers papilionate, 2 petals joined to form a keel, 2 petals as lateral wings, 1 as a banner; fruit an inflated pod Crotalaria 51. Flowers mimosoid, not as above; fruit thick but not inflated Cassia 50. Leaves bi- or tripinnate Polyscias fruticosa 49. Vines or creepers 52. Leaflets usually 5 or 7 per leaf, more than 1 inch long; plants seldom climbing . . Derris elliptica 52. Leaflets more numerous but smaller; high climbers; seed small, red, with black spot Abrus precatorius 43. Leaves simple or merely lobed or parted, not divided into distinct leaflets 53. Leaves markedly lobed (not merely toothed) 54. Trees, giant softwooded herbs, or herbs, with thick milky latex 55. Herbs; upper leaves with basal red patches Euphorbia heterophylla 55. Trees or tree-like herbs 56. Leaves pinnately lobed (lobing quite variable, some trees with nearly entire leaves) Artocarpus 56. Leaves palmately parted Carica papaya 54. Vines or herbs; sap not milky 57. Climbing vines 58. Leaves mostly 3-lobed; flowers large, rotate, with an ornately laciniate calyx, not tubular Pas si flora foetid a 58. Leaves 3- or mostly 5-parted; flowers tubular, trumpetlike; calyx not laciniate Ipomoea digitata 57. Herbs or prostrate creepers (sometimes slightly woody) 59. Prostrate creepers; fruit a burr Triumfetta procumbens 59. Erect (somewhat woody) herbs 60. Fruit a burr; leaves mostly 3-lobed Triumfetta semitriloba 60. Fruit a hairy capsule; leaves mostly 5-lobed Abelmoschus moschatus 53. Leaves not at all lobed, sometimes toothed 61. Trees or large notably woody shrubs 62. Sap milky white, or noticeably yellowish latex 63. Sap yellowish; leaves with numerous curved parallel lateral veins; fruit a hard woody sphere of golfball size Calophyllum inophyllum I! Flora of Romonum Island — Stone 103 63. Sap milky; leaves not as above 64. Old leaves turning red just before falling; fruit a small 3-celled capsule; latex poison- ous Excoecaria agallocha 64. Old leaves usually turning yellow; fruit a small "fig”; not poisonous 65. Small, somewhat shrubby tree; leaves usually asymmetric at base; dioecious; figs orange Ficus tinctoria 65. Large trees with aerial roots; leaves symmetric; figs pink Ficus virens 62. Sap clear, watery 66. Mangrove trees with prominent aerial proproots or ascending breather-roots 67. Leaves spirally arranged, longer than broad; stipules present 68. Flowers with scarlet corolla, tubular; leaves narrowly obovate, often notched at tip; seed germinating after falling Lumnitzera lift or ea 68. Flowers with inconspicuous white or orange petals (but calyx may be deep red); seed germinating on tree, radicle growing to a length of a foot or more before falling 69. Calyx of 7-14 narrow lobes, usually red (rarely white) ; flowers shortly stalked, pendent Bruguiera gymnorrhiza 69. Calyx of 4 short deltoid lobes, usually green 70. Inflorescence branched, of several flowers Rhizophora mucronata 70. Inflorescence short, of few flowers Rhizophora apiculata 61. Leaves opposite, nearly as broad as long or broader; stipules absent Sonneratia 66. Not mangrove trees as above 71. Stamens fused into a tube surrounding the style; corolla tubular, yellow, of 5 petals; hibiscus- like flowers 72. Leaves broadly cordate, grayish beneath, tip not much drawn out Hibiscus tiliaceus 72. Leaves narrowly cordate, green, tip long drawn out Thespesia populnea 71. Stamens not fused as above 73. Leaves pale beneath, covered closely by minute peltate scales; fruit a keeled, woody, boat- like structure Heritiera litoralis 73. Leaves not as above 74. Twigs thorny; foliage with odor of lime Citrus 14. Not thorny; not with lime odor 75. Leaves alternate or spiralled 76. Leaves distichous (alternating in 1 plane) 77. Flowers bisexual 78. Leaves coarsely toothed; flowers yellow; a scrambling shrub Colubrina asiatica 78. Leaves entire; trees 79. Fruit muricate (with soft blunt thorns), edible ... Annona muricata 79. Fruit not muricate Cananga odorata 77. Flowers unisexual Glochidion 76. Leaves not distichously arranged 80. Leaves peltate or nearly so 81. Flowers bisexual; leaf entire with red spot at junction of petiole; fruit a black berry set inside a lantern-like calyx Hernandia 81. Flowers unisexual; leaf with 3 large teeth and many small ones; without red spot; fruit a small capsule Macaranga carolinensis 80. Leaves not at all peltate 82. Leaves concave, saucerlike Polyscias Scutellaria 82. Leaves flat 83. Flowers unisexual; leaves coarsely toothed Acalypha 83. Flowers bisexual; leaves entire 84. Corolla orange Cordia 84. Corolla white, greenish, pink, or cream 85. Leaves mostly 1-2 inches long, crowded on stems, fleshy; flow- ers white with separate petals Pemphis 85. Leaves larger, mostly 4-16 inches long 86. Leaves elliptic, glabrous; fruit a mango . . Mangifera indica 86. Leaves obovate 87. Leaves pale, often softly hairy with indistinct veins; flow- ers in short cymes; berry white Scaevola 87. Leaves darker, with distinct veins; fruit not white 88. Flowers in narrow spikes, less than % inch wide; fruit a red drupe Terminalia 88. Flowers in long racemes or clusters; fruit a green or brownish, angular, boxlike structure 89. Flowers pink, in long pendent racemes; fruit about 3 inches long, with rounded angles; tree of fresh- water swamps Barringtonia racemosa 104 PACIFIC SCIENCE, Vol. XXI, January 1967 89. Flowers white or faintly pinkish, in large not pen- | dent inflorescences; fruit 4-5 inches long, boxlike with 4 (rarely 5) fairly sharp angles, seaside j tree Barringtonia asiatica ! 75. Leaves opposite 90. Stipules present; flowers white, regular I 91. Fruit compound, fleshy, whitish, lumpy; flowers 4 or 5 petaled . . Morinda 91. Fruit not compound; flowers 8-petaled Guettarda speciosa 90. Stipules absent; flowers whitish or pale lavender, very small, 2-lipped . . Premna 61. Not trees; herbs (erect or prostrate) or vines 92. Leaves with petioles much longer than the cordate blades; low herbs with inconspicuous umbels of small flowers Centella asiatica 92. Not as above 93. Sap milky 94. Small, more or less prostrate herbs 95. Hairy and purplish leaves Euphorbia hirta 95. Glabrous Euphorbia thymifolia 94. Erect, sometimes slightly woody herbs; all leaves pale green, entire . . Euphorbia chamissonis 93. Sap not milky 96. Leaves opposite 97. Stamens long, protruding from the corolla; woody climbing or scrambling vines Clerodendrom inerme 97. Stamens included; small, somewhat woody, shrubby herbs 98. Flowers purple or blue, borne on spikes Stachytarpheta 98. Flowers white or yellow, not in spikes 99. Garden herbs with very pungent minty odor Ocimum sanctum 99. Wild or weedy plants without strong odor 99a. Leaves entire; flowers white Hedyotis bi flora [ 99^. Leaves coarsely toothed, flowers yellow W edelia bi flora 96. Leaves alternate 100. Flowers yellow with clawed petals; erect herb of freshwater swamps . . . Ludwigia octovalvis 100. Flowers greenish-white, minute, petals not clawed; herb of dry ground Phyllanthus amarus TAXONOMIC CHECK LIST PTEROPSIDA Class filicinae (Ferns) Acrostichum aureum L. A giant fern of swamps, usually mingled with ^mangrove species; it may reach 10 or 12 ft in height. The fertile fronds are slightly smaller than the sterile, which may be 18 inches wide. Asplenium nidus L. "nnuk”2 The birds’ -nest fern. A large species, usually epiphytic, with long strap-shaped fronds form- ing a rosette; sporangia in oblique linear sori. Cy cl os or us goggilodus (Schkuhr) Link A fern of swamps (fresh-water) and taro patches. Sometimes called C. gongylodes. Davallia solida (Forst. ) Sw. ; "peceen attu” (5281 )3 2 The vernacular names given are in the orthography used by Prof. Goodenough, and fuller rules on pro- nunciation will be found in his works. It should be noted here, however, that c is equivalent to / as in just; and that doubled vowels indicate extension of the sound. 3 These numbers refer to the author’s collections. A common epiphyte with a long, scaly rhizome closely attached to trunks or branches, bearing broadly deltoid tripinnatifid fronds. Nephrolepis exaltata (L.) Schott; "amaare” (5275) Terrestrial, rarely epiphytic; fronds pinnate. Phymatodes scolopendria (Burm.) Ching; "wenniimey” (5273) Terrestrial or epiphytic; fronds deeply pin- nately parted Also called Micros ovum (" Micro - sorium”) scolopendria. Class angiospermae (Flowering Plants) Subclass MONOCOTYLEDONAE PANDANACEAE Pandanus carolinensis Martelli; "fach” HYDROCHARITACEAE Blyxa octandra (Roxb.) Planch, ex Thw. The flowers of this aquatic plant are borne at the end of narrow scapes and are minute. the originals of which are deposited in the College of Guam Herbarium; duplicates have been sent to the Bishop Museum, Honolulu, and to the U. S. National Herbarium. Flora of Romonum Island — Stone 105 Enhalus ac oroides (L.f.) Rich, ex Chatin In lagoons; more common than Thalassia . Thalassia hemprkhii (Ehrb.) Aschers. GRAMIMEAE PANICOIDEAE group Tribe andropogoneae Ischaemum muticum L; "fetinin wuumw” (5285) Chrysopogon aciculatus (Retz.) Trin. Saccharum officinarum L; sugarcane Tribe paniceae Cenchrus echinatus L; burgrass Digitaria prunens (Fisch. ex Trin.) Buse var. microbachne (Presl) Fosb. Oplismenus comp os it us (L.) Beauvois P asp alum orbiculare Forst. Thuarea involuta (Forst. f.) Roemer and Schultes POOIDEAE group Tribe festuceae Centotheca lappacea Desvaux.; "fetinin wum- wune” (5284) Tribe arundineae Tricboon karka (Retz.) Roth in Roem. (5289) Hitherto generally known as Phragmites karka (Retz.) Trin. ex Steudel. Unfortunately this name cannot be maintained; see Stone (1964). Tribe eragrosteae Eragrostis amabilis (L.) Wight and Arnott Dactyloctenium aegyptium (L.) Willd. Eleusine indica (L.) Gaertner Tribe leptureae Lepturus repens (Forst.) R. Brown CYPERACEAE Cyperus javanicus Houtt. Cyperus kyllingia Endl. Cyperus odoratus L. (5305) Cyperus sp. Eleocharis geniculata (L.) Roemer and Schultes (5300) Fimbristylis cymosa R. Br. (5314) Scleria sp. (5306) PALMAE Cocos nucifera L. ; coconut palm Metroxylon amicarum (Wendland) Beccari; ivory-nut palm Generally in standing water or wet locations in valleys. Nypa fruticans Wurmb. The nipa palm. Easily recognised by its trunk- less appearance in swamps. ARACEAE Alocasia macrorrhiza (L.) Schott ex Schott and Endlicher Colocasia esculenta (L.) Schott; taro Cyrtosperma chamissonis (Schott) Merrill BROMELIACEAE Ananas comosus (L.) Merrill; pineapple Occasionally in cultivation. AGAVACEAE Cordyline fruticosa (L.) Goepp. Hitherto known as Cordyline terminalis (L.) Kunth (see Stone, 1964). AMARYLLIDACEAE Crinum asiaticum L. ; spider-lily Hymenocallis littoralis (Jacq.) Salisb.; seaside- lily Zephyranthes rosea (Sprengel) Lindley TACCACEAE Tacca leontopetaloides (L.) O. Kuntze (5309) DIOSCOREACEAE Dio scorea bulbifera L. ? ; yam (5366) 106 MUSACEAE Musa balhisiana X acuminata (M. paradisiac a F.) ; banana ZINGIBERACEAE Alpinia purpurata (Vieill.) K. Schumann; red ginger ORCHIDACEAE Dendrobium sp. ; nikocopwcopw” (5282) Epiphytic. A native species. Subclass DICOTYLEDONAE PIPERACEAE Peperomia pellucida (L.) HBK. Piper sp., "enes” (5274) MORACEAE Artocarpus altilis (Park.) Fosb. ; breadfruit Ficus tinctoria Forst f. (5298, 5310) Ficus virens Ait.; "aaw” (5291) Hitherto known as F. carolinensis Warb. (see Corner, 1965). URTICACEAE Procris pedunculata (Forst. f.) Wedd.; "kimm- wit” (5292) POLYGONACEAE Polygonum minus var. procerum (Danser) Steward? (5303) [If this is the same as the Polygonum in Guam.] ANNONACEAE Annona muncata F. ; soursop Cananga odorata (Fam.) Hook. f. and Thom- son; ylangylang LAURACEAE Cassytha pliformis F. HERNANDIACEAE PACIFIC SCIENCE, Vol. XXI, January 1967 LEGUMINOSAE Abrus precatorius L.; prayerbead Cassia occidentalis L. ; coffee senna Canavalia maritima (Aublet) Thouars; seaside peavine, "ceecon” (5272) Crotalaria sp. ; "afanafan” (5283) Denis elliptica (Roxb.) Bentham; "wimp” (5276) Derris trifoliata Loureiro; "wunenipot” or "wupenipot” Desmodium umbellatum (L.) DC. Vigna marina (Burm.) Merr. RUTACEAE Citrus auranti folia (Christm.) Swingle; lime, "nayimis” (5277) EUPHORBIACEAE Acalypha trukensis Pax and Hoffman; "mon- now” (5270) An endemic small tree, fairly common throughout Truk. Euphorbia chamissonis (Klotszch and Garcke) Boissier (5313) Euphorbia heterophylla F. Euphorbia hirta L. Excoecaria agallocha L. (5294) The sap of this tree is reputedly dangerous, especially to the eyes. It may be recognized by its prevalence in or near mangrove swamps or | rocky seaside locales, the tendency for the ma- ture leaves to turn red before falling, and the ! small catkins of flowers. The sap is notably ! milky. Glochidion ramiflorum Forst.?; "afor” or "ofor” (5365) Macaranga carolinensis Yolkens; "tuupw” or "kuruwen” (5271) Endemic in the Caroline Islands. Phyllanthus amarus Schum. and Thonn. ANACARDIACEAE Man gif era indie a L. ; mango SAPINDACEAE Allop hylus timorensis (Bl.) DC. (5311) Hernandia sonora L. Flora of Romonum Island — Stone 107 RHAMNACEAE Colubrina asiatica (L.) Brongniart TILIACEAE Triumfetta procumbens Forst. f. (5308) Triumfetta semitriloba Jacq. ? ; "sacawer” (5280) MALVACEAE Abelmoschus moschatus (L.) Medik; "nikono- koon” (5288) Hibiscus tiliaceus L. (5290) Malvastrum coromandelianum (L.) Garcke; "siyoyinen” (5286) Thespesia populnea (L.) Solander ex Correa STERCULIACEAE Heritiera littoralis Dry. (5299) Copiously fruiting, Jan. 28, 1965. GUTTIFERAE Calophyllum inophyllum L. ; kamani or Alexan- drian laurel CARICACEAE Carica papaya L. Only a few seen. PASSIFLORACEAE Passi flora foetida L. var. hispida (DC.) Killip; "pwompworn” (5279) LYTHRACEAE Pemphis acidula Forst. Apparently rare on Romonum ; only one indi- vidual seen. RHIZOPHORACEAE Bruguiera gymnorrhiza (L.) Lam. Also called B. conjugata (L.) Merrill. The usual form has the calyx scarlet; a form with a pure white calyx was described from Namo- nuito (Stone, 1959). This requires a nomencla- tural adjustment, as follows: Bruguiera gymnorrhiza forma alba B.C. Stone, comb. nov. B. conjugata (L.) Merr. forma alba B.C. Stone, Pacific Sci. 13:102 (1959). Type: Namonuito, Pisarach Islet, 2 July 1957, Stone 2144 (bishop museum). N.v. "ong.” Rhizophora apiculata Bl. ; "ciyaan iimw” Rhizophora mucronata Lam.; ''ciyaan wuumw” (5297) Since Rhizophora stylos a Griff, has been reported from Guam ( Stone 4437, GUAM and leiden) by Ding Hou (in litt. ) , it may turn up in other parts of Micronesia also. The three species are distinguished in Flora Malesiana (Ser. I, vol. 5, part 4, p. 450), 1958, as follows : 1. Inflorescences 2 -flowered, shorter than the petiole, in the axils of leaf scars of last year’s or last season’s growth; bracteoles at the base of the flower completely connate; petals glabrous R. apiculata 1. Inflorescences 2— 16-flowered, longer than the petiole, in the axils of current year’s or season’s growth; bracteoles connate only at base; petals hairy 2. Style obscure or very short, to 1.5 mm long R. mucronata 2. Style filiform, 4-6 mm long . R. stylo sa Since only the first two species are recorded in Kanehira’s check list of the Micronesian flora (1935), it will be of interest to see if the occur- rence of R. stylosa elsewhere in Micronesia can be established. SONNERATIACEAE Sonneratia caseolaris (L.) Engler (5295) COMBRETACEAE Lumnitzera littorea (Jack) Voigt (5293) Terminalia samoensis Rechinger LECYTHIDACEAE Banin gtonia asiatica (L.) S. Kurz Barringtonia racemosa (L.) Blume (5302) 108 PACIFIC SCIENCE, VoL XXI, January 1967 ! ONAGRACEAE Ludwigia octovalvis (Jacq.) Raven, Kew Bull. 15:476 (1962). (5304) Hitherto known as fussiaea suffruticosa L. ARALIACEAE Polyscias fruticosa (L.) Harms. Polyscias Scutellaria (Burm. f.) Fosb. Polyscias pinnata Forst. cultivar "tricochleata” UMBELLIFERAE Centella asiatica (L.) Urban CONVOLVULACEAE Ipomoea digitata L. (5367) Ipomoea indica (Burm. f.) Merrill The same as I. congesta R. Br. BORAGINACEAE Cordia subcordata Lam. (5312) VERBENACEAE Clerodendron inerme (L.) Gaertner Premma obtusifolia R. Br. (5307) Stachytarpheta jamaicensis (L.) Vahl; "sakura” (5287) The vernacular name, obviously Japanese (sakura = cherry), indicates the relative recency of this plant’s introduction. Informants placed the first appearance of the species in the 1920’s. LABIATAE Ocimum sanctum L. ; "wariig” (5278) Cultivated as an herb used with fish. The herbage is very rank. ACANTHACEAE Blechum Ibrownei RUBIACEAE Guettarda speciosa L. Hedyotis bi flora (L.) Lam. (5367 -a) Morinda citri folia L. GOODENIACEAE Scaevola taccada (Gaertn.) Roxb. Variously called S. frutescens or S. koenigii in older literature. Also known as S. sericea Vahl. COMPOSITAE Wedelia bifora (L.) DC. (5296) V ernonia cinerea (L.) Less. ECOLOGICAL NOTES ON THE VEGETATION OF ROMONUM Major Patterns of Vegetation Very little, if any, of the original vegetation is left intact. Instead the island presents a pic- ture of the long-existing interaction of man on the insular environment. Because of the small size and low elevation of the island, every square foot has probably had, from time to j time at least, the imprint of the human foot or the effects of the agricultural hand. In fact, throughout Truk, it is difficult to envision what the original lowland vegetation was like except in the areas which, because of their marginal nature — such as mangrove swamps, freshwater swamps, and sandy beach areas — have been considered useless or too difficult to change. Of course, in the more advanced areas (e.g., Moen) even these areas are now much altered through the use of modern techniques and machinery. On Romonum, however, we may look to these marginal areas for at least a partially persistent element of pre-human vegetation. Outside these marginal areas, Romonum con- sists largely of cultivated trees, usually rather well spaced, and consisting primarily of coconut palms and breadfruit trees. These two species are the only large trees in some localities, espe- cially in the immediate neighborhood of houses. On the hill in the eastern part, and toward the central part of the island, mango trees are also found in considerable numbers. Wherever "vil- lages” are located quite near the coast, there are small numbers — sometimes single individuals — of various arborescent species, especially Her- nandia sonora, Ficus virens, Metroxylon ami- carum, Calophyllum inophyllum, Hibiscus tiliaceus, or Thespesia populnea . Although there are exceptions, the mangrove trees — • Flora of Romonum Island — Stone 109 Fig. 3. View of the south coast of Romonum looking slightly eastward. (The pier of basalt rocks is that shown on the map just short of the western tip of the island.) The appearance is very characteristic, with the numerous coconut palms. The tree at the left is a Hernandia. Rhizophora, Bruguiera, Lumnitzera, Sonneratia, Excoecaria — and the littoral Heritiera are sel- dom found very near houses. To summarize, then, the major visual aspect of the forested portions of the island is the predominance of Cocos , Artocarpus, and scattered individuals of Man gif era, Ficus, and occasional other trees. Other than this fairly homogeneous and largely man-made "forest” type, several other major features are evident. These are the man- grove forest; the freshwater swamps; the sandy beach; and the basalt-boulder coast. Mangrove Formation Tree species: Rhizophora mucronata, R. apic- ulata, Bruguiera gymnorrhiza, Sonneratia case- olaris, Excoecaria agallocha, Nypa fruticans, Lumnitzera littorea. Other characteristic species: Acrostichum aureum. Marginal species: Hibiscus t iliac e us ; species of the freshwater swamps, which to some extent interpenetrate the mangrove area. In general, the Micronesian mangrove for- mations are not as rich in species, or so produc- tive in individuals, or so notable for large trees, as are the formations in the Malaysian or Carib- bean areas. In turn, the mangrove areas in Truk are rather less rich in species than those of Palau, farther west and consequently nearer the vast Philippine mangrove regions. Such species as Scyphiphora hydrophyllacea (Rubiaceae), Doli- chandrone spathacea (Bignoniaceae; occasional at margins of mangrove areas), and Ceriops candolleana (Rhizophoraceae), although found in Palau (and Yap) do not occur in Truk, 110 PACIFIC SCIENCE, Vol. XXI, January 1967 Fig. 4. The swamp fern, Acrostichum aureum, in the foreground; behind, a marginal zone of Trichoon karka, with intermixed coconut palms. Ponape, Kusaie, the Marianas, or the Marshall Islands. Many other mangrove formation plants, common enough in Malaysia and the Philip- pines, do not occur anywhere in Micronesia (e.g., Aegiceras, Kandelia, other species of Rhizophora) . Xylocarpus granatum occurs else- where in Truk, but I did not find it in Romo- num. Rhizophora stylosa Griff, has been found in Ponape ( Stone 1773) and in Guam ( Stone 4437) and probably occurs in Truk also. The mangrove formation of Romonum is on the whole rather poorly developed. The one extensive area is on the southern side of the island, from about the middle to within a few hundred yards of the east end. Small parts of the periphery have been converted to taro {Colo ca- st a) or Cyrtosperma plots. The larger central area of the swamp is composed of scattered and fairly small individuals of the various species, seldom over 10 ft high, interspersed with clumps of Acrostichum. Freshwater Swamps All of these are now to some extent planted with taro or Cyrtosperma , but they also include, to quite varied extent, various other species. Some swamps have been essentially cleared of vegetation, which has then been replaced by the cultivated aroids. Others, perhaps abandoned at various times of earlier cultivation, have be- come overrun with "weedy” species or with in- vading native elements. In time the freshwater swamp trees, Banin gtonia racemosa and Met- ro xy l on ami c arum, become a conspicuous fea- ture (Fig. 6). Also, the periphery of such swamps is constantly undergoing slight changes, depending on the adjacent area; advancing on Flora of Romonum Island — Stone 111 Fig. 5. Ivory-nut palms ( Metroxylon amicarum) in the west freshwater swamp. Below, left and right, dumps of Cyrtosperma chamissonis. Center, Trichoon karka. or retreating from them in accordance with such factors as rainfall, changes in tides, or man’s activities. Tree species: Banin gtonia racemosa, Metrox- ylon amicarum . Some other trees are also to be found in or at the edges of these swamps; they exhibit varying degrees of tolerance to standing water. Some, for example Glochidion, may en- dure the swamp conditions for an appreciable time, but succumb eventually, and meanwhile present an unhealthy appearance, the leaves be- ing few and often chlorotic. Typically, only the Barringtonia and Metroxylon are bona fide members of such swamp communities, and even the Metroxylon is not restricted to such com- munities but, for example, as on Tol Island, may be found in moist rocky valleys. Other characteristic species: the aroids, either actively cultivated or persisting from former cultivation ( Colocasia , Cyrtosperma , and, rarely, Alocasia ) ; the tall reed Trichoon karka; Lud - wigia octovalvis ; Polygonum minus; Cyperus odor at us ; other Cyperaceae on occasion (Cy- perus sp., Eleocharis geniculata, Scleria ) ; the fern Cyclosorus goggilodus ; the aquatic Blyxa octandra. Marginal species: Hibiscus tiliaceus, Aero sti- ck um aureum. Sandy Beach Formation (Cocos and Artocarpus must be included also.) Tree species: (1) Canopy trees — Hernandia sonora, Calophyllum inophyllum (scarce on Romonum), Barringtonia asiatica, Thespesia populnea , Pandanus (rarely). (2) Understory or smaller trees, or large shrubs — Scaevola tac- cada, Guettarda speciosa, Premna integrifolia, 112 Fig. 6. The ivory-nut palm, Metroxylon amt c arum. Morinda citrifolia, Allophylus timorensis, Ficus tinctoria, T erminalia samoensis, Cordia subcor- data. Climbing vines or scrambling low shrubs: Clerodendron inerme, Piper sp., W edelia bi- flora, Colubrina asiatica, Cassytha filiformis. Prostrate vines: Canavalia maritima, Vigna marina. Herbs or shrubs of low stature (generally less than 1 ft high) : Euphorbia chamissonis , Trium- fetta procumbens, the grasses Lepturus repens, Thuarea involuta, and the sedge Fimbristylis cymosus. Erect herbs: Tacca leontopetaloides, Crinum as i at i cum, Nephrolepis. Epiphytes: Phymatodes scolopendria (also terrestrial on occasion), Davallia solida, Den- drobium sp., Asplenium nidus. These weedy grasses frequently are found in sand: Eragrostis amabilis, Cenchrus echinatus, Dactyloctenium aegyptium, and, less commonly, Eleusine indica. Euphorbia hirta and E. thy mi- folia, Passiflora foetida, and Stachytarpheta jamaicensis are all found around houses near the beach. Except for the absence of certain species (e.g., PACIFIC SCIENCE, Vol. XXI, January 1967 Messerschmidia argentea, Soulamea amara, Suri- ana maritima, etc.) the plants enumerated here would be found on the reef islets of Truk and on most atolls in Micronesia as well. I Rocky Coastal Vegetation Trees: Ficus virens, Heritiera litt oralis, Bar- ringtonia asiatica, Thespesia populnea, Hibiscus tiliaceus. Shrubs: Allophylus timorensis, Desmodium umbellatum, Pemphis acidula. Herbs, low shrubs, or vines: Denis trifoliata i' (generally prostrate, but also a climbing vine), Nephrolepis, Procris pedunculata, W edelia bi- flora. Areas Under Cultivation, Past or Present This term is preferable to a specific one in- dicating a formation, since so little evidence of a recognizable indigenous formation is left. As such it is a loose heading under which may be assembled the various "villages” (really small 1 groups of houses) with their immediate en- virons, as well as the entire inner or central portion of the island (including the hill area) in which only a few scattered houses occur, but throughout which there are nearly continuous signs of either present or former cultivation. This is often not intensive nor extensive, but may consist of one or a few fruit trees (limes, j soursops, Carica papaya, the edible pandans, mango) scattered under virtually continuous cover of coconuts, breadfruits, and occasional mangoes or large Ficus, and mixed with such | relatively persistent, aggressive, or fast-growing noncultivated species as Acalypha indica, Maca- ranga carolinensis, and Glochidion Pramiflorum; with ornamental species such as Cananga odorata or Cordyline fruticosa; or with occasional native species which may be remnants of an earlier type of vegetation or perhaps are randomly opportunistic individuals of other formations, usually at the margins of their area (e.g., Premna, Pandanus, Hernandia) . The ground-cover species (in the sense of being at ground level — these may be scatttered rather than continuous) include a number of weeds, such as Euphorbia heterophylla, Malva- strum, Triumfetta semitriloba, and the weedy grasses. Most evident is the very extensive cover Flora of Romonum Island — Stone 113 Fig. 7. Pbymatodes scolopendria growing epiphytically on a branch of Hernandia sonora on the west beach of Romonum. formed, usually jointly, by Ischaemum muticum and Denis elliptic a. In small clearings may be found Cassia occidentalism Ipomoea digitata , I. indie as Dio scored, and Abrus precatorius. In gardens around houses, or on old house sites, and also sometimes at random in various parts of the higher parts of the island, will be found ornamental species and hedgerow species, i.e., Polyscias (various species), Zephyranthes rosea, Hymenocallis, bananas, variegated leaf pandans, Ocimum sanctum (used as a flavor- ing), Cordyline fruticosa, the aroids, and Cananga. In the west-central part of the island are the remains of the former Japanese colonial ad- ministration unit, with school, baths, well, generator plant, and other structures now mostly reduced to mere foundations or walls. In this area there are several ornamentals not found elsewhere. REFERENCES Bridge, Josiah. 1948. A restudy of the reported occurrence of schist on Truk, eastern Caroline Islands. Pacific Sci. 2(3) :215-222. Corner, E. J. H. 1965. Check-list of Ficus in Asia and Australasia with keys to identifica- tion. Gardens’ Bulletin (Singapore) 21(1): 1-186. Gressitt, J. L. Insects of Micronesia: Introduc- tion, pp. 1-257. B. P. Bishop Museum, Honolulu. Hess, H. H. 1946. Drowned ancient islands of the Pacific Basin. Am. J. Sci. 244:772-791. Imazeki, R. 1941. Materials of the Micronesian higher fungi. J. Jap. Bot. 17:175-184. 114 PACIFIC SCIENCE, VoL XXI, January 1967 Jatta, A. 1903. Licheni esotlci dell’ Erbario Levier raccolti nelF Asia Meriodionale e nelF Oceania. Malpighia 17:3-15. Kanehira, R. 1935. An enumeration of Micro™ nesian plants. J. Dept. Agr. Kyushu IJniv. 4(6):237-464. Kobayasi, Y. 1939. Fungi Austro- Japoniae et Micronesiae, I. Bot. Mag. Tokyo 51:749-803. III. Ibid. 53:158-162. Miller, H. A., H. O. Whittier, and C. E. B. Bonner. 1963. Bryoflora of the atolls of Micronesia. Nova Hedwigia, Heft 11:1-93, pis. 1-31. Okamura, K. 1916. List of marine algae col- lected in Carolines and Marianas Islands, 1915. Bot. Mag. Tokyo 30:1-14. Stone, Benjamin C. 1959. The Flora of Namonuito and the Hall Islands. Pacific Sci. 13:88-104. 1961 (with Peter J. R. Hill). The Vegetation of Yanagi Islet, Truk, Caroline Islands. Pacific Sci. 15:561-562. 1964. A review of the new botanical names published in Salford’s "Useful Plants of Guam.” Microneska 1:123-129. Sydow, H., and P. Sydow. 1921. Die Pilze Micronesiens aus der Sammlung Ledermann. Engl. Bot. Jahrb. 56:430-432. Tayama, R. 1940. Geomorphology, geology, and coral reefs of Truk Islands. Jubilee Publ. Comm. Prof. H. Yabe . . . 2:709-723. [English abstract in Jap. J. Geol. Geogr. 17: 60-61 (1940).] Taylor, W. R. 1950. Plants of Bikini and other northern Marshall Islands, pp. 1-227. Univ. of Michigan Press, Ann Arbor. A Hitherto Unrecorded Midge Gall of Myrsine australis (A, Rich.) Allan B. C. Arnold1 Stem galls on Myrsine australis were reported by Lamb (I960) in his checklist of New Zea- land Zoocecidea, but there appears to be no record yet of the conspicuous bud galls (Fig. 1) found near the branch extremities of this handsome tree, which is readily recognized by its red, mottled leaves, red stems, and dark red bark (Allan, 1961). When cut open, the dark bud galls may be seen to contain small white midge larvae or pupae in various stages of metamorphosis. The present report is concerned chiefly with anatomical modifications of the leaf bud by the midge larvae. I am indebted to the University Grants Com- mittee for a grant in aid of the work. METHODS AND MATERIALS Galls of Myrsine australis were examined on trees in the forest over a period of 16 months, and specimens for histological examination were fixed in Formo-acetic-alcohol. Serial sec- tions were cut at 10^ and stained in Safranin and Fast Green (Johansen, 1940). Freehand sections of living galls were made to observe the details of nutritive tissue which is slightly distorted by the fixative. Pupae were removed from galls and left in small stoppered glass vials to transform into adult midges. Artificial formation of the galls was at- tempted by removing small larvae from galls and placing them on tender buds of stem cut- tings. The cuttings were kept under bell jars in the south light of the laboratory window, with the base of the cuttings immersed in tap water in small flasks. OBSERVATIONS New galls were first apparent in December as dull olive-green, budlike structures (Fig. 1). 1 Department of Botany, University of Canterbury, Christchurch, New Zealand. Manuscript received Au- gust 31, 1965. Fig. 1. Bud galls of Myrsine australis. Within two months they had become almost black outwardly, masking the chlorophyll and red pigmentation of the internal tissues. Dur- ing a mild spell of weather in winter, a few additional young galls were formed in June, but this would seem to be an exceptional event. Under natural conditions in the forest, galls appear to live 10-12 months. By the following spring most galls are shrivelled and dry. Mature galls range from I/4 to % inch in length. Frequently the terminal gall of a branch- let may dominate the subjacent lateral galls (Fig. 2). The midge larvae are small and white with prominent salivary glands containing polytene banded chromosomes. The larvae transform within the gall into small black pupae with prominent eyes which are crimson at first and 115 Fig. 2. Large terminal gall and three smaller galls. later dark purple. The adult fly is shining black with simple, fragile wings. In the laboratory some flies emerged in autumn, probably in response to the high indoor temperatures. When young larvae were removed from galls and placed on the soft tips of new shoots in the laboratory, they made no attempt to enter the new host plant. (In other experiments by the author with the gall moth Morova subfasciata Walk., the larval insect was capable of re-enter- ing new shoots of the host after removal from the galls.) The attempt to elicit new galls on Myrsine australis with living larvae of the midge was unsuccessful, therefore. It is readily apparent that each gall is com- posed of two or three modified leaves which are fused together. A trace of the leaf blades may be seen at the top of the gall. Evidently the gall is derived chiefly from swollen petioles, which are curved and fused together to form a small urnlike structure occupied by one or more larvae. There is some variation in the degree of fusion of the modified leaves which form the PACIFIC SCIENCE, Vol. XXI, January 1967 walls of the gall. Where the leaf margins are merely closely pressed together, a boundary is recognizable ; in other cases the cells of each contributing swollen leaf are completely inter- knit, leaving no demarcation. Some galls are partitioned into two compart- ments, and others contain only a single loculus. The number of larvae per gall ranges from one to four, the larger galls generally containing more larvae than the smaller ones. In living galls the vaselike cavity is lined by nutritive tissue which may bear finger-like cells Fig. 3. Longitudinal section through a large gall. cn, Chlorophyllous nutritive parenchyma; m, midge larva. Midge Gall of Myrsine australis — Arnold 117 Fig. 4. Detail of Figure 3. v, Vascular strand; cn, chlorophyllous nutritive tissue; c, larval cavity; e, epi- dermis of gall. projecting inwards. The concave base of the cavity is composed of chlorophyllous nutritive parenchyma (Fig. 3) which continues to multi- ply as the larvae feed. This growth ceases soon after the larvae discontinue feeding. No remnant is left of the shoot apical meri- stem which originally gave rise to the two or three primordial leaves. In the development of galls at stem tips, increase in length of the branch is therefore curtailed at the same time as the leaf primordia are converted, under the in- fluence of the larvae, into a gall. The internal anatomy of the gall is rather more stemlike than leaflike, lacking as it does the characteristic stomatal arrangement and associated organisation of palisade layers and spongy mesophyll. A very prominent cuticle extends over the gall epidermis and is continuous with the equally thick cuticle of the normal epidermis of the supporting stem. Apart from the nutritive tissue surrounding the larvae, the histological features of the galls are the same as are found in the normal stem, namely: red pigmented cells, secretory cells, schizogenous cavities and canals with yellow or reddish brown contents, and normal vascular tissue (Metcalfe and Chalk, 1950). DISCUSSION A study of the life cycle of the gall midge was not undertaken, and the record of emer- gence of adult flies in autumn in the laboratory appears to be unseasonal, and related to higher average temperatures than those prevailing out of doors. It would seem that the main morphogenetic changes accompanying gall formation are a suppression of the activity of the marginal meristems of the leaf primordia togteher with the eventual destruction of the shoot apical meristem from which the leaf primordia arose. There is evidence of limited growth of the leaf apex. The bulk of the tissue appears to be con- 118 PACIFIC SCIENCE, Vol. XXI, January 1967 tributed by the swollen and fused petioles. These leaf parts are the nearest to the larva which occupies the former site of the shoot growing point. Substances emanating from the larva might tend to stimulate growth in basal parts of a rudimentary leaf and these regions might un- dergo cell division and growth at the expense of other parts of the unformed leaves. Fusion of the basal parts of rapidly multiplying leaf primordia could readily occur, assisted possibly by wound hormones and larval secretions. These would be free to operate without the overriding influence of the shoot apical meristem. In paraffin sections stained in Safranin and Fast Green, the salivary glands of larvae show characteristic giant cells with banded polytene chromosomes. The synthetic activities of the salivary gland cells not only may assist feeding of the larvae but also may provide the secretions which are responsible for the transformation of presump- tive leaf primordia to galls (Mani, 1964). SUMMARY A bud gall of My r sine australis (A. Rich.) Allan caused by a gall midge is reported for the first time. The galls are composed of modified leaves j which form the walls of an urnlike structure | enclosing the midge larvae or pupae. The larvae feed on proliferating chlorophyl- lous tissue which lines the larval cavity. During the development of the gall from a shoot bud, the apical meristem is destroyed by the larvae, and the leaf rudiments undergo | transformation and fusion. REFERENCES Allen, H. H. 1961. Flora of New Zealand, Vol. 1. Gov’t. Printer, Wellington. Johansen, D. 1940. Plant Microtechnique. McGraw-Hill Book Co., Inc., New York. Lamb, K. P. I960. A check list of New Zea- land plant galls (Zoocecidia) . Trans. Roy. Soc. N. Z. 88:121-139. Mani, M. S. 1964. Ecology of Plant Galls. Dr. j W. Junk, The Hague. ; Metcalfe, C. R., and L. Chalk. 1950. Anat- omy of the Dicotyledons, Vol. II. Oxford j University Press. An Unusual Example of Pseudoseisms1 Resulting from Military Exercises2 Harold L„ Krivoy, Charles G. Johnson, and Robert Y. Koyanagi3 ABSTRACT: Aerial bombing of the target island of Kahoolawe, Hawaii, during several hours on 19 and 20 December 1961 and on 13 February 1962 generated acoustic disturbances that were felt by people and recorded by seismometers on the island of Hawaii. The azimuth of arrival of the pseudoseisms was calculated from the accurate seismographic responses. Special atmospheric conditions are suspected as prime agencies in the propagation and focusing of these phenomena; lack of specific data in this field, however, leaves lative. Many communities have been disturbed by sonic booms created by high-speed aircraft. This problem has not yet become serious on the island of Hawaii, which has only recently established jet facilities. Nonetheless, series of disturbances resembling sonic booms were felt and recorded on Hawaii during several hours of the evening of 19 December 1961, the morning of 20 December 1961, and the evening of 13 February 1962. Since the affected neigh- borhoods are in a zone of active volcanism, prompt differentiation between artificial and natural events is a problem of immediate con- cern to the population and to the authorities. From the results presented here, it seems possible that careful study of the character and timing of sonic disturbances recorded on Hawaiian seismographs by experienced seismol- ogists may permit prompt identification of the source. Other experiments under more con- trolled conditions have already revealed im- portant facts about energy distribution, refrac- tion paths, etc. between the source and the receptors. EVENTS LEADING TO THIS STUDY Between 19:00 and 20:00 (hst) on 19 December 1961, an unusual variety of sensa- 1 This term is used in Gutenberg and Richter’s published description (1931) of a similar incident. 2 Publication authorized by the Director, U. S. Geo- logical Survey. 3 U. S. Geological Survey, Flagstaff, Arizona, Den- ver, Colorado, and Hawaiian Volcano Observatory, Hawaii, respectively. the matter of atmospheric structure specu- tions were both felt and recorded on the island of Hawaii. Personnel of the U. S. Geological Survey’s Hawaiian Volcano Observatory at Hawaii Volcanoes National Park, who were at their homes about 3 miles from the observatory, were aware of explosive shocks. When they reached the observatory they inspected the recordings being written by the high-gain, short-period, vertical seismographs at Desert, Uwekahuna, Ahua, and Mauna Loa stations. They recognized that the "pattern” of data as written by the Desert seismograph had been repeated by the Ahua seismograph after a lapse of about 17 seconds. (See Fig. 1 for station locations.) The observatory personnel assumed that these were sonic disturbances, and that they were not seismic events which would have swept the 15 -km net in less than 2 seconds. At about the time the shocks were felt in the National Park area they were also felt by residents in the communities of Volcano, Hilo, and Kau- mana. Civil Defense officials, alerted by many reports, called the observatory and were advised of the results of the seismogram analysis. During the evening, all possible local noise sources were checked carefully. Explosive vol- canism, though rare at Hawaiian volcanoes, is always a possibility. In this case, the lack of reports from Kona seemed to rule out eruption of the dormant volcano Hualalai. A radio check with personnel at the U. S. Weather Bureau, Mauna Loa Slope Observatory, 6 miles from the summit caldera of Mauna Loa, indicated that no one there had felt, heard, or sensed the dis- 119 120 PACIFIC SCIENCE, Vol. XXI, January 1967 01 2 3 4 5 KILOMETERS Fig. 1. Locations and elevations, in meters, of U. S. Geological Survey seismometers ( circles ) on Kilauea and a constructed sonic wave front based on data picked up by the seismometers. turbances, thus eliminating Mauna Loa as a source. Gunfire from the U. S. Army base at Pohakuloa, which is often heard by Hilo and Volcano residents during training exercises, was ruled out as a possible cause when it was learned that the base was then inactive. INTERPRETATION OF RECORDED DATA All four seismometers are timed from the same master clock, and accurate comparison of events recorded by these instruments is possible. Figure 2, a copy of the Desert seismogram, shows the disturbances described above, as well as those of a few local earthquakes for com- parison. As the Desert seismometer seemed to be most sensitive to the sonic arrivals, its record was used in compiling a master list, which in- cludes every suspicious event recorded during the disturbed period. The Uwekahuna seismo- , graph has a response similar to that of the ! Desert, although it is recorded optically. All events on the Uwekahuna record are shown in Table 1. Records from the Mauna Loa and Ahua instruments are similar to those from Desert. As stated above, events recorded on Desert were recorded on Ahua 17 seconds later. The next morning, 20 December 1961, ob- servatory personnel noticed renewed activity on the Desert record. People in the Hilo area, who perhaps were alerted by the events of the pre- vious evening, called the observatory, and, as j was the case the evening before, reported that j they had experienced an audible sensation rather than a physical one, and that windows vibrated strongly and wall clocks shook. With the expec- tation that Mauna Loa should be the first local station to record these events, the gain on that Pseudoseisms from Military Exercises — Krivoy, Johnson, and Koyanagi 121 TABLE 1 Arrival Times at Summit Stations of Sonic Disturbances from the Events of 19 December 1961 (Numbers in parentheses are recorded double-amplitudes, in millimeters) desert uwekahuna ahua mauna loa 18-29-06.3(3) 29-35.3(3) 31-37.2(6) 31-42.8(14) 35-42.8(4) 39-54.3(12) 54.6(5/2) 40-36.9(7) 40-39.8(10) 40-58.3(34) 59.1(4) 43-16.8(23) 17.5(3) 43-46.7(38) 47.5(2) 44-24.8(19) 26.3(1) 44-56.3(21) 57.0(2) 47-32.8(8) 33.5(5/2) 47-56.9(7) 57.8(4) 48-28.8(47) 30.0(8) 48-49.5(21) 50.1(2) 51-41.5(5) 42.0(5/2) 52-07.5(14) 08.0(3) 52-40.6(10) 41.4(2) 55-48.1(11) 48.6(5/2) 56-21.0(4) 21.5(4) 56-46.3(5) 46.7(4) 59-09.2(2) 59-48.5(5) 49.4(7/2) 19-26-42.1(10) 40.6(4) 27-14.1(18) 16.3(2) 27-49.5(27) 49.5(5) 30-57.0(2) 31-40.3(5) 40.6(4) 32-37.4(12) 37.7(6) 35-14.0(5) 16.3(3/2) 36-08.1(5) 07.2(2) 36-58.5(3) 38-56.1(10) 56.4(4) 39-51.0(22) 50.5(5) 41-01.1(19) 01.2(6) 48-53.8(3/2) 20-34-35.8(2) 48.6(1) 40-11.6(5) 18.1(5) 41-16.0(14) 34.5(21) 44- 04.6(16) 42.5(4) 45- 14.1(6) 22.1(5/2) 48- 14.5(2) 45.6(7) 49- 06.9(2) 47-53.2(2) 58.7(5) 24.9(6) 51-31.5(1) 04.9(3/2) 56-05.4(3) 12.0(5) 38.3(4) 55-45.0(5/2) 57-03.6(5/2) 10.1(3/2) 00-07.5(3) 57.6(8) 31.5(5/2) 28-06.4(8) 55.6(1) 53.0(3/2) 33.7(2) 04.0(2) 25.7(3) 35-31.8(2) 39- 13.4(5) 40- 08.5(11) 39-15.2(3/4) 18.6(5) amplifier was increased greatly. The gain on the Ahua amplifier was also increased. The data listed in Table 2 generally bear out the observa- tions suggested by Table 1. In Figure 3 are shown other locations at which there are instru- ments similar to that at Uwekahuna. It is interesting that such instruments on Maui and at Kamuela, Naalehu, Hilo, and Pahoa on Hawaii all failed to record these events. Further- more, residents and police at these places, as well as along the west and northeast coasts of Hawaii, who were questioned later, reported no unusual sensations. On 13 February 1962, 15 additional events were recorded. The observatory had been alerted to expect these events; and so, while they were recording, the staff phoned about six of the Hilo residents who had reported alarming sen- sations during the morning of 20 December 1961. Of those called, only one thought that a similar event had just taken place, although no concern or alarm was engendered. This third 122 PACIFIC SCIENCE, Vol. XXI, January 1967 Fig. 2. Copy of original Desert seismogram showing events described in text. Pseudoseisms from Military Exercises — Krivoy, Johnson, and Koyanagi 123 Fig. 3. State of Hawaii east of Oahu, showing locales discussed in the text. The area enlarged in Figure 1 is outlined herein. Stipples indicate direction and travel time of advancing wave front. series of events was recorded between 18:03 and 20:25 and most people who were called were watching television. A fourth series was recorded on 29 October 1963. It was sensed at the summit of Kilauea, but no calls were re- ceived from other places. The third and fourth series of events pre- sented a picture quite like the first two and have been only summarized in this presentation. Average differences between arrival times at Mauna Loa and at each of the other stations, for each of the four series of disturbances, are presented in Table 3. The fourth column of Table 3 corresponds to the fourth series (of five events) which was confirmed to be sonic in nature but was not recorded at Mauna Loa. Relative arrival times of the 29 October 1963 series at Desert, Uwekahuna, and Ahua are similar to those of the three earlier series. ORIGIN OF THE DISTURBANCES After the first series of disturbances on 19 and 20 December 1961, it was ascertained that Navy bombing exercises had been conducted during the suspect time-intervals. The target islands Kahoolawe and Kaula Rock had been "dive bombed" by aircraft from the carrier USS "Coral Sea," using 500- and 1,000-lb bombs which exploded on contact (rather than above or within the ground). Figure 3 shows the Hawaiian Islands east of Oahu and indicates sectors of airspace traversed during successive 1 -minute intervals by sonic wave fronts which originated at Kahoolawe. The region on Hawaii 124 PACIFIC SCIENCE, Vol. XXI, January 1967 TABLE 2 Matched Arrival Times at Summit Stations of the Events of 20 December 1961 (Numbers in parentheses are recorded double-amplitudes, in millimeters) DESERT UWEKAHUNA AHUA MAUNA LOA 09-56-42.5(21) 43.7(4) 57-00.9(5) 06.5(5/2) 10-26-56.6(14) 56.2(4) 27-12.9(6) 19-3(3/2) 27-15.5(3) 27-45.0(13) 45.0(4) 28-02.0(4) 08.0(2) 29- 18.8(24) 30- 03.0(3/2) 19.5(4) 36.2(3) 28-42.8(1) 30-48.9(14) 51.0(5) 31-08.0(3) 14.5(2) 31-52.0(5) 52.0(2) 11-11-07.3(6) 07.5(4) 27.5(3) 10-29.5(1) 40-54.7(12) 56.0(6) 41-12.0(1) 16.8(1) 41-23.3(17) 23.0(7) 41-40.5(1) 40-46.3(3/2) 41-44.8(17) 44.1(6) 42-02.3(2) 07.0(3) 42-12.9(16) 13.1(3) 30.7(3) 36.3(2) 47-40.6(5) 41.6(4) 57.9(2) 04.6(5/2) 48-40.9(20) 41.5(2) 59.1(1) 04.8(1) 49-04.2(16) 05.1(7) 21.3(2) 49-33.0(12) 34.6(11) 51.6(7) 48-57.6(2) 53-27.5(11) 27.8(2) 54-19.0(15) 19.5(4) 36.4(3) 53-42.9(3) 54- 38.2(10) 55- 12.0(4) 59-52.8(9) 38.6(2) 55.9(1) 01.8(1) 12-00-45.8(20) 01-03.5(6) 09.5(1) 01-12.0(14) 00-36.1(1/2) 06-31.0(11) 31.2(2) 49.5(2) 05-53.0(3/2) 07-28.0(5) 07- 32.2(4) 08- 02.5(4) 46.3(1) 11-44.5(21) 45.0(3) 12-03.0(5) ' 08.8(3) 12-44.4(20) 13-02.5(1) 08.6(3/2) 19-40.8(23) 41.8(8) 58.9(19) 04.7(3/2) 27-54.3(17) 52.2(2) 58-10.0(3) 16.2(3) 59-26.0(14) 26.7(3) 43.8(8) 58-49.6(5) 13-01-12.5(18) 13.7(9) 30.7(9) 00-36.6(4) 04-27.7(18) 28.7(7) 45.9(2) 03-51.7(2) 05-46.5(31) 47.5(6) 06-04.2(4) 10.1(6) 06-42.7(13) 43.1(3) 07-00.6(2) 06.3(1) 11-17.6(21) 18.6(6) 35.8(6) 10-41.2(4) 12-02.9(15) 03.8(3/2) 21.1(5/2) 11-26.6(7) 13-03.6(9) 04.6(2) 21.5(1) 12-27.5(3) TABLE 3 Time of Arrival of Sonic Wave Fronts at Each Station Compared with Arrival at Mauna Loa, for Each of Four Different Bombing Dates STATION ORDER OF ARRIVAL 19 dec. 1961 20 dec. 1961 13 feb. 1962 29 oct. 1963 Mauna Loa (1st) 0 0 0 not recorded Desert (2nd) + 35.9 sec + 36.4 sec + 36.6 sec "+36.3 sec” (assumed) Uwekahuna (3rd) + 0.7 sec + 0.6 sec + 0.2 sec • — 0.2 sec Ahua (4th) + 16.9 sec + 17.2 sec + 16.7 sec + 16.7 sec Pseudoseisms from Military Exercises — Krivoy, Johnson, and Koyanagi 125 in which the disturbances were recorded and felt is approximately 190 km from the Kahoolawe source. Kaula Rock is 400 km to the west of Kahoolawe. An approximate schedule supplied by the Navy did not specify the exact time of any individual explosion or the total number of bombs dropped. By virtue of overlap in attack periods on the two islands, however, there is weak evidence, based on schedules of exercises for 19 and 20 December 1961, that all dis- turbances recorded and felt on the island of Hawaii originated at the closer target, Kahoo- lawe. On 13 February 1962, both targets were bombed simultaneously, but because the bomb- ing log is skeletal, no conclusions can be drawn. No log was provided for the exercise of 29 October 1963; but it was confirmed that bomb- ing had taken place and it was implied that only Kahoolawe was involved. Bombing exercises on these targets during 1961 only, and qualitative results from them obtained by searching the seismograms follow: DATE HOURS RESULTS 24 Feb. 06:40-14:30 No recordings, eruption tremor in progress. No sensations reported. 16 May 05:30-13:00 Neither recordings nor sensations. 22 Aug. 05:00-12:00 Same as above. 15:30-20:00 Same as above. 23 Aug. 09:35-13:30 Same as above. 18 Nov. 05:40-13:10 Weakly recorded on Ahua only, no sensations re- ported. 20 Nov. 10:30-13:40 Neither recordings nor sen- sations. 17:45-19:20 Weakly recorded on Ahua only, no sensations. 21 Nov. 09:30-13:35 Very weak on Mauna Loa and Ahua only, no "felt” reports. 19 Dec. 06:30-12:30 Weakly recorded, no sen- sations reported. 17:45-20:00 Many reports and good records (Table l). 20 Dec. 09:30-13:00 Some reports and good records (Table 2). GRAPHICAL ANALYSIS: FIGURE 1 Because of the almost simultaneous arrival of the sonic waves at Uwekahuna and Desert, it was convenient to use a graphical technique to estimate the azimuth to the source on the basis of data summarized in Table 3. On Figure 1, therefore, an assumed wave front has been drawn at the instant it passes Desert seis- mometer. The seismic stations which time such wave fronts are arrayed along a chord which is approximately 20 km long and about 20 km from the source. With this geometry, the chord is within 0.3 km of the circular segment it inter- sects; therefore, straight-line wave fronts have been assumed. Seismograph recordings of sonic disturbances differ considerably from those of local earth- quakes. There are no definite phase identifica- tions for sonic arrivals as recorded by short- period seismometers; instead, the maxima have been read on each record. Some events were recorded as featureless bursts barely resolvable above the normal background noise. The Desert record as illustrated in Figure 2 usually gave the clearest and largest arrival. We can only specu- late on the possible interference of wave groups following slightly different paths from the source to the receivers and on the effect such interference would have on the times of maxima at different receivers. However, the relatively low velocity of sound in air reduces the degree of precision necessary in timing sonic arrivals compared with that for seismic waves, for example. If we assume that the correlation of the sonic arrival between receivers was in error by 5 or 10 seconds, and if we apply all of this error at either extreme of the hypothetical 20-km record- ing chord, errors of only 1 0 of arc would result. These errors are so small that confusing a source on Kahoolawe with one on Kaula Rock seems unlikely. If we make these simplifications and allow- ances for error, and if we assume further that constant velocity prevailed in the seismic record- ing zone and over the 200-km propagation path, a direction of N 44° -46° W may be read from the diagram on Figure 1. This solution is excel- lent for a Kahoolawe source. (The Kaula Rock target is about N 64° W of the seismic pickup location.) VARIATIONS IN APPARENT VELOCITY The traveltimes (averaged from Table 3) for each leg in the wave front’s passage are shown in Figure 1. The spacing provided by 126 PACIFIC SCIENCE, Vol. XXI, January 1967 i existing instrument locations indicates two dis- tinct travel paths : the high elevation path, from Mauna Loa to Desert/Uwekahuna, and the lower elevation path, from Desert/Uweka- huna to Ahua. Apparent velocity for the high elevation path is 0.33 km/sec, which, on the basis of the slope distance of 12.6 km, results in a ground velocity of 0.35 km/sec. These results are consistent with sonic velocities in air (Chemical Rubber Publishing Co., 1947:1928) of 0.33 km/sec at 0° C, and 0.34 km/sec at 20° C. Similarly, the lower elevation wave front path yields an apparent velocity of 0.37 km/sec, a condition observed at ambient temperatures of about 70° C. Blumenstock (1961) summarized weather data collected in Hawaii and concluded that the winter mean temperature was 20° at the Na- tional Park housing area (see Figure 1) and that it decreased 2° for every 1,000 ft of eleva- tion. He observed the remarkably "equable tem- perature conditions" in Hawaii — that is, the small range between winter and summer means at any one observation point — but he also stressed the great variations in temperature and in rainfall caused by very local topographic situations. Additional, near-surface temperature vari- ables which may . bear on our present problem are the diurnal temperature and wind-direction patterns. In table 3, the 19 and 20 December 1961 figures represent average traveltimes for a large number of events in each of two groups. The 19 December events occurred at night; the 20 December events occurred during the day- time. The apparent velocities which occurred in the two events are: (1) Mauna Loa to Desert (12.6 km slope distance): 35.9 sec, or 0.351 km/sec for the evening events; 36.4 sec, or 0.347 km/sec for the daytime events. (2) Desert to Ahua (6.4 km distance) : 17.6 sec, or 0.364 km/sec for the evening events; 17.8 sec, or 0.360 km/sec for the day- time events. As shown above, the gross velocity increases as the sound front moves from the slopes of Mauna Loa to the flatter terrain at the summit and flank of Kilauea, and the velocities are systematically lower in the daytime than at night. At one atmosphere pressure, the velocity of sound in air increases 0.012 km/sec between 0° C and 20° C. Therefore, the natural expecta- tion would be the reverse of our findings, i.e., velocities expectedly would be slightly greater during the daytime than during the evening, when temperatures are lower. Again, Blumen- j stock’s findings (1961:6) can be invoked for ; a mechanism which might explain this seeming i contradiction: "The usual regime is to have up- slope winds by day and downslope winds by night.” In our situation downslope winds * (nighttime) would augment velocities across our recording range; upslope winds (daytime) would provide relative decreases in apparent velocities. The velocity increase we seek to ex- plain by this mechanism is 4 m/sec or about 8 miles/hr — a modest windspeed vector which 1 is not unrealistic. Thus, some of the diurnal changes in sonic ? traveltime shown in Table 3 can be explained by assumptions of expectable change due to diurnal wind-velocity conditions. However, such changes can be only partially responsible for the difference between apparent velocity !i over the Mauna Loa-to-Desert leg and that over ; the Desert-to-Ahua leg. As we have pointed 1 out, such an assumption would require an un- j realistic ambient temperature of 70° C for the low-elevation, high-velocity segment. ANGLE OF INCIDENCE OVER THE RECORDING RANGE t It has been demonstrated above that the sonic ! travel path from Mauna Loa station down to Desert and Uwekahuna fits into a reasonable model for sound-wave rays moving parallel to ‘ the ground across that particular path. By con- I trast, the lower elevation segment of the record- ing range — that between Desert and Ahua — offers evidence of increased velocity which | cannot be explained by temperature alone. It might be explained by a favorable component ; of wind velocity, but that would require wind velocities in excess of 50 mph, a condition rarely ! observed in Hawaii. The best situation pro- ducing such a velocity increase, as well as one,; which would also provide for energy focusing, j would be encountered if the sonic rays impinged upon the low-elevation stations at a steeper angle of incidence. Pseudoseisms from Military Exercises— Krivoy, Johnson, and Koyanagi 127 A plausible hypothetical model is therefore considered, and with vastly oversimplified parameters. At an ambient temperature of 20° C, which is given by Blumenstock as the winter mean for the Park housing area, sonic velocity is 0.345 km/sec. In fact, we observe a velocity of 0.37 km/sec for the range south and east of the housing area. If all of this excess is assumed to be due to incidence angle, it may be computed: sin a — .345 = 69°, where a is the angle between the incident ray and the vertical. Energy from Kahoolawe can then be imagined to impinge upon the Mauna Loa and Kilauea recording range as follows: On the slopes of Mauna Loa, wave fronts move downhill and are normal to the surface of the ground; on the flatter, low-elevation terrain, wave fronts are about 20° from the vertical. Such increases in the angle of incidence would improve coupling between air and ground, and if such improve- ment occurs as theorized above, the maximiza- tion of available energy and "seismic” mani- festations reported by residents would be the expected results. THE GROSS PROPAGATION PATH The many uncertainties discussed for the limited region of acoustic recording are clearly multiplied in a consideration of the size and complexity of the air space through which the energy is refracted. At present there are few concentrated data which describe atmospheric conditions over Hawaii. For example, although daily weather observations are made at Hilo, these are limited to that place and concern only operational altitudes for aircraft. Perkins et al. (I960) illustrated many theo- retical and actual instances of the focusing of sonic energy due to meteorological conditions. Their work seemed to involve more limited source-to-target distances ; on their computed graphs data are restricted to areas having a maximum altitude of 10,000 ft and to lateral distances of about 100,000 ft. Variables dis- cussed in that report were those of temperature and wind velocity; the stratification thus pro- duced caused favorable refracting conditions and, in turn, focusing. Such conditions in Hawaii are known only generally, but salient features which would be propitious in generat- ing the special phenomena we have recorded are summarized: (1) Wind velocities which increase with al- titude in the Hawaii direction from Kahoolawe. This condition is prevalent in the winter when trade winds (blowing westward) abate and counter trade winds blow near the ground. Thus, the propagation pattern discussed above would be enhanced during the Hawaiian winter and it would normally be inoperative during the rest of the year. (2) High velocity (higher temperature) propagation paths which serve to refract energy in the Hawaii direction. This condition prevails most of the time (Blumenstock, 1961) in the form of a sharp temperature inversion overlying the Hawaiian area at altitudes between 5,000 and 7,000 ft. (3) Jet streams in a sheath above 40,000 ft, the least understood, but an important, feature. These jets, which supposedly blow toward the east (thus contributing to situation (1), above), can make radical changes in direction and can attain great velocities. Although more accurate information about possible zoning or velocity/temperature stratifi- cation between Kahoolawe and Kilauea would be helpful, it is still possible to come up with an approximation suggested by Perkins et al. (I960), who suggested a single gradient case. This oversimplification would call for Kahoo- lawe and Uwekahuna to lie at opposite ends of a chord connecting them. The chord, therefore, would be 190 km in length. And the circular path intersected by this chord would describe the simplified refracted energy path. If it is further assumed that tangents to this circular path at either end make an angle of 20° with the (horizontal) chord (i.e., if we interject the previously computed angle of incidence), a circular path 280 km in radius results. Such a path would reach a height of 17 km before refracting downward. This suggestion of a major refracting condition somewhere near an altitude of 55,000 ft is in good agreement with available knowledge about the altitude of the tropopause over Hawaii (described briefly in (3) above). 128 PACIFIC SCIENCE, Vol. XXI, January 1967 REFERENCES Berning, Warren W. 1948. Investigation of the propagation of blast waves over relatively large distances and the damaging possibilities of such propagation. Aberdeen Proving Ground, Maryland, Ballistic Research Lab. Rept. 675. (Reprinted by the U. S. Bur. Mines, August 1950.) Blumenstock, David I. 1961. Climate of the States — Hawaii. U. S. Dept. Commerce, Weather Bureau Climatography of the United States, No. 60-51. Gutenberg, B., and C. Richter. 1931. Pseudo-seisms caused by abnormal audibility of gunfire in California. Gerlands Beitrage zur Geophysik 31:155-157. Perkins, Beauregard, Jr., Paul H. Lorrain, and William H. Townsend. I960. Forecasting the focus of air blasts due to meteorological conditions in the lower atmosphere. Aberdeen Proving Ground, Maryland, Ballistic Re- search Lab. Rept. 1118. Gravity and Geological Studies of an Ultramafic Mass in New Zealand1 Alexander Malahoff ABSTRACT: A gravity and geologic survey was carried out over a portion of the Nelson ultramafic belt of the South Island. In this region, the ultramafic rocks out- crop over a 5-mile-wide belt and abut against the Alpine greywacke along the right lateral transcurrent Alpine Fault. The dunite and peridotite of the ultramafic belt as well as the overlying geosynclinal sediments strike north. At their southern ex- tremity, these rocks are faulted by the northeast-southwest striking Alpine Fault against the massive Alpine greywackes to the south of the fault. There is a com- plete discordance of the stratigraphic elements between the two sides of the fault. The basal Permian ultramafic belt (Wairau ultramafic mass) to the north of the fault is horizontally layered and shows inch-scale layering comparable to that ob- served by Hess in the Stillwater complex of Montana. Stratigraphically above the Wairau ultramafic mass and also on the northern side of the fault lies a vertically dipping, 31,000-ft-thick sequence of serpentinite, spilite, grey slate, red and green slate, and tuffaceous sandstone. The density of the rocks surrounding the Wairau ultramafic mass varies between 2.65 gm/cc and 2.75 gm/cc, while that of the peridotite and dunite varies between 3.2 gm/cc and 3.3 gm/cc. A total thickness of 7,000 ft for the Wairau ultramafic mass was computed, using the average den- sity contrast of 0.5 gm/cc between the ultrarnafics and the country rock. Gravity analysis also shows that the Alpine Fault dips 67° southeast along the contact between the ultrarnafics and the Alpine greywacke. It is thought that the Wairau ultramafic mass was emplaced as a vertical dike when the surrounding rocks were horizontal and that the dike and the surrounding rocks have been rotated by 90° so that the dike is now horizontal and the beds are vertical. Comparisons between the stratigraphic sequence studied here and an almost identical sequence on the southern side of the Alpine Fault in Otago province supports the previously postulated 300-mile-long transcurrent displacement between the two areas along the Alpine Fault system of New Zealand. Studies of displace- ment of post-glacial river terraces along the Alpine Fault in Nelson show an average right lateral movement of 0.36 inches per year along the fault since the last glacia- tion. The origin of emplacement of ultramafic rocks has always been a prime geologic problem in world geology. There are two ultramafic belts in New Zealand and these are separated by a 300-mile-long displacement along the Alpine Fault system of New Zealand. The New Zea- land ultramafic belts have an added interest be- cause of this prominent fault movement. They provide an accessible source for geophysical and geological investigation, in the country where dunite was first described. 1 Hawaii Institute of Geophysics Contribution No. 159. Manuscript received October 1, 1965. Of particular interest to the geophysicist and geologist alike are the two areas where the ultramafic rocks abut against the fault planes of the Alpine Fault. In these regions, a genuine physical cross section is obtained across the ul- tramafic rocks and their associated formations where the Alpine Fault system has cut across the formations. One of these two regions is located in South Nelson in the northern part of the South Island of New Zealand and has been named, in this paper, the Tophouse district. A reconnaissance geological survey was carried out by the author 129 130 PACIFIC SCIENCE, VoL XXI, January 1967 in 1961-1962 over the Tophouse district in order to map the geologic boundaries of the ultramafic belt and all visible fault traces of the Alpine Fault system. Another purpose of the geological survey was to compare the lithol- ogy, structure, and thickness of the geologic formations in the Tophouse district immediately north of the Wairau Fault with similar forma- tions south of the Alpine Fault in Otago, 300 miles southwest of the Tophouse district. A gravity survey was carried out simultane- ously with the geological survey over the Top- house district in order to map the maximum thickness of the ultramafic rocks, and their at- titude as they abut against the Wairau Fault, and to investigate whether there is any sub- surface extension of the ultramafic belt south of the fault in the Tophouse district. Gravity surveying appears to be one of the best geo- physical methods to use in the study of ultra- mafic belts because of the high density contrast usually measured between the peridotite and the country rock into which the peridotite has been intruded. It is hoped that this paper will serve as a use- ful contribution to the gravimetric study of ultramafic rocks throughout the world. GEOLOGY Outline of Stratigraphy The rocks of the Tophouse district (Fig. 1 and Table 1) consist of two classes, pre-Ter- tiary rocks and extensive post-glacial deposits. The pre-Tertiary rocks are divided into three fault blocks (Fig. 2) : (1) the Brook Street vol- canics west of the Waimea Fault and probably underlain by rocks of the Rotoroa igneous com- plex; (2) the Maitai and Te Anau series, east of the Waimea Fault and north of the Wairau Fault; and (3) the Alpine greywacke, south of the Wairau Fault. The relationship between the rocks of the Tophouse district and those of the ultramafic belt in Otago south of the Alpine Fault is shown in Table 2. Outline of Structure The Wairau Fault, the major one of the dis- trict, is a right lateral fault downthrown to the north. It is probably a branch of the Alpine Fault, together with which it forms a 300- mile-long transcurrent fault system which sepa- j rates the Maitai and Te Anau rocks of the Top- house district from those of Otago (Wellman, j 1956:25). The strike of the Brook Street volcanics has changed as a result of stresses associated with movement along the Wairau Fault. At Top- house, the Brook Street volcanics strike at 360°, j 1. e., parallel to the Waimea Fault. At Lake Rotoiti, these volcanics have been regionally j bent to strike at 60° and fault swarms have jj developed at intermediate angles between the j strike of the Waimea and Wairau faults. The Waimea Fault has not been active during the post-glacial period. The Maitai and Te Anau formations have 1 undergone strike-faulting, but the Wairau ul- 1 tramafic mass shows little sign of structural de- j formation. Detailed geology of the Tophouse district is shown on the geological map, Figure { 2, and the stratigraphic units are presented in '■ Table 1. Brook Street Volcanics The Brook Street volcanics of the Te Anau series are part of the southern end of a sequence j of Upper Paleozoic volcanics which extend from d’Urville Island to Tophouse. The con- tinuous Waimea Fault between the Brook Street volcanics and the Maitai series makes strati- graphic relationship at Tophouse uncertain. No j! fossils were found in the Brook Street volcanics j in the Tophouse district; however, they are jj considered Upper Paleozoic here, the age as- !j signed by Bruce (1962:166) to the Brook !j Street volcanics of Nelson. The bulk of the rocks are massive metasoma- tized spilites and green-grey keratophyres. Along § the Waimea Fault, there is an outcrop of a j green volcanic conglomerate. The spilites are dense, hard and nonvesicular. The dip of the j volcanics has been determined from alignment 1 of xenoliths and mineral grains. The spilite ex- posed in the Motupiko Valley is equigranular jj and fine grained, with 20% subcalcic augite I (2V = 20°) and with 60% plagioclase feld- spar. The augite crystals are corroded and set in a highly altered groundmass of epidote and chlorite. The green color of the Brook Street volcanics results from the alteration of mafic minerals to chlorite. Ultramafic Mass in New Zealand — Malahoff 131 Fig. 1 Holocene sediments Tertiary sediments Triassic sediments Tasmon Intrusives BUI Brook Street vo Maitoi Series Te Anau volcanics Te Anau ultramaf ics TeAnau sediments Iconics Rock types north of the Tophouse district. Inset shows geographic location of Tophouse district. Fig. 2. Geology of the Tophouse district. Cenozoic covering strata lithology I H 1 Post-glacial river deposits | S 1 Grey wacke scree 1 Tm 1 Uplifted Pleistocene gravels (Tophouse gravels* covered by moraine in places(approximat* boundary only) Paleozoic and Lower Mesozoic rocks BEDDING Inclined Ve r t i ca I Distinct boundary Indistinct boundary Fau Its Faults recently act ive FORMATION, GROUP Stevens Formation siltstone, sands tone, congl. Wa i ua Format ion red and green argil I i te Greyville Formation grey banded ms. and ss. Whits hard limestone j > Rangi toto F Calcareous sandstone I S p 1 1 i te, delsr i te Red and green volcanic breccia S e r pa n t i n i te Dunite, peridotite SYSTEM, SERIES > Ma i t a i se ries PERIOD Lower Perm ian Wa ira u ES v Ultramafics Dunite scree, dunite > Mass Hydrogrossula in serpe n t i ne R o d i n g i t e A m ph i bol i te x Te A na u S series Basal P® rmian Rocks west of Waimea Fault B | Brook Street Volcanics Rocks south of Wairau Fault R Mt. Robert Group W Raglan Group Grtywoeke Alpine Facies (?) C a rbon if erous (?) T riassic 134 PACIFIC SCIENCE, Vol. XXI, January 1967 TABLE 1 Stratigraphic Units of the Tophouse District SERIES AND AGE FORMATION MAP SYMBOL LITHOLOGY MAXIMUM THICKNESS IN FEET* s mountain scree Wairau surface H gravels, 10-ft terrace 100? Holocene Tophouse surface T fluvial gravels, silts, 300? 200-ft terrace Tm glacial moraine 200? Major Unconformity Tertiary t terrestrial sediments 600? Major Unconformity Stevens formation YSa tuffaceous siltstone (5,000 ft thick) YSb tuffaceous sandstone 11,000 (6,000 ft thick) Maitai Waiua formation Yw green and purple banded Series, mudstone and sandstone 2,500 Upper Greville formation YG grey banded argillite 9,000 Permian Rangitoto formation YRA. white limestone (?) (2,000 ft thick) YR grey calcareous sandstone 3,800 (1,800 ft thick) Cessation of Igneous Activity Es spilite, dolerite 4,000 (extensively haematized) EB haematized red and green volcanic breccia 1,000 Te Anau Series, Stratigraphic Relationship Uncertain Basal Permian E|3 serpentinite 2,000 Ey dunite, harzburgite, pyroxene peridotite 7,000 Wairau Eys peridotite and ultramafic peridotite scree mass E8 serpentine containing 500 hydrogrossular Ee amphibolite, rodingite 5,000? Rocks West of the Waimea Fault Te Anau Brook Street B keratophyre, andesite, Series, volcanics volcanic agglomerate, 12,000 Carbonif- spilite erous (?) Rocks Southeast of the Wairau Fault R Mount Robert group- p Triassic Alpine facies greywacke, argillite W Raglan group- p greywacke, argillite As measured across outcrops in the Tophouse district. TABLE 2 Ultramafic Mass in New Zealand — Malahoff 135 * o o < H O o o w W u o P3 Z < s pe5 W Ph Z W w H w PQ Z O H < i-l w pej O U y X < (4 O H < H CO o o o o o o o o o W < eej Q w y c w) > « 8 oo G _■. Oh-S tr u 2 & Oh h 3 £ wo H -D -G d vT 2 o S g u Ja © ►. O J3 •— * ^ to „ tl ^ © 3 +J H3 a 8 g n, ^ o ^ 4-4 '1=5 a, g a, Oh o. ^ Lfi ^ m C^> 5 8 8 H "d H3 n3 © nd "d s ~ w 8 13 ^ ^ -g > a, a, C w W y lndurated Basaltic Sand Tan-Gray Limy Mud and Sand RSf Indurated Brown Basaltic Sand UNCONFORMITY— fSSf-Cobble-Pebble Conglomerate f.’f.'jfea.Soil nnnl- fc\Hard Basalt - 1 100 ■“ xlinker LEGEND Reef Limestone The width of the log is directly proportional to the hardness of the rock. ^~bana ana bur h^J"Gravel, Boulders |6jj Brown, Gray and White Muds ® Son 0 Green and Black Mud -Lignite Fig. 3. East-west geologic section across Ewa Coastal Plain, Oahu. 156 drilled on the Ewa Coastal Plain are not the only holes that have been drilled on that plain. Hundreds of wells have been drilled from Barber’s Point to Honolulu and from the mountains to the sea during the past 40 years (Stearns and Vaksvik, 1935, 1938). However, no samples were saved from these holes. Logs of these wells are frequently ambiguous as to terminology; "coral” and "limestone” are used interchangeably, and it is suspected that any white rock penetrated by the drill was logged as "coral.” Using recovered cores from Ewa Nos. 1 and 2, however, it has been possible to go back to these earlier logs and reinterpret them. In this manner a great deal of additional informa- tion was obtained that could be used to extrap- olate the data from the Ewa holes (Figs. 2 and 3). Mr. William Ebersole assisted in the project The holes were drilled by Layne International Company of Honolulu. LITHOLOGIC TYPES The sedimentary rock in the two Ewa cores consists mainly of various types of reef lime- stone alternating with shallow-water muds. A few soils, a layer of beach rock, two beds of lignite, basaltic sands, and stream conglomerates were encountered. The igneous rock recovered in the lower portion of each core consists of alternate flows of pahoehoe and aa basalt. Lithologic terms used throughout the following discussion are defined below: Reef limestone. A sedimentary rock con- sisting of the remains of various corals, mainly Porites, calcareous algae, molluscs, etc., essenti- ally in position of growth. Much of the original skeletal material has been replaced by secon- dary calcite and/or dolomite. Mud. A marine or fresh water sediment con- sisting of particle diameters mainly in the silt and clay size range, i.e., 1/16 mm to about 1/1000 mm, and composed of various detrital minerals resulting from terrestrial weathering. Marine shells may be present. The various types of muds are described in terms of their colors; the muds in the Ewa cores owe their colors to the following constituents: (a) oxides of iron and aluminum (brown mud) ; (b) iron PACIFIC SCIENCE, Voi. XXI, April 1967 sulfides and organic detritus (black mud) ; (c) calcium carbonate particles mixed with brown or black mud (gray mud) ; (d) clay size particles of calcium carbonate presumably reef detritus (white mud) ; (e) clay minerals and ferrous iron (green mud). Beach rock. A sedimentary rock consisting of calcareous beach sand cemented by calcium carbonate. Beach rock is commonly found form- ing within the beaches of tropical islands and owes its origin to the seepage of carbonate-rich ground waters through a beach composed of calcium carbonate particles. Beach rock is formed only at or within the tidal range and positively indicates a former shore line. Reef breccia. A sedimentary rock composed of the angular fragments of an organic reef. The broken fragments may be y2-4 inches in diameter and are commonly mixed with sand and silt-size reef debris. Lignite. Fossil plant remains altered by pres- sure to a highly friable, soft, black sedimentary rock. A low grade of coal. Conglomerate. A sedimentary rock composed of rounded cobbles and pebbles intermixed with finer material. Clinker. Rough, spinose, vesicular fragments j of lava produced by lava flow. Pahoehoe basalt. Lava with a smooth or ropy j surface spread chiefly through tubes and char- ; acterized by round vesicles. Aa basalt . A lava flow with a rough clinkery surface and base. Deflated and stretched vesicles characterize the solid part of the flow. Marl. A calcareous clay. Soil. The term is used in a general way to i mean regolith on the basalt and any sediment altered by weathering. Cobbles, pebbles, gravel, sand, silt, clay. The usage herein follows the usual dictionary defini- tions. MEGASCOPIC DESCRIPTION OF CORES Ewa No. 1 (Table 1) The drilling site for this hole was located as far seaward on the Ewa Coastal Plain as it was feasible to drill, within 200 yards of the sea on the 158th meridian on the eastern end of the property of the U. S. Coast and Geodetic Station, 91-270 Fort Weaver Road, Ewa Beach, Deep Cores of Oahu — Stearns and Chamberlain 157 Oahu, opposite Ewa Beach Park. The terrace at this locality was a flat, low, emerged coral reef of undetermined age, partially covered with a thin discontinuous soil layer. The ground level is -}- 6.1 ft above mean sea level; all depths in the core are measured from ground level equaling zero. Ewa No. 2 (Table 2) The second hole on the Ewa Coastal Plain was also located on the 158th meridian but 2 miles inland from Ewa No. 1. The exact locality was within the confines of the West Loch of Pearl Harbor Naval Base, at a point a few hundred yards south of the West Loch shoreline. The ground level at the hole is 19.7 ft above mean sea level; all depths in the core are measured from ground level equaling zero. PRELIMINARY SEISMIC DATA On July 13 and 14, 1965 seismic refraction studies were made of the ocean bottom to the south of drilling site Ewa No. 1. The following results were obtained at a distance of 8.4 km seaward from the coast along the 158th merid- ian (21°15'N, 158°00'W): (1) water depth — 0.36 km; (2) depth from sea surface to the upper surface of the basalt basement — 1.1 km; (3) sound velocity in the sedimentary section = 2.8 km/second. These data show the sedimentary column to be 2,920 ft in thickness at a distance of 4.6 miles offshore from Ewa No. 1 in a water depth of 1,182 ft. The sediment-basalt interface was found to be essentially parallel to the sea water- sediment interface. PRELIMINARY GEOLOGIC HISTORY OF THE EWA COASTAL PLAIN The following description applies to the outer edge of the Ewa Plain in the vicinity of Ewa No. 1 : 1. Prolonged weathering and erosion of the upper surface of the Koolau basalts. Formation of thick soil deposits, deep incision of stream valleys, and deposition of stream cobbles, peb- bles, and basaltic sand along the coast. 2. Gradual submergence. 3. Accumulation of thick deposits of shallow marine lagoonal sediments behind a barrier reef. Stream-transported muds and silts pre- dominate, with occasional layers of basaltic sand and gravel. Inorganically precipitated CaC03 common to these sediments indicates a restricted oceanic circulation. Much of the mud is high in organic carbon, indicating swamp conditions. Typical lagoonal-deltaic sedimentary facies. 4. With continued submergence, the water deepened sufficiently to allow the lagoonal deposits to be superseded by calcareous muds and coral debris. These sediments indicate the barrier reef structure was in close proximity. The upper portion of this section grades into a hard reef limestone horizon at — 786 ft msl in the core. 5. Following the growth of these corals the progradation of the land was sufficient to shift the coral reef facies seaward, allowing at first the accumulation of gray calcareous mud and coral debris, and finally, the progradation of the land was sufficient to bring basaltic river sands and silts and dark-gray to black organic muds into the area. The environment again became swampy-lagoonal and eventually peat deposits accumulated, now represented by the lignite and soils found at — 624 ft msl. At this depth a major unconformity occurs which probably marks the Pleistocene-Pliocene boun- dary. 6. Following the deposition of the lignite beds the sea level rose, allowing the coral facies to shift landward. At first calcareous muds containing coral debris accumulated, but these were followed by the growth of marine coral- line reefs more than 50 ft in thickness. 7. On top of this reef is found at first cal- careous mud followed by brown mud and sands and soils, indicating a progradation of the lagoonal facies. At — 406 ft msl a minor unconformity occurs and continues upward through brown muds and basaltic sands and soils to a major unconformity at — 358 ft msl that most likely corresponds to the Kahipa- Mamala submarine shelf around Oahu (Stearns, 1966). 8. Subsequent to the development of this 158 PACIFIC SCIENCE, Vol. XXI, April 1967 TABLE l Description of Core from Ewa No. l Hole DEPTH (in feet) ROCK TYPE DESCRIPTIVE NOTES 0-2 Loose coral and sand Hole started in exposed emerged coral reef. 2-41.8 Reef limestone Coral in upright position of growth and reef debris. 41.8-43.8 Tuff (?) Stratified fine grained material with thin horizontal calcite layers, possibly an altered tuff. 43.8-165 Reef limestone Mostly Porties coral and nullipores. The corals are in upright position of growth. Starting at 96 ft the cavities in the reef limestone contain coatings of a red clay which X-ray and mineral analysis indicate is sediment derived from a basaltic terrane. The red mud becomes only a trace below 135 ft. 165-166 Brown mud with fragments of coral 1 inch across 166-167 Brown compact mud Brown mud full of fossil molluscs and fragments that represents a discontinuity. 167-203 Reef limestone Hard white and cream colored reef with no soil in cavities. 203-209 Beach rock Thin-bedded beach rock; contains about 10% basaltic grains, the rest Foraminifera, shell, and coral grains, very well rounded and cemented by calcite into hard limestone. 209-250 Brown mud Brown mud containing 1-20% limestone grains. 250-270 Reef limestone Hard fragmental reef with some brown mud in cavities. Much of it is recrystallized limestone. 270-283 Brown mud Dark-brown organic mud mixed with coral fragments mostly 1 to 2 inches across. 283-290 Altered reef limestone Soft, slightly muddy, powdery limestone; apparently altered top of reef containing thin limonite streaks. 290-311 Reef limestone Fragmented limestone, mostly recrystallized and broken by drilling. Some heads of Forties and a few molluscs. 311-314 Muddy limestone Recrystallized reef with 50% brown mud in interstices, mud content increasing downward. 314-315 Brown mud About 90% brown mud and 10% limestone fragments. 315-331 Reef limestone Muddy reef limestone, partly recrystallized. 331-333 Gravel (?) Partly rounded reef limestone %-! inch across, mixed with similar size and shape basalt pebbles. Basaltic pebbles constitute 1% of deposit. 333-337 Reef limestone Similar to 315-331 ft, with a few subrounded pebbles of basalt. 337-348 Reef limestone Mud becomes whiter progressively with depth and decreases in quantity at 341 ft, where fragments in- crease in size. 348-350 Reef limestone Same as above but white. Possibly the white mud is due to grinding action of bit. 350-355 Reef limestone Brown mud filling interstices in a reef. Bit breaks it all up and makes a fragmental deposit with brown coating. Similar to 315-331 ft. 355-358 Limey mud breccia White and brown mud with mostly small limestone frag- ments less than 1 inch across, a few l1/^ inches across. Deep Cores of Oahu — Stearns and Chamberlain 159 TABLE 1 ( Continued ) Description of Core from Ewa No. l Hole DEPTH (in feet) ROCK TYPE DESCRIPTIVE NOTES 358-363 Brown mud Pure dark-brown mud with irregular tubular cavities p2 mm across lined with limonite. No bedding visible. Last 6 inches is light brown. 363-364 Gray sand Gray-brown fine sand; over 50% basaltic grains eroded from a basaltic terrane or water-laid lithic tuff deposit. No glass particles obvious. 364-371 Calcareous mud Calcareous mud with limonitic streaks. 371-376 Reef limestone Muddy fragmental reef limestone, much altered. 376-383 Calcareous mud White calcareous mud. 383-394 Reef limestone Muddy fragmental reef limestone probably broken by bit. 394-401 Mud and coral fragments Layers of mud and mixed mud and coral fragments. 401-412 Brown mud Brown mud with scarce limy grains. A 4-inch layer of fine weathered basaltic sand from 403.5 to 404 ft. 412-415 Reef limestone Muddy reef limestone, with about 1 ft of altered lime- stone at top with laminations. 415-435 White limy mud White chalky mud with a few hard chunks. Contains minute borings of marine organisms. Probably a chemical precipitate. Lumps of hard limestone at 424 ft and from 426-435 ft. Some are altered coral fragments. 435-444 Reef limestone Recrystallized reef limestone and white mud. Transition into material above. Mostly fragments broken up by drilling ( ? ) . At 440 ft mud becomes browner and at 444 ft becomes predominant over coral. 444-453 Brown mud and limestone fragments Mud and coral reef limestone fragments. 453-464 Gray mud Same as above but fewer rock fragments. Possibility that some rock fragments are chiefly crystallized calcite in place. 464-464. 5 Organic mud Brown layer with 1-inch layer of black lignite at bottom. 464.5-472 Gray mud Gray lagoonal mud; in places 4 inches of it is hard cemented mud-limestone. Contains one oyster shell and a few other types of molluscs. 472-493 White mud White calcareous mud with hard crystalline calcite lumps toward the bottom. Spherical and oval grains suggest altered Foraminifera. 493-497 Reef breccia (?) Fragmental limestone containing large oyster shells and other molluscs. The mud matrix is darker than above. The whole deposit resembles a fine-grained reef talus deposit. 497-572 Reef limestone Fragments of reef limestone %-3 inches across, probably broken by bit; probably highly permeable structure. Oyster shell at 523 ft. Mostly recrystallized. At 535- 540 ft several zones of smaller sized fragments and white mud. Oyster at 546 ft. Porites at 545 ft. At 545-555 ft much recognizable coral, less altered than above. 160 PACIFIC SCIENCE, Vol. XXI, April 1967 TABLE 1 ( Continued ) Description of Core from Ewa No. 1 Hole DEPTH (in feet) ROCK TYPE descriptive notes 572-575 Brown mud and lime A brown mud full of streaks and nodules of lime and a few shells. At 575 ft a hard 1-inch layer of dark greenish claystone full of shells in excellent state of preservation. 575-585 White mud and lime White calcareous mud full of nodules becoming indurated at 579-580 ft, then calcareous mud again. 585-590 Reef limestone Highly altered fragmental reef. 590-597 White limy mud White calcareous mud with nodules scattered throughout. 597-609 Greenish mud Greenish calcareous mud, highly fossiliferous, con- taining irregular red iron oxide streaks. 609-617 White mud At 611 ft in the white mud is a 1-inch layer of fine grained tuff. A few chunks of very hard chemically precipitated limestone. 617-629 Green and black mud Greenish calcareous mud, highly fossiliferous to 623 ft, then organic mud becoming blacker with depth. 629-631 Lignite Firm lignite full of fossil plant remains; no shells. 631-635 Gray-green mud Calcareous organic mud full of shells. 635-646 Gray mud Gray mud with shells. 646-660 Dark-gray mud Mud is becoming more organic ; still highly fossiliferous. 660— 668 Green and black mud Greenish-brown to olive black mud. Recognizable weathered basaltic grains. Few fossils. 668-675.8 Black sand Thin-bedded, compact, fine basaltic sand and silt. 675.8-676.5 Fine sand Fine calcareous sand and silt, highly fossiliferous. 676.5-686 Gray mud Gray calcareous mud with two beds of black mud at 678 and 678.5 ft. Shells are in thin zones. 686-706 Tan mud Tan mud full of fossils. A layer of calcareous sand with abundant rounded grains of basalt and one un- weathered feldspar crystal possibly indicating tuff source at 692-693 ft. Laminated at 699 ft. 706-727 White mud Chalky white mud; fossils scarce. 727-727.4 Brown mud Firm brown mud. 727.4-728 Gray mud Calcareous gray mud. 728-735.5 Reef limestone Hard reef limestone; some layers contain grains of lime. 735.5-7 64 White mud and limestone nodules Hard limestone fragments in white mud, possibly a breccia transitional to reef below. 764-792 Reef limestone Highly altered fragmental reef limestone with a few shell molds and pockets of clay. White mud layer at 776-778 ft. 792-811 Gray mud Gray calcareous mud, some limy streaks, and scarce solid nodules and concretions. 06 T CO ON Reef limestone Reef limestone. 816-851 Gray mud Gray calcareous mud with hard nodules and con- cretions up to % inch across. Some indurated layers. At 846-851 ft some mixed gray and green mud. Deep Cores of Oahu — Stearns and Chamberlain 161 TABLE 1 ( Continued ) Description of Core from Ewa No. l Hole DEPTH (in feet) rock type DESCRIPTIVE NOTES 851-858 Olive black mud Highly fossiliferous mud with sharp basal contact. 858-931 Gray and black mud Gray calcareous mud with no nodules and some shells. At 874 ft a !/2-inch layer of basaltic sand with some layers of gray-black and black mud, highly fossiliferous and full of basaltic grains and microscopic fossils. Some silt layers at 911-921 ft. 931-942 Olive black mud Similar to above, poor in fossils, but uniformly dark.. 942-948 Gray mud Similar to above except for color. 948-950 Basaltic sand Indurated basaltic sand and clay; basalt grains diverse and weathered. 950-966 Gray mud Indurated gray calcareous mud. No fossils observed. 966-979 Brown mud Indurated brownish-black mud. 979-980.5 Brown sand Indurated fine basaltic sand and silt. 980.5-981.5 Brown clay Indurated brown silty clay. 981.5-984 Brown sand and gravel Indurated brown basaltic sand with scarce pebbles up to p2 inch across. 984-991 Brown mud Indurated brown fossiliferous silty clay containing a 1-inch piece of Poriies coral embedded in clay at 988 and 990 ft. 991-1015 Gray mud Indurated gray mud with limy zones and nodules. Oyster at 1,004 ft and more at 1,009 ft. Another oyster at 1,015 ft. 1015-1043 Tan and gray mud Hard very indurated brown mud, in places fossiliferous. Oyster shell, other fossils at 1,025 ft. 1043-1054 Brown sand Typical brown basaltic indurated sand ; grains mostly weathered limonite stained areas. 1054-1061.5 Conglomerate Cobbles and pebbles up to 6 inches across, mostly dense blue basalt with a layer of silty clay, sand, and small pebbles at 1,055.5-1,057 ft. The sandy layers may be the matrix washed by drilling. Sand again at 1,061- 1,061.5 ft. 1061.5-1072 Brown clay Well indurated brown silty clay. No fossils noted. Basaltic grains visible. Becomes sandy at 1,071 ft for 1 ft. 1072-1077.5 Weathered basalt Weathered basaltic aa clinker typical of a subsoil, con- sisting of partly decomposed clinker in a softer matrix, with creamy montmorillonite in the interstices. The top soil has been eroded away by the stream which emplaced the basal conglomerate. 1077.5-1088.3 Basalt Solid basalt with large stretched vesicles typical of an aa lava. One unbroken core is 31V2 inches long. The rock is nonporphyritic. 1088.3-1089 Basaltic clinker Partly weathered red clinker. 1089-1097 Basaltic clinker Red aa clinker. 1097-1107 Pahoehoe Very vesicular olivine pahoehoe with slightly weathered surface. 162 PACIFIC SCIENCE, Vol. XXI, April 1967 TABLE 2 Description of Core from Ewa No. 2 Hole DEPTH (in feet) ROCK TYPE DESCRIPTIVE NOTES 0-10 Artificial fill Crushed blue basalt and coral fill for drill platform. 10-18.8 Calcareous soil Tan calcareous muddy soil with secondary calcified lumps. 18.8-36.5 Brown sandy soil Chiefly basaltic grains; a few pebbles of basalt inch across. Some secondary calcite nodules. Changes to plastic brown clay downward. 36.5-46 Gray marl Lumps of lime and concretions in gray mud ; probably weathered surface of underlying reef. 46-102 Reef limestone Hard reef limestone with shell molds, much recrystallized. 102-118 Reef detritus Reef limestone with red mud in the interstices to 111 ft, and then changes to gray mud. 118-120 Brown sand Fine silt and sand becoming coarser toward the bottom. Sand contains 50% well-rounded basaltic grains. 120-122 Coarse calcareous sand and gravel Subangular reef detritus. 122-129 Brown mud Fine mud. No lime present. 129-141 Indurated limy mud Grayish-brown indurated mud with sharp break at top. Suggests very different environment. Contains tiny holes, possibly root holes with limonitic stain. Silty at 137-139 ft, with concre- tions. Some holes are lined with concentric structure. 141-161 Brown mud Brown mud with mottled soil structure. 161-162 Brown silt and sand Brown basaltic sand. 162-163 Brown mud 163-165 White mud White calcareous mud with some calcareous fragments; much recrystallized calcite with %-inch crystals. 165-182 Reef limestone Recrystallized reef limestone in fragments. 182-184 Limestone fragments in brown mud Broken reef fragments in mud. 184-184.2 Black organic mud 184.2-192 Brown mud 192-194 Fine sand Basaltic fine sand. 194-199 Brown mud 199-204 Muddy fine sand 204-208 Gravel and sand Basaltic pebbles partly weathered in a sandy matrix. 208-239 Brown mud Brown mud with pure calcite lumps V2 inch across. 239-241 Gravel and sand Brown dirty basaltic sand with tiny round pebbles up to *4 inch across. 241-262.5 Brown mud 262.5-263 Sand and gravel Brown nearly completely weathered basaltic gravel (l/4 inch or less) and sand. 263-268 Brown mud 268-271 Coarse sand Coarse sand (basaltic) and mud. 271-274 Brown mud Brown basaltic mud. 274-286 Brown and Brown mud with soft secondary lime deposits and irregular white mud masses. A few hard calcite lumps. Lumps become more numerous at 285 ft. 286-288 Brown mud 288-290 Fine sand Fine basaltic sand and mud. Sand grains are mostly decomposed and have a variety of colors. Deep Cores of Oahu — Stearns and Chamberlain 163 TABLE 2 ( Continued ) Description of Core from Ewa No. 2 Hole DEPTH (in feet) ROCK TYPE DESCRIPTIVE NOTES 290-294 Brown mud Brown basaltic mud. 294-297 Fine Sand Brown basaltic sand and mud. 297-300 Brown mud Brown basaltic mud. 300-301 Brown sand Brown basaltic sand showing %-inch and %-inch layers of coarse and fine sand. 301-307 Brown mud Brown basaltic mud. 307-310 Fine sand Brown basaltic sand and mud. 310-326 Brown mud Brown basaltic mud, very mottled starting at 328 ft. Much limonitic from 333 to 334 ft along fractures. 326-330 Brown sand Basaltic sand and silt much weathered. 330-339 Brown mud Brown laminated mud. 339-344 Brown sand Brown calcareous (secondary) sand, in places indurated with lime. 344-381 Brown silt and mud Brown silt with soil structures in places. 381-400 White and gray marl Mottled gray, white, and brown mud. Contains oyster shells and lime nodules. 400-415 Brown limestone Recrystallized reef. At 405 ft 3 inches of brown mud mixed with coral fragments. 415-418 Brown marl and limestone Fragments of recrystallized reef mixed with brown mud. 418-420 Brown and gray marl Fragments of limestone mixed with brown and gray clay. 420-425 Gray limestone Large fragments of limestone, probably recrystallized reef, oyster shells, etc., mixed with some gray and brown mud between 420 and 423 ft. At 423-425 ft large sections of limestone. 425-428 Gray-brown marl 428-438 Brown mud (soil) Plastic brown silty clay. 438-447 Gray marl At 443 ft a 4-inch layer of red soil. 447-462 Brown mud (soil) Stratified; soil structure contains well weathered basalt pebbles. 462-486 Brown muddy sand At 467-469 ft and at other depths nearly pure coarse medium basaltic sand; remainder of section muddy sand; at 483-486 ft no sand, just brown mud. 486-499 Brown gray marl Oyster shells abundant. The core from 487 to 498 ft was lost in a drilling mishap, but the cores at 487 and 498 ft were of the same lithology, and so possibly the missing 11 ft are also brown gray marl. 499-504 Brown silty mud 504-517 Red basaltic, Numerous highly weathered basalt cobbles; e.g., at 508, 510, residual soil 511, 512 ft, all about 4-6 inches. At 513 to 514.5 ft one large pahoehoe boulder IV2 ft in diameter was cored. From about 515 ft the soil grades imperceptibly into soft weathered aa basalt. 517-535 aa basalt At 517 ft rotten aa basalt; no distinct upper surface; grades continually into soil above. From about 518 ft blue weathered aa basalt; large elongated vesicles. 535-542 aa clinker (soil) ? Weathered aa clinker, lower portion highly weathered into soil structures. 542-544 aa basalt Blue; fractured, somewhat weathered aa basalt. 164 PACIFIC SCIENCE, Vol. XXI, April 1967 surface, the coral reef facies shifted landward, allowing the accumulation of a thick coralline limestone reef. 9. The growth of this reef was followed by a progradation of the lagoonal facies during which time nearly 100 ft of brown lagoonal mud accumulated. The upper surface of this mud is capped by several feet of bedded beach rock, indicating a still stand at — 203 ft msl. 10. Above the beach rock is another reef limestone section indicating a migration of coral facies landward again. After a short time this trend reversed itself, for at — 160 ft msl brown mud and soil occur, indicating a pro- gradation of the lagoonal facies. Both of the above two unconformities may correspond with the Penguin Bank stand of the sea. 11. Above this level the coral reef facies advanced inland and dominated the remaining portion of the core except for one soil horizon at — 38 ft MSL that may correspond to the Waipio stand of the sea. 12. The reef making up the present surface of the Ewa Plain appears to belong to the Waimanalo -|- 2 5 -ft stand of the sea, inasmuch as reef limestone can be traced from Ewa No. 1 to Ewa No. 2 where it overlies lagoonal muds. The surface of this reef probably has been eroded by the sea as it retreated from the -j- 25-ft level to the last glacial low stand. CURRENT RESEARCH AND FUTURE PLANS It will be several years before the Ewa cores have been thoroughly examined. Even then the cores will continue to be used for comparison with cores obtained elsewhere. The interpreta- tion of results of the preliminary core examina- tion, especially of those sections dealing with the geologic history, doubtless will be modified as the research proceeds. The cores are stored at the Hawaii Institute of Geophysics, Univer- sity of Hawaii, Honolulu. The present plans for the examination of the Ewa cores include palaeontologic, mineralogic, and chemical analyses of selected samples along the core. Specifically the following types of studies are currently underway: 1. Soil analyses. Chemical and mineralogic studies of samples from suspected soil hori- zons; climatic and other environmental inter- pretations of proven soils. 2. Geochemical analyses. Absolute dating of various horizons within the cores by means of radioactive decay of certain elements. Methods used will include K-Ar, C14, and a new helium method. Paleo-temperature measurements will also be made. 3. Palaeontologic analyses. Macro- and rnicro- palaeontologic studies of fauna and flora; de- terminations of geologic age by the use of these fauna and flora; paleoecologic studies. 4. Sedimentologic analyses. Textural studies of the sediments and sedimentary rock; studies of the sedimentary environments. 5. Mineralogic and petrologic analyses. Optical, chemical, and X-ray determinations of minerals and rocks, including analyses of the underlying basalts. The results of these analyses will give a partial answer to such questions as the tectonic history of the Hawaiian Archipelago and the nature and magnitude of the eustatic changes in sea level recorded in the Ewa cores. However, additional information will be necessary before the complete stratigraphic and paleoecologic history can be unravelled. Most of the sedi- ments in the Ewa cores indicate either a lagoonal or back reef environment. A very ex- tensive barrier reef undoubtedly lay to seaward of the present Ewa sites throughout most of the geologic period recorded in the cores. Without cores through this reef the stratigraphic inter- pretation of the present Ewa cores is handi- capped. Consequently, plans are underway for a research program to drill two more holes in the Ewa area, both offshore and in the area of the anticipated barrier reef. The first hole would be drilled in about 400 ft of water three miles off the beach along the 158th meridian. This deeper hole would be drilled over the 1,800-ft shelf, possibly Miocene in age (Men- ard et al., 1962). Sediment thickness in this area is about 2,000 ft, based on seismic work. With the complete cores recovered from these two offshore holes it would be possible to trace completely the sedimentary facies changes from terrestrial to lagoonal to barrier reef both horizontally and vertically throughout the Pleistocene Epoch and possibly the bter Tertiary Period. Deep Cores of Oahu — Stearns and Chamberlain 165 REFERENCES Menard, H. W., E. C. Allison, and J. W. Durham. 1962. A drowned Miocene ter- race in the Hawaiian Islands. Science 138: 896-897. Stearns, H. T. 1966. Geology of the State of Hawaii. Pac. Books, Palo Alto, Calif. 266 pp. and K. N. Vaksvik. 1935. Geology and ground-water resources of the Island of Oahu, Hawaii. Hawaii Div. of Hydrography. 479 pp. 1938. Records of drilled wells on the Island of Oahu, Hawaii. Hawaii Div. of Hydrography. 213 pp. The Larval Development of the Crab, Cyclograpsus cinereus Dana, under Laboratory Conditions1 John D. Costlow, Jr.2 and Elba Fagetti3 Early descriptions of larvae of the Grap- sidae, based largely on material from the plank- ton and frequently limited to the first zoeal stage, suggested considerable uniformity in the morphological characteristics of larvae of this group of crabs (Hyman, 1924). Subsequent descriptions, based on plankton material as well as on material obtained from rearing the larvae in the laboratory, have served to point out cer- tain differences which do exist in the larvae of this group (Aikawa, 1929; Hart, 1935). To date, however, a very limited number of larvae of crabs belonging to the family Grapsidae have been described. Within the subfamily Sesarminae larvae of two species of the genus Sesarma have been de- scribed from rearing under laboratory conditions (Costlow and Bookhout, I960, 1962). These two species, S. cinereum and S. reticulatum , while relatively common along the east coast of North America and even extending as far as Venezuela, are not known from the west coast of North or South America. On the west coast of Chile, crabs of the sub- family Sesarminae are limited to two species of Cyclograpsus : C. cinereus Dana and C. punc- tatus Milne Edwards (Garth, 1957). The larvae of these species have not been described, either from rearing or from the plankton, and nothing is known about the effect of environmental fac- tors on the development of the larval stages. The present study has had two main objec- tives: one, to rear the larvae of Cyclograpsus cinereus Dana and provide a description of all developmental stages; and two, to determine if salinity and temperature affect the survival and duration of the larval stages. 1 These studies were supported by a grant (GB 1699) from the National Science Foundation and, in part, through the Exchange of Scientists Program of the Pan American Union, Organization of American States. Manuscript received February 9, 1966. 2 Duke University Marine Laboratory, Beaufort, North Carolina. 3 Estacion de Biologia Marina, Vina del Mar, Chile. METHODS Ovigerous Cyclograpsus cinereus females were obtained in the vicinity of the Marine Biological Station, Montemar, University of Chile, and flown by air to the Duke University Marine Laboratory, Beaufort, North Carolina. The crabs were transported in sea water, salinity 34.4 ppt, and extreme temperature changes were avoided by packing them in thermos con- tainers. On arrival at Beaufort, the females were retained at 35 ppt, 20° C. At the time of hatching the larvae were removed, segregated into series of 50 or 100 larvae, and maintained at the temperature-salinity combinations shown in Table 1. Within each temperature-salinity series the larvae were further subdivided, 10 zoeae per bowl, and were fed Artemia nauplii and Arbacia eggs. Each day the larvae were moved to freshly filtered sea water in clean bowls and fresh food was added. At this time the bowls were examined for exuviae, the dead larvae were removed, and the number was re- corded. When the megalops stage was reached the larvae were maintained individually in plas- tic compartmented boxes and fed only Artemia nauplii. From mass cultures the larvae and exuviae were preserved in 5% formalin at known stages of development. Drawings were made to scale with the aid of a camera lucida and the chroma- tophore pattern was determined from living larvae. RESULTS Larval Stages There are five zoeal stages and one megalops stage in the complete development of C. cine- ; reus. The major characteristics of each larval stage are as follows: first zoea (Fig. 1, A-I) : The cephalotho- rax has a gibbose dorsal spine which curves caudally (Fig. 1 , A). The rostral spine is short and the carapace is devoid of lateral spines. 166 Cyclograpsus cinereus — Costlow and Fagetti 167 TABLE 1 Comparison of Original Number of Larvae of Cyclograpsus cinereus, Maintained at Different Combinations of Salinity and Temperature, Survival to Megalops and Crab Stage, and Time Required for Developme nt Through All Larval Stages TIME OF DEVELOPMENT (days) SALINITY (ppt) TEMP. °C ORIGINAL NUMBER PERCENT TO MEGALOPS PERCENT TO CRAB HATCH TO megalops MEGALOPS DURATION HATCH TO CRAB 30 20 100 70.0 45.0 24-33 25.7 15-31 20.3 40-57 45.7 35 20 100 61.0 38.0 25-29 26.0 15-29 22.0 40-56 47.7 30 25 50 30.0 2.0 19-24 20.7 13 30 35 25 50 44.0 10.0 18-21 19.7 14-16 15.0 33-35 34.2 35 30 50 0.0 0.0 — — — * Average figures given in bold type. The eyes are not stalked. The ventrolateral edge of the carapace bears 8 small spines. The abdomen (Fig. 1, B) consists of 5 segments plus the telson. The second abdominal segment has a short lateral hook directed anteriorly. The telson formula is 3 plus 3 spines and the furcal rami are denticulate. The antennule (Fig. 1, C) bears 2 long aesthetes and 2 shorter, unequal setae. The antennal peduncle (Fig. 1, D) is unsegmented and terminates as a large, denticulate spine. The exopodite is shorter and terminates as 2 un- equal spines. The bisegmented endopodite of the maxillule (Fig. 1, F) bears 4 terminal setae, 1 subterminal seta on the second segment and a single seta on the first segment. The basipodite has 5 terminal plumose spines and the coxopodite bears 4 terminal and 1 sub- terminal setae. Four unequal soft, plumose setae fringe the distal border of the scaphognathite of the maxilla (Fig. 1, G), and the apical tip terminates as 1 plumose hair. The bilobate endopodite bears 2 plumose setae on each lobe and its margins are covered with numerous fine hairs. The bilobed coxopodite and basipo- dite bear 5-4 and 2-4 plumose setae respec- tively. The basopodites and exopodites of max- illipeds one and two are as shown in Figure 1, H and 1. The 5 -segmented endopodite of the first maxilliped has a setation of 2, 1,1, 2, 5 (Fig. 1, H) . The 3-segmented endopodite of the second maxilliped (Fig. 1, I) has a setation of 0,1,5. The pattern of chromatophores is consistent for all five zoeal stages. The location of melano- phores in the carapace is as follows: (a) 1, median and dorsal to eyes; (b) 1 pair, ventro- lateral border; (c) 1 pair, median-lateral, poster- ior margin. Melanophores of the zoeal append- ages are as follows: 1 at basis of the antenna; 1 in mandible; 1 in labrum; and 1 in basipodite of first maxilliped. In the abdominal segments 1 melanophore is found dorsal to the gut in segments one through four, 1 in the midventral region of segments two through five (in first and second zoea), and through six (in third to fifth zoea), and 1 pair in the posterior-lateral corner of segment five. SECOND ZOEA (Fig. 2, A-I) : The eyes are stalked, the dorsal spine is no longer gibbose, and the lateral spines are present (Fig. 2, A). The ventrolateral borders of the carapace bear 6 small spines plus 6 setae and there are 2 spines on the posterior, middorsal margin. Antennule now bears 4 aesthetes, approximately equal in length (Fig. 2, C). Changes in seta- tion of the appendages are limited to the fol- lowing: basipodite of maxillule has 5 spines 168 PACIFIC SCIENCE, Vol. XXI, April 1967 Fig. 1. Side view (A) of first zoeal stage of Cyclograpsus cinereus Dana. B, Ventral view of abdomen; C, antennule ; D, antenna; E, mandible; F, maxillule ; G, maxilla; H, first maxilliped; I, second maxilliped. plus 2 setae and a spine projects from the pro- topodite (Fig. 2, F). Scaphognathite of maxilla (Fig. 2, G) has 5 proximal and 3 distal plu- mose setae. Setation of the exopodite of the first and second maxillipeds has increased to 6 (Fig. 2, H and /). Setation of the endopodite of the first maxilliped is now 2, 2, 1,2, 5, and of the second maxilliped, 0,1,6. third zoea (Fig. 3, A-I) : Small spines on the ventrolateral borders of the carapace have decreased to 5 and setae increased to 10 (Fig. 3, A). The sixth abdominal segment is added and the number of telson spines has increased to 8 on the inner surface (Fig. 3, B) . Changes in setation of the appendages are limited to the following: 1 small spine is added on the lateral margin of basopodite of maxillule (Fig. 3, F) ; scaphognathite of maxilla has 8 proximal and 6 distal plumose setae and each lobe of baso- podite bears 5 setae (Fig. 3, G) ; 1 seta has been added to lateral margin of coxopodite of maxilla (Fig. 3, G). Setation of exopodite of first and second maxillipeds has increased to 8 (Fig. 3, H and I). Setation of endopodite of first maxilliped is now 2, 2, 2, 2, 5. fourth zoea (Fig. 4, A-I) : Small spines on ventrolateral borders of carapace decreased to 4 and setae increased to 14. Rudiments of pereiopods present under carapace (Fig. 4, A). Dorsal surface of first abdominal segment now Cyclograpsus cinereus — Costlow and Fagetti 169 Fig. 2. Side view of (A) of second zoeal stage of Cyclograpsus cinereus Dana. B, Ventral view of abdo- men; C, antennule ; D, antenna; E, mandible; F, maxillule; G, maxilla; H, first maxilliped; I, second maxil- liped. bears 5 spines and pleopod buds are present on segments two through six (Fig. 4, B) . Setation of the appendages is as follows : antennule (Fig. 4, C) bears 5 terminal and 1 subterminal aesthetes; the unsegmented endopodite bud of the antenna (Fig. 4, D) is approximately half the length of the antennal spine. Setation of basopodite of maxillule has increased to 11 (Fig. 4, F). Scaphognathite of maxilla (Fig. 4, G) has 16 proximal and 9 distal plumose setae and setation of coxopodite has increased to 9. Setation of exopodite of first and second maxil- lipeds (Fig. 4, H-I) has increased to 10. Setation of endopodite of first maxilliped is now 2, 2, 2, 2, 6 (Fig. 4, 7). fifth zoea (Fig. 5, A-I) : Minute spines on ventrolateral borders of carapace increased to approximately 10, and 18 setae are present (Fig. 5, A). Rudiments of unsegmented pereiopods now visible under carapace. Pleopods of abdom- inal segments two through six partially seg- mented (Fig. 5, B ). Changes in setation of the appendages are as follows: swollen basal region of antennule (Fig. 5, C) bears 2 small, plu- mose setae, endopodite bud is present and sub- terminal aesthetes have increased to 4. Endo- podite of antenna (Fig. 5, D) partially seg- mented and longer than exopodite. Setation of basopodite of maxillule (Fig. 5, F ) increased to 13 and coxopodite now bears 9 setae. 170 PACIFIC SCIENCE, VoL XXI, April 1967 Fig. 3. Side view (A) of third zoeal stage of Cyclograpsus cinereus Dana. B, Ventral view of abdomen; C, antennule; D, antenna; E, mandible; F, maxillule; G, maxilla; H, first maxilliped; I, second maxilliped. Cyclograpsus cinereus — Costlow and Fagetti 171 Fig. 4. Side view (A) of fourth zoeal stage of Cyclograpsus cinereus Dana. B, Ventral view of abdomen; C, antennule; D, antenna; E, mandible; F, maxillule; G, maxilla; H, first maxilliped; l, second maxilliped. 172 PACIFIC SCIENCE, VoL XXI, April 1967 Fig. 5. Side view (A) of fifth zoeal stage of Cyclograpsus cinereus Dana. B, Ventral view of abdomen; C, antennule ; D, antenna; E, mandible; F, maxillule; G. maxilla; H, first maxilliped ; I, second maxilliped. Cyclograpsus cinereus — Costlow and Fagetti 173 Fig. 6. Dorsal view (A) of megalops of Cyclograpsus cinereus Dana. B, Telson; C, antennule; D, an- j tenna; E, mandible; F, maxillule; G, maxilla; H, first maxilliped; I, second maxilliped; J, third maxilliped; I K, pleopods. 174 PACIFIC SCIENCE, Vol. XXI, April 1967 Scaphognathite of maxilla (Fig. 5, G) has ap- proximately 30 proximal and 12 distal plumose setae, setation of basopodite increased to ap- proximately 16, and the coxopodite now bears 1 1 setae. Setation of exopodite of first and second maxillipeds (Fig. 5, H—l) has increased to 12. Setation of endopodite of first maxilliped is now 2, 3, 2, 2, 6. megalops (Fig. 6, A-K) : Cephalothorax without rostral spines and provided with hairs at lateral edges (Fig. 6, A). Abdomen 6-seg- mented, with telson (Fig. 6, B) bearing 4 dorsal and 3 terminal plumose setae. Antennule (Fig. 6, C) with 2 flagella on second segment. Unsegmented flagellum with 3 terminal setae and 1 subterminal seta. The 4-segmented flagel- lum bears 3 aesthetes on the second segment, 4 aesthetes plus 1 seta on the third segment, and 5 aesthetes on the terminal segment. Antenna (Fig. 6, D) with 11 segments and setation as shown in Figure 6, D. Mandible (Fig. 6, E ) with 2 -segmented palp bearing 9 plumose setae on terminal segment. Unsegmented endopodite of maxillule (Fig. 6, F) with 2 terminal and 4 lateral setae. Basopodite with approximately 21 spines and setae and the coxopodite with 6 terminal and 5 lateral setae. Scaphognathite of maxilla (Fig. 6, G ) fringed with approximately 70 plumose setae (Fig. 6, G). Unsegmented endopodite bearing 2 unequal terminal setae and 1 subterminal plumose seta. Lobes of baso- podite bearing 13 and 11 plumose setae, respec- tively, and lobes of coxopodite with 6 and 12 setae. The first, second, and third maxillipeds are as shown in Figure 6, H, I, and /. Setation of exopodites of pleopods of abdominal seg- ments two to five varies from 17 to 20. Endo- podites of all pleopods have 3 median hooks (Fig. 6, K ). Uropods bisegmented with 1 seta on first segment and 10 setae on second segment. Chromatophore pattern of carapace as fol- lows: one, dorsomedian surface, one on each median-lateral border, a pair on ventral surface of rostrum, joined at the median line, and one at each posteriolateral corner. Chromatophore pattern of appendages: one on each eyestalk, one on basopodite of antennule, one on basopo- dite of cheliped, mandibles, and labrum. With- in the abdominal segments melanophores are located as follows: one in first segment, dorsal to gut, extending into cephalothorax; one at posterioventral margin of segments two through five, the chromatophore of segment five ex- tending into proximal region of segment six. MOLT I MOLT H 30ppt. 35Ppt. 3°ppl- 35Ppb 1 MOLT in MOLT Et MOLT 3 Z 30ppt. 35ppt. 30ppt. 35ppf. 30ppt. 35ppt. PERCENT MOLTS — roto^LnO^ODioc O.Oo OoO OOOOC 1 1 ulil i l i .l . .r i: l O cO fO i/"> lO r- O— OJ O— OJ COOlO — C\J aOO'O O <0 rf vT» tO f-* 00 o> O — (\J ^ 0*0— OJf*W iD COCT) 0 — — — CVJ CVJ’CM OJ OJ (\J OJ OJ OJ OJ fO fOfOfr> — OJOJOJ OJOJ OJ OJOJOJ OJ 0J(\1 CYCLOGRAPSUS CINEREUS DAYS OF DEVELOPMENT ■ 20* c □ 25* C Fig. 7. Comparison of the time of larval molts for zoeae and megalops of Cyclograpsus cinereus Dana reared in the laboratory at different salinities and temperatures. Black , 20°C ; white, 25°C Cyclograpsus cinereus — Costlow and Fagetti 175 Larval Development The frequency of molting during the five zoeal and one megalops stages is shown in Figure 7. The first and second molts of most larvae were confined to two days at all four salinity-temperature combinations. By the time of the third molt, however, the uniformity was not as apparent and the later molts were spread over three, four, and five days. Figure 8 shows the time required for development of the five zoeal stages and megalops of C. cinereus when maintained at combinations of two salinities and two temperatures. At 20° C, the duration of all larval stages is longer than that observed for larvae reared at 25 °C This difference is more pronounced as development continues and is quite apparent for the megalops stage. Within the two salinities used, 30 and 35 ppt, there is no apparent difference in the time required for development of the larval stages (Fig. 8). The average time for total develop- ment to the first crab (Fig. 8) at 25 °C was 30 days at 30 ppt and 34.2 days at 35 ppt. At 20°C and 30 ppt, 45.7 days were required and at the same temperature and 35 ppt comparable development required 47.7 days. Mortality of the individual larval stages is shown in Figure 9. Although larvae were main- 50 45 40 35 30 30 ppt. 35ppt . £ 25 in Q O 20 co > < ° 15 10 TEMPERATURES. Fig. 8. Comparison of the duration of five zoeal stages and megalops of Cyclograpsus cinereus Dana reared in the laboratory at different salinities and temperatures. G2 meg. £2 st. 5Z H st. nz: □ st. n r Dffl st. it H st. i Fig. 9- Comparison of the mortality of the five zoeal stages and megalops stage of Cyclograpsus cinereus Dana reared in the laboratory at different salinities and temperatures. 176 PACIFIC SCIENCE, Vol. XXI, April 1967 tained at 30 and 35 ppt, 30°C, all died within the first two days following hatching and are not included in Figure 9. Mortality of most larval stages was higher at 25 °C than at 20 °C Mortality of any one zoeal stage at 20 °C does not exceed 16%, while at 25 °C mortality with- in one stage (Stage V) was as high as 36%. At 25 °C, there was a tendency for the mortality to be higher in the later larval stages than in the early zoeal stages. Survival to the crab ranged from 2 to 10% at 25 °C and from 38 to 45% at 20°C. DISCUSSION Larval Stages A comparison of the larvae among species of the same subfamily as Cyclograpsus cinereus is limited to descriptions of Sesarma cinereum (Hyman, 1924; Costlow and Bookhout, I960), Sesarma reticulatum (Hyman, 1924; Costlow and Bookhout, 1962) and Sesarma picta (Aikawa, 1937). The number of zoeal stages varies considerably within those species of Sesarminae which have been described. C. cinereus has five zoeal stages, S. cinereum has four zoea (Costlow and Bookhout, I960) and S. reticulatum has three zoeal stages (Costlow and Bookhout, 1962). Aikawa (1937), while describing the first stage zoea of S. picta from the plankton, does not indicate the total num- ber of larval stages of this species. The first zoea of C. cinereus is quite distinctive in that the dorsal spine is gibbose. In all subsequent zoeal stages, when the dorsal spine is straight only C. cinereus has lateral spines. All the larval stages of C. cinereus may be further differen- tiated from zoeae of the three species of Sesarma by the absence of a knob on each lateral surface of the third abdominal segment of larvae of C. cinereus. Among the Chilean species of the family Grapsidae the larvae of only one, Grapsus grapsus L., have been described to date (Aikawa, 1937). Only the first zoea was de- scribed and may be differentiated from the first zoeal stage of C. cinereus by the presence of the lateral knob on the third abdominal seg- ment of G. grapsus as well as by the gibbose dorsal spine of C. cinereus. The megalops of C. cinereus differs from the megalops of S. cinereum and S. reticulatum in several respects. The main differences can be observed in the rostrum, telson, antennule, and second maxilliped. The rostrum of C. cinereus megalops is not depressed as in S. cinereum (Costlow and Bookhout, I960), and does not have a rostral spine as in S. reticulatum (Cost- low and Bookhout, 1962). The telson of C. cinereus bears only 3 setae on the distal margin, while the telsons of S. cinereum and S. reticu- latum bear 8 setae and 6 setae, respectively, plus lateral spines. In C. cinereus the unsegmented flagellum of the antennule is present, while in S. cinereum and S. reticulattim the unsegmented flagellum is absent and replaced by a single seta. The epipodite of the second maxilliped is present in the megalops of C. cinereus but absent in S. cinereum and in S. reticulatum. Of all the other megalops of the family Grapsidae which have been described to date, none are found in Chilean waters. However, it should be noted that the megalops of C. cinereus bears a greater resemblance to mega- lops of Hemi grapsus nudus and H. ore gone sis, which belong to the subfamily Varuninae and were described from the Pacific coast of Canada by Hart (1935), than to megalops of the sub- family Sesarminae. A more detailed comparison of the mor- phology of larvae of C. cinereus with larvae of other closely related forms in Chilean waters must await additional descriptions. Larval Development The adults of C. cinereus are normally con- fined to the area from Ancon, Peru to Calbuco, Chile on the western coast of South America (Garth, 1957). One extra limital locality, Panama, has also been recorded (Rathbun, 1910, 1918). The habitat of the adults is in the upper level of the intertidal region, where they live under stones in the coarse sand. Ovigerous females have been observed in the Montemar region throughout the year with the exception of February. The principal spawning period, however, appears to be from July through November, when more than 60% of the population is ovigerous. During this period the water temperature increases from 12°C to 14°C. The salinity of the water in which larval development occurs is quite stable, ranging Cyclograpsus cinereus — Costlow and Fagetti 177 from 34.1 ppt to 34.5 ppt (Antezana, Fagetti, and Lopez, 1965). In the experimental conditions of the labora- tory, duration of the five zoeal stages and one megalops stage appears to be relatively un- affected by the limited range of salinity used. The larvae did develop to the crab faster at 25 °C than at 20 °C, as would have been ex- pected. Survival, however, was consistently higher at 20°C than at 25 °C or at 30°C. The larval development of certain other species of Brachyura, normally considered to be estuarine, has been shown to be directly affected by salinity (Costlow and Bookhout, 1962 ; Costlow, Book- out, and Monroe, I960, 1966). The results of the present study, however, suggest that the development of larvae of C. cinereus is not strongly influenced by the relatively small salin- ity fluctuations to which the larvae would be subjected during their planktonic existence in the waters off the western coast of Chile and Peru. SUMMARY AND CONCLUSIONS The larval stages of Cyclograpsus cinereus Dana have been reared in the laboratory from hatching to the first-stage crab. The larvae were maintained in combinations of three tempera- tures, 20°C, 25 °C, and 30°C, and two salini- ties, 30 ppt and 35 ppt, and were fed recently hatched Artemia nauplii and fertilized Arbacia eggs- There are five zoeal stages and one msgalops under laboratory conditions. The larvae, as well as the setation of the functional appendages, have been described and figured. Descriptions of larvae of closely related species from Chilean waters are not available, but the larvae of C. cinereus Dana can be differentiated from the other grapsid larvae described to date. Approximately 46 days were required for development to the crab at 20°C At 25°C, development was completed in 30 to 34 days. Larvae completed development to the first crab in salinity-temperature combinations other than 35 ppt, 30 °C. A higher percentage of the larvae survived at 20°C than at 25 °C. Survival at 30 ppt and 35 ppt was simliar, suggesting that development under natural conditions is not affected by minor fluctuations in salinity. REFERENCES Aikawa, H. 1937. Further notes on Brachyura larvae. Rec. Oceanogr. Wks. 9:87-162. Antezana, T., E. Fagetti, and M. T. Lopez. 1965. Observaciones bioecologicas en De- capodos comunes de Valparaiso. Montemar 12(1) :l-60. Costlow, J. D., Jr., and C. G. Bookhout. i960. The complete larval development of Sesarma cinereum (Bose) reared in the laboratory. Biol. Bull. 118(2) :203-2l4. — - 1962. The larval development of Sesarma reticulatum Say reared in the laboratory. Crustaceana 4(4) : 281-294. and R. Monroe, i960. The effect of salinity and temperature on larval development of Sesarma cinereum (Bose) reared in the laboratory. Biol. Bull. 118: 183-202. — 1966. Studies on the larval development of the crab, Rhithropano- peus harrisii (Gould). I. The effect of salin- ity and temperature on larval development. Physiol. Zool. (In press.) Garth, J. S. 1957. The Crustacea Decapoda Brachyura of Chile. Report of the Lund University Chile Expedition, 1948-1949. Lund Univ. Arsskr. N.F. 53(7) :1-130. Hart, J. 1935. The larval development of British Columbia Brachyura. I. Xanthidae, Pinnotheridae and Grapsidae. Canadian J. Res. 12:411-432. Hyman, O. W. 1924. Studies on larvae of crabs of the family Grapsidae. Proc. U. S. Natl. Mus. 65, Art. 10:1-8. Rathbun, M. J. 1910. The stalk-eyed Crus- tacea of Peru and adjacent coast. Proc. U. S. Natl. Mus. 38:531-620. 1918. The grapsoid crabs of America. U. S. Natl. Mus. Bull. 97:1-461. A Comparison of Euphausiid Shrimp Collections Made with a Micronekton Net and a One-Meter Plankton Net1 Charles W. Jerde In an evaluation of variable factors affecting the apparent geographic range and estimated abundances of euphausiids, Brinton (1962) compared euphausiid catching ability of a 1-m diameter net, made principally of 0.65 mm mesh, with a 45 -cm diameter net made of 0.33 mm mesh. He found that adult and juvenile euphausiids were taken by the larger net in numbers as great or greater than were obtained with the 45-cm net, but that only about half as many larvae were taken with the coarser meshed meter net as with the 45-cm net. Collections with the 45-cm net contained almost as many species as the collections with the 1-m net, which filtered a volume of water 5 times as great (Brinton, 1962). On Scripps Tuna Oceanography Research cruises 64-1 and 64-2 (off southern Baja Cali- fornia) an attempt was made to sample con- secutively to the same depth with a micronekton net and a 1-m diameter plankton net, in order to compare euphausiid catches between the two nets. This paper is an evaluation of the euphau- siid catching ability of the two nets. The author is indebted to Dr. Edward Brin- ton for his assistance in the identification of the euphausiids. The constructive advice of Dr. Maurice Blackburn, Dr. E. W. Fager, Dr. Mil- ner B. Schaefer, and Dr. Paul Smith was grate- fully received. METHODS A description and figures of the micronekton net are found in Blackburn and associates (1962) ; the net with a 2.3 m2 mouth opening 1 This work formed part of the Scripps Tuna Oceanography Research Program of the Institute of Marine Resources and Scripps Institution of Oceanog- raphy, University of California. Partial support was provided by the U. S. Bureau of Commercial Fish- eries under Contract 14-17-0007-306. Manuscript re- ceived June 20, 1966. Contribution from the Scripps Institution of Oceanography, University of California, San Diego. is made of nylon netting of uniform mesh (apertures measuring about 5.5 mm by 2.5 mm) throughout and has a detachable cod end of #5 6 XXX grit gauze (mesh aperture 0.31 mm). The micronekton net was towed in oblique hauls, from an average depth of 131 m to the surface at 5 knots for an average period of 50 minutes; depth of haul was deter- mined by a bathythermograph attached to the upper edge of the square mouth opening (1.5 m by 1.5 m). A flow meter was not used with the micronekton net, and volume of water fil- tered was estimated from size of mouth opening, ship speed, duration of tow, and a filtration coefficient of 0.757 which had been determined by Blackburn (MS). Estimated volume of water filtered per tow with the micronekton net ranged from 14,000 to 16,000 ms. The 1-m net (Ahlstrom, 1948) has a mouth opening of 0.785 m2 and is made of #30 XXX grit gauze (mesh 0.65 mm) in the forward sec- tion of the net, with #56 XXX grit gauze (mesh 0.31 mm) in the rear section and cod end. It was towed in oblique hauls, from an average depth of 133 m to the surface at 1-2 knots for an average period of 14 minutes. Maximum depth of haul of the 1-m net was j estimated from the amount of wire out and the wire angle ; a calibrated flow meter placed at the center of the mouth opening was used to esti- mate volume of water filtered, which ranged from 385 to 468 m3. On the average, the micro- nekton net filtered 34.4 times as much water as the meter net at each station. Euphausiids were picked from the entire col- lection of each tow at 10 stations. "Wet” dis- placement volume of each entire euphausiid sample was determined according to the method of Ahlstrom and Thrailkill (1963). All euphau- siids in these plankton samples were counted, with the exception of those in cruise 64-2 col- lections at stations 4l and 56; from these two collections, aliquots of 1/2 and *4, respectively, 178 Comparison of Euphausiid Collections — Jerde were counted. Because the samples taken with the micronekton net were very large, it was necessary to use aliquots in all cases; these ali- quots ranged from 2.75% to 50% depending upon the size of the sample. A Folsom plankton splitter (McEwen, Johnson, and Folsom, 1954) was used for fractionating the samples, with the exception of the micronekton sample at station 41. For this sample the animals in a gallon jar were kept in suspension by agitation, and a portion of animals and fluid was poured out; "wet” displacement volume of the animals was determined, and subsequently the euphausiids were measured and counted. After the actual catch of euphausiids was estimated, the numbers were standardized for each size category to numbers per 500 m3 of water (Table 1). Blackburn (MS) estimates that the amount of water actually filtered by the micronekton net at a speed of 5 knots, using the above mentioned filtration coefficient, is 1000 m3 per 3.69 minutes. In this study micronekton standardized volumes, or numbers, per 500 m3 were calculated by the following formula: actual vol. or number ml or number/500 m3 = X 1-85 number of minutes Brinton (1962) has denoted as plentiful species those which occur in concentrations greater than about 25 specimens per 1000 m3 of water. Of the euphausiid species which mature at 7> 9 mm, only one, Euphausia eximia, was plentiful in the 64-1 and 64-2 collections, and this species was the predominant euphausiid in the samples. For each collection, in the por- tion of the sample counted, the length of each E. eximia was measured to the nearest mm, from the tip of the rostrum to the tip of the telson. In addition to other station data, the percentage of each sample which was counted and measured is noted in Table 1. Excluding station 41, the remainder of each sample was scanned under the microscope for rare species. DISCUSSION AND SUMMARY Wilcoxon’s signed-rank test (Tate and Gel- land, 1957), a nonparametric statistical method, was employed to test for differences in euphau- siid catching ability between the two nets. The data in Table 1 indicate that there is no sig- 179 nificant difference between the nets with respect to estimated volume of total euphausiids per 500 m3. However, it is clear that the nets differ with regard to ability to catch different species and ontogenetic stages. It is evident that the micronekton net does not quantitatively sample larval or juvenile Euphausia eximia , and that those animals which are less than 13 mm long escape readily through the larger mesh. In the size range 13-21 mm there appears to be no significant difference in number of E. eximia per 500 m3, but there may be such a difference in the 22-28 mm size range; the micronekton net appears to catch more euphausiids in this size range than does the 1-rn net. This difference in the 22-28 mm category may be interpreted as evidence of avoidance of the 1-m net by the larger euphausiids. However, when all adults (13-28 mm) are grouped together there is no significant difference between the nets with re- gard to the estimated density of E. eximia. Evi- dence of avoidance of towed nets by zooplank- ton has been presented by Fleminger and Clutter (1965). In terms of the number of euphausiid species found at a station, there was no significant dif- ference between the two nets when adults alone were considered (Table 1). When larvae and juveniles, as well as adults, were used to deter- mine the total number of species present at a station, there was a significant difference be- tween the catches of the two nets. The 1-m net caught more euphausiid species than the micro- nekton net, because it retained more larvae and juveniles than the micronekton net (Table 1) and also retained more adults of the smaller species (adult at <9 mm in length, Table 2). Table 2 shows a comparison of the two nets with respect to presence or absence of adults of different euphausiid species at nine stations. For the larger species (adult at 7> 9 mm) the micro- nekton net as a sampling device is as good as or better than the 1-m net with regard to pres- ence or absence of species (Table 2). Of the smaller species, with the exception of E. dis- tinguenda (Table 2), presence of adults was observed more often in the 1-m net than in the micronekton net. Thus, for qualitative euphau- siid studies, the 1-m net provides almost as much or more information for one-third of the ship time. EUPHAUSIID CATCHING ABILITY OF A MICRONEKTON NET COMPARED WITH A ONE-METER 180 PACIFIC SCIENCE, Vol. XXI, April 1967 8 E ooin^oj»-tO'!^Noovooocx)cn^’^oiOsfrjc^ CO CM inrHCOiHrHCn'Ocn^CMCO^^'HCOiH VO CO Comparison of Euphausiid Collections — Jerde 181 TABLE 2 Comparison of the Micronekton Net and One-Meter Net with Respect to Presence or Absence of Adults of Euphausiid Species at Nine Stations NUMBER OF STATIONS WHERE ADULTS WERE COLLECTED SPECIES IN METER NET ONLY IN MICRONEKTON NET ONLY IN BOTH NETS IN NEITHER NET Large species (adult at ^ 9 mm) Euphausia eximia — — 9 — Euphausia gibboides — — 5 4 N ematobrachion flexipes — 5 3 1 Nematoscelis difficilis — 4 4 1 Nematoscelis gracilis 1 3 — 5 Nyctiphanes simplex — 2 2 5 Small species (adult at < 9 mm) Euphausia diomedeae 1 — — 8 Euphausia distinguenda 1 3 1 4 Euphausia mutica 1 — 1 7 Euphausia recurva 1 - — 1 7 Euphausia tenera 1 • — 1 7 Stylucheiron affine 6 — 1 2 Stylocheiron longicorne 2 — — 7 Thysanoessa gregaria 1 — — 8 REFERENCES Ahlstrom, E. H. 1948. A record of pilchard eggs and larvae collected during surveys made in 1939 to 1941. U. S. Fish and Wildlife Serv. Spec. Sci. Rept. 54:1-76. and J. R. Thrailkill. 1963. Plankton volume loss with time of preservation. Calif. Coop. Oceanic Fish. Invest. Rept. 9:57-73. Blackburn, M. (MS). Micronekton of the eastern tropical Pacific Ocean. and associates. 1962. Tuna oceanog- raphy in the eastern tropical Pacific. U. S. Fish and Wildlife Serv. Spec. Sci. Rept. Fish. 400:1-48. Brinton, E. 1962. Variable factors affecting the apparent range and estimated concentra- tion of euphausiids in the North Pacific. Pacif. Sci. 16(4) :374-408. Fleminger, A., and R. I. Clutter. 1965. Avoidance of towed nets by zooplankton. Limnol. Oceanog. 10(1) :96-104. McEwen, G. F., M. W. Johnson, and T. R. Folsom. 1954. A statistical analysis of the performance of the Folsom plankton splitter, based upon test observations. Arch. f. Mete- orol. Geophysik Bioklimatol., ser. A, vol. 7. Tate, M. W., and R. C. Clelland. 1957. Nonparametric and Shortcut Statistics. Inter- state Printers and Publishers, Danville, Illi- nois. A New Genus and Two New Species in the Families Volutidae and Turbinellidae (Mollusca: Gastropoda) from the Western Pacific Harald A. Rehder1 ABSTRACT : Sigaluta pratasensis, new genus, new species, in the family Volutidae is described from the South China Sea, off Hong Kong. Phenacoptygma Dali, 1918 is removed from the Volutidae and placed in the synonymy of Surculina Dali, 1908, which in turn is removed from the Turridae and assigned to the Turbinellidae near Benthovoluta Kuroda and Habe, 1950, on the basis of its radula. It is pro- posed that the families Turbinellidae ( olim Xancidae) and Vasidae be of co- ordinate rank. A new species of Benthovoluta , B. gracilior, is described from the Sulu Sea, Philippines. In the process of arranging the specimens of the family Volutidae in the Division of Mol- lusks, U. S. National Museum, two new species of deepwater mollusks from the western Pacific were found. These were dredged by the U. S. Bureau of Fisheries steamer "Albatross I” in the South China and Sulu Seas during her 1907-09 cruise in the Philippine Islands. One of these species turns out to belong to the genus Benthovoluta, recently placed by Kuroda (1965:50-51) in the family Turbinel- lidae. For the other a new genus in the family Volutidae must be erected. VOLUTIDAE Sigaluta,2 new genus Shell moderately large, ovate with only few whorls (about 4) ; nuclear whorls large; shiny with glazelike surface. Aperture ovate; outer lip slightly flaring and somewhat thickened, with shallow rounded sinus at junction with body whorl ; columella straight, bearing 2 strongly ascending spiral folds. type species: Sigaluta pratasensis, new species This interesting new genus is represented in our collection by only two shells, and as the 1 Division of Mollusks, U. S. National Museum, Smithsonian Institution, Washington, D. C. Manu- script received March 9, 1966. 2 From the Greek sigaleios (glossy) + Voluta. soft parts unfortunately were not retained, the exact allocation of the genus must await the discovery of fresh living material. On the basis of the general appearance of the shell, nucleus, and columella plaits, I am placing this genus temporarily in the subfamily Cymbiinae, tribe Meloides, as defined by Pilsbry and Olsson (1954:16-17). Sigaluta pratasensis, new species Figs. 1-4 description: Shell of moderate size (54- 61 mm, 2 ^-2 1/2 inches long) and solidity, narrowly ovate; nuclear whorls 2%, large, bul- bous, smooth, shining; transition between nuclear and postnuclear whorls marked by faint line of demarcation and slight increase in diam- eter of first postnuclear whorl ; postnuclear whorls l%-2, smooth and shining as if glazed; body whorl rather strongly descending on penultimate whorl; suture glazed over; outer lip slightly thickened with a whitish callus and marked with a shallow sigmoid sinus below suture and a broad, very shallow sinus at obliquely truncate base. Aperture narrowly ovate, acuminate at top, broadly truncate at base ; columella straight, bearing 2 strongly ascending spiral plaits. Color from light yellow- brown (#76) (with a slightly grayish cast) to light-gray olive-brown (#94) (iscc-nbs Color Names, Kelly and Judd, 1965). locality: West of Pratas Reef, South China Sea, in 208 fathoms (380 m) ; U.S.B.F. "Alba- 182 New Gastropoda from Western Pacific — Rehder 183 tross I” Sta. 5301, 20° 37' N, 115° 43' E, gray mud and sand bottom. August 8, 1908. measurements: SPECIMEN LENGTH WIDTH Holotype (usnm 237018) 53.9 mm 28.5 mm Paratype (usnm 637251) 60.95 mm 29.6 mm Turbinellidae Swainson, 1840 synonyms: Ptychatractidae Stimpson, 1865; Xancidae Woodring, 1928 Benthovoluta Kuroda and Habe, 1950 Kuroda, T. and T. Habe, 1950:37. Kuroda, T., 1965:50-52. Figs. 1 and 2. Sigaluta pratasensis n. gen., n. sp.; approx. X %• 1, Paratype. 2, Holotype. Fig. 3. Sigaluta pratasensis n. gen., n. sp.; apical view; approx. X %• TYPE SPECIES, BY ORIGINAL DESIGNATION: Phenacoptygma ? kiiensis Kuroda (=V oluta hilgendorp von Martens, 1897). Kuroda (1931:48) described the type species, under the name Phenacoptygma kiiensis, locat- ing it doubtfully in that eastern Pacific genus and suggesting that "Mitra” plicifera Yoko- yama (Yokoyama, 1920:48) from a Pliocene formation near Tokyo Bay was related. In 1950 Kuroda and Habe, in proposing the new genus Benthovoluta, placed kiiensis Kuroda in the synonymy of V oluta hilgendorp von Martens, although designating P. kiiensis as the type species of the genus. At the same time they listed "Mitra” plicifera Yokoyama as an addi- tional synonym of hilgendorp. Judging from the figure and description of the Pliocene plicifera, I would suggest that it represents a distinct species, with more numerous axial ribs on the spire whorls, which are less convex than in hilgendorp. The species described below seems to repre- sent a third species of this genus, although without a knowledge of its soft parts its alloca- tion to this group must be largely speculative. Habe (1952:132) depicted the radulae of a number of Japanese marine mollusks without comments. Among them was a figure of the teeth of Benthovoluta hilgendorp. Kuroda (1965:50-51) called attention to the fact that Habe’s figure is unlike that of any volutid radula and suggested that Benthovoluta be placed in the family Turbinellidae, near Metz- geria Norman, 1879, a monotypic boreal genus. A comparison of Sars’ figure (Sars, 1878, pi. Fig. 4. Sigaluta pratasensis n. gen., n. sp.; view showing columella fold; approx. X %• 184 PACIFIC SCIENCE, Vol. XXI, April 1967 IX, fig. 13; Thiele, 1929:343, fig. 409) of the radula of Metz geria alba (Jeffreys, 1873) (syn. Meyeria pusilla 'M. Sars’ G. O. Sars, 1878) with the figure given by Habe reveals indeed a very close similarity; while the radulae of Turbinella jus us Sowerby, 1825, as published by Dali (1885:346, pi. XIX, fig. 1; Abbott, 1950:202, pi. 89, fig. 2) and of T. laevigata Anton, 1839 (Thiele, 1929:342; Abbott, 1950: 202, pi. 89, fig. 3) also show a similarity, al- though the relationship is less close. A more strikingly close relationship is re- vealed by a study of the radulae of what Dali described as Daphnella ( Surculina ) cortezi (Dali, 1908:292) from off San Diego, Cali- fornia, and TLeucosyrinx galapagana (Dali, 1919:5, pi. 3, fig. 2) from the Galapagos Islands. The type species of Surculina, Daph- nella ( Stirculina ) blanda Dali (1908:291, pi. 3, fig. 1) is certainly congeneric with S. cortezi, for which Dali in 1918 proposed the generic name Phenacoptygma (Dali, 1918:138), plac- ing it in the family Volutidae. This genus was placed in the subfamily Calliotectinae by Pils- bry and Olsson (1954:19). The genus Surculina Dali, 1908 (Dali, 1908:260-261), with its synonym Phenaco- ptygma Dali, 191 8, therefore also must be placed in the family Turbinellidae. Surculina was considered to be a subgenus of Leuco syrinx by Grant and Gale (1931:509- 510), who assigned both blanda and gala- pagana to this subgenus. Powell (1942:21) follows this allocation, placing the group in the subfamily Cochlespirinae. In order to make this relationship more clear, and because the type species of Surculina ap- parently has never been figured and the figures of the other species may not be readily acces- sible to all students, I am illustrating all three species of Surcidina (Figs. 7-9). In addition, I am figuring the radula of S. cortezi (Fig. 10) and, for comparison, that of Benthovoluta hilgendorji (Fig. 11). Another genus that probably belongs here is Ptychatractus Stimpson (1865:59) with three species: the type of the genus, P. ligatus Mighels and Adams, 1842, from the Gulf of Maine; P. occidentalis Stearns, 1873, from Alaska; P. calif ornicus Dali, 1908, from Mon- terey Bay to San Diego, California. A rather poor figure of the radula is given by Stimpson (1865: pi 8, fig. 8). I have used the family name Turbinellidae instead of Xancidae or Vasidae for the follow- ing reason. In 1957, in Opinion 489 of the International Commission on Zoological Nomenclature, the generic name Turbinella Lamarck, 1799 was validated and placed on the Official List of Generic Names, and Xancus Roding, 1798 was suppressed and placed on the Official Index of Rejected and Invalid Names. Concurrently, the family name Turbinellidae Swainson, 1840 was placed on the Official List of Family Group Names in Zoology. A perusal of the history of this case (Hemming, 1957:155-178) reveals the fact that whereas six persons are cited as supporting the use of Turbinella, eight opposed it. Of these eight, six were professional mala- cologists (one a paleontologist) , while three malacologists (only one of them a professional worker) supported the proposal, siding with a botanist, an ichthyologist, and an anthropol- ogist. In spite of the preponderance of opposition against the proposal, and the clear evidence of the very limited use of Turbinella in recent scientific literature, the proposal was approved and Opinion 489 was issued as summarized above. Disturbed by the action of the International Commission in passing a ruling so contrary to the majority of considered opinion, many mala- cologists have refused to follow the recom- mendation, and have continued to use Xancus and Xancidae, apparently hoping for an even- tual reversal of this Opinion. This procedure, however, appears to me to be unwise. If we wish to have any kind of stability in nomen- clature, and if the decisions of the International Commission on Zoological Nomenclature are to have any meaning, we must accept the final decisions of the Commission, particularly as regards names placed on the official lists. What scientific workers must do in the future is to act promptly to prevent decisions by the In- ternational Commission on Zoological Nomen- clature that are contrary to the evidence and majority opinion. In his monograph of several genera of the family Vasidae in the Indo-Pacific, Abbott New Gastropoda from Western Pacific — Rehder 185 (1959:15) proposed to divide the family into two subfamilies, Vasinae and Xancinae, on the basis of differences in the radula and shell characters. On the basis of the rather funda- mental differences in the radula, and in order to minimize to some extent the disturbance to nomenclature caused by the action described above, I suggest that these subfamilies be raised to the rank of families. In this way we can retain the well-known family name Vasidae H. and A. Adams, 1853. Benthovoluta gracilior, new species Figs. 5 and 6 description: Shell of moderate size (50- 60 mm, about 2 inches long), fusiform, white, with a thin, light or straw yellow periostracum which under the microscope is seen to be minutely rough and lamellately scabrous, especially in the area between suture and pe- ripheral angulation of the ribs. Nuclear whorls 1%, smooth, bulbous; postnuclear whorls about 10% in holotype and largest paratype; first 5 whorls show about 8 strong ribs, markedly angulate at the periphery and crossed by 3 or 4 spiral cords below the periphery; area above the periphery smooth in earliest whorls, but showing spiral threads that gradu- ally increase in strength ; later whorls with more ribs (13-14 in penultimate whorl), which are less strongly angulated, and with fine spiral cords over the entire surface; last Figs. 5 and 6. Benthovoluta gracilior n. sp. ; approx. X 1. 5, Paratype. 6, Holotype. Fig. 7. Surculina blanda (Dali, 1908). Holotype; X 2. Fig. 8. Surculina cortezi (Dali, 1908). Holotype; X 1. Fig. 9- Surculina galapagana (Dali, 1919). Holotype; X 3. 186 PACIFIC SCIENCE, Vol. XXI, April 1967 showing two views of the lateral and the rachidian. Fig. 11. Radula teeth of Benthovoluta hilgen- dorfi (von Martens, 1897). Copied from Habe, 1952. whorl considerably longer than spire, with a long, straight, open anterior canal; outer lip broken in all specimens, but apparently simple, thin; columella with 2 low spiral folds, the upper fold larger. locality: Off Cagayan Islands, northern Sulu Sea, Philippines, in 508 fathoms; U.S.B.F. "Albatross I” Sta. 5423, 9° 38' 30" N, 121° 11' E, gray mud and coral sand bottom; March 31, 1909. Six specimens collected. measurements: SPECIMEN LENGTH WIDTH Holotype (usnm 637252) 54.75 mm 13.4 mm Figured Paratype (usnm 238408) 57.75 mm 12.9 mm remarks: This species differs from both B. hilgendorf von Martens and B. plicifera Yoko- yama in being more slender, with a thinner shell, and with the axial ribs more angulated at the periphery. REFERENCES Abbott, Robert T. 1950. The genera Xancus and Vasum in the western Atlantic. John- sonia 2(28) :201-219, pis. 89-95. • 1959. The family Vasidae in the Indo- Pacific. Indo-Pacific Mollusca 1(1): 15-32, 10 pis. Dall, William H. 1885. On Turbinella pyrum Lamarck, and its dentition. Proc. U. S. Natl. Mus. 8:345-348, pi. 19. 1908. Reports on the scientific results of the expedition to the eastern tropical Pacific, in charge of Alexander Agassiz, by the . . . "Albatross," from October, 1904, to March, 1905, .... XIV. The Mollusca and the Brachiopoda. Bull. Mus. Comp. Zook Plarvard College 43(6) :205-487, 22 pis. 191 8. Changes in and additions to molluscan nomenclature. Proc. Biol. Soc. Washington 31:137-138. 1919. Descriptions of new species of mollusks of the family Turritidae from the west coast of America and adjacent regions. Proc. U. S. Natl. Mus. 56(2288) :l-86, pis. 1-24. Grant, U. S. IV, and R. Gale Hoyt. 1931. Catalogue of the marine Pliocene and Pleisto- cene Mollusca of California and adjacent regions. Mem. San Diego Soc. Nat. Hist. 1:1-1036, 32 pis. Habe, Tadashige. 1952. Illustrated Catalogue of Japanese Shells, 1 (18) : 1 2 1—1 32, pi. 18, 28 text figs. Hemming, Francis. 1957. Opinion 489. Vali- dation under the plenary powers of the generic name ” Turbinella ” Lamarck, 1799 (Class Gastropoda), as the name for the sacred chank shell of India. Opinions and declarations rendered by the Intern. Comm. Zool. Nomen cl. 17:155-178. Kelly, Kenneth L., and Deane B. Judd. 1955. The iscc-nbs Method of Designating Colors and a Dictionary of Color Names. Natl. Bur. Standards Circular 533, iv + 158 pp. Supplement of Color Name Charts (reprinted in 1965). Kuroda, T. 1931. Two new species of Volu- tacea. Venus 3(1) :45-49, 3 figs. 1965. On the generic position of Benthovoluta (Gastropoda). Venus 24(1): 50-52. and Tadashige Habe. 1950. Volutidae in Japan. Illustr. Cat. Japanese Shells, 1(5) : 3 1-38, pis. 5-7, 6 text figs. Pilsbry, Henry A., and Axel A. Olsson. 1954. Systems of the Volutidae. Bull. Am. Paleontol. 35(152) :1— 36, 4 pis. Powell, A. W. B. 1942. The New Zealand Recent and fossil Mollusca of the family Turridae with general notes on turrid nomen- clature and systematics. Bull. Auckland Inst, and Mus. 2:1-188, 14 pis., 6 text figs. New Gastropoda from Western Pacific — Rehder 187 Sars, G. O. 1878. Bidrag til kundskaben om Norges Arktiske fauna. I. Mollusca regionis Arcticae Norvegiae. Christiania. XIII -f- [3] + 466 pp., 1 map, 52 pis. Stimpson, William. 1865. On certain genera and families of zoophagous gasteropods. Am. J. Conch. 1(1) : 5 5-64, pi. 8, 9. Thiele, Johannes. 1929. Handbuch der sys- tematischen Weichtierkunde. Erster Teil. Gustav. Fischer, Jena. 376 pp., 470 figs. Yokoyama, Matajiro. 1920. Fossils from the Miura Peninsula and its immediate north. J. Coll. Sci. Imp. Univ. Tokyo 29(6): 1-193, 19 pis., 1 map. Pogonophora from the Northeastern Pacific: First Records from the Gulf of Tehuantepec, Mexico Oluwafeyisola S. Adegoke1 Northeastern Pacific records of species of the Phylum Pogonophora Johansson, 1937 are few. In all, eight species have been recorded. The first was by Kirkegaard (1956^, 1961) who described Lamellisabella ivanovi from the Gulf of Panama. In two successive records, Ivanov (1961, 1962) described Galathealinum brachiosum, and Heptabrachia ctenophora and H. canadensis , respectively, from the west coasts of Canada and Oregon. Hartman (1961) re- corded abundant occurrences of Siboglinum veleronis Hartman from the La Jolla Canyon off the coast of southern California. Southward (1962) next described Galathealinum arcticum from Arctic waters off the northern coast of Yukon, Alaska; and more recently, Cutler (1965) described two new species of Sibog- linum, S. albatrossianum and S. ecuadoricum, and an undetermined specimen, from collec- tions dredged off Cape San Francisco, Ecuador, by the U. S. Fish Commission steamer "Alba- tross” in 1888. The occurrences of a few dark-brown, cylin- drical collar segments, measuring about 2.9- 4.0 mm across, and 3.3 mm long, from West Cortes, East Cortes, and Long Basins, and from the San Diego Trench, were recorded by Hart- man and Barnard (I960). These were later referred to the genus Galathealinum Kirke- gaard, 1956 by Hartman (1961:546), who also mentioned a new record of another species of Siboglinum closely resembling S. veleronis from "Velero IV” Station 7231, off San Eugenio Point, Lower California, Mexico. ACKNOWLEDGMENTS The materials on which this study is based were made available to the author by the kind permission of Professor J. Wyatt Durham and 1 Department of Paleontology, University of Cali- fornia, Berkeley, California. Present address: Cali- fornia Institute of Technology, Pasadena, California. Manuscript received April 18, 1966. Mr. J. H. Peck, Jr. of the Museum of Paleon- tology, University of California, Berkeley. The j author is grateful to Dr. Gwyn Thomas of the j Geology Department, Imperial College, Lon- don, for calling his attention to the fact that the specimens might represent an undescribed j species, and to Professor H. A. Lowenstam of ; the California Institute of Technology, Pasa- dena, for his critical reading of the manuscript and for his many helpful suggestions. The illustrations were prepared by the writer and Jurrie J. van der Woude of the California Institute of Technology. The author’s wife, [; Adekunbi Adegoke, assisted in preparing the manuscript. - MATERIAL STUDIED The new species, Galathealinum mexicanum | sp. nov., described below is the fourth species j to be described in the genus Galathealinum . It was dredged by the Vermillion Sea Expedition I (1958) from a depth of 3531-3603 m in the ! Gulf of Tehuantepec, Mexico (Univ. Calif, j Mus. Paleo. locality B-7469). It is the first pogonophoran record from the Gulf of Te- 1 huantepec. Only the dried remains of the tubes are 1 preserved. These dark brown tubes are thick- I walled and rigid, and have preserved their true ' cylindrical shape. They taper slightly and uni- formly, have an average diameter of over 2.0 mm, and are prominently subdivided along j the entire length into segments, each about twice as long as the average diameter of the 1 tube. The exterior of the tubes is covered by a thin, feltlike layer composed of fine and coarse fibers. The coarse fibers are more prominent j 1 and more numerous near segmental junctions. These characters place this species within the genus Galathealinum as defined by Kirkegaard (1956) and Ivanov (1963). Although most Recent pogonophoran genera and species are established primarily on the 188 Pogonophora from Tehuantepec — Adegoke basis of soft part anatomy, remains of well- preserved tubes are also known to be sufficiently reliable for "specific diagnosis” (see Hartman, 1961:546; Ivanov, 1963:120). The latter author firmly established the validity of this contention by constructing a dichotomous key for the identification of most Recent pogonc- phoran species from remains of their tubes (Ivanov, 1963:456-461). The characters of the tubes described and figured in this report readily distinguish them from the three pre- viously described species of Galathealinum. It is hoped that, in the future, preserved material from the same area will provide data on the nature of the soft part anatomy of the new species. SYSTEMATIC DESCRIPTION Phylum pogonophora Johansson, 1937 Order thecanephria Ivanov, 1955 Family polybrachiidae Ivanov, 1952 Genus Galathealinum Kirkegaard, 1956 Galathealinum Kirkegaard, 1956, Galathea Rept. 2:79-83. type species: Galathealinum hruuni Kirke- gaard. Galathealinum mexicanum sp. nov. Figs. 1-7 diagnosis: Galathealinum with elongate, cylindrical, segmented tube; circular cross-sec- tional outline; average tube diameter 2.0-2. 5 mm; covered along entire length by thin, felt- like layer of fine and coarse fibers ; individual coarse fibers at segmental junction about 15— 22p thick; segment length approximately twice the diameter. description: This species is represented by fragments of dried tubes only. The tubes are brownish-gray to dark-gray, elongate and slen- der. The longest fragment (holotype, Fig. 1) is 147.5 mm long. Maximum diameter 2.5 mm, minimum diameter 1.96 mm. Externally, the tube is divided into numerous prominent seg- ments (see Figs. 1-6). These segments have a i circular cross-sectional outline and are of rather i uniform length, each measuring 3. 8-4.9 mm. The widened funnel-like frills that are prom- inently shown at the nodes of the described 189 Fig. 1. Galathealinum mexicanum Adegoke sp. nov. Tubes from Univ. Calif. Mus. Paleo. locality B-7469, Gulf of Tehuantepec, Mexico. Holotype, ucmp 32882. Entire specimen showing long, curved, segmented tube, X 3/4. species of Galathealinum are only poorly pre- served on the dried tubes of the present species (Figs. 3, 7). The coarse fibers generally asso- ciated with these frills are well developed, how- ever, and are more abundant at the nodes than in the much longer internodes. The exterior of the tube is coated by a thin, friable, feltlike layer, mostly composed of numerous, very fine, light-brown fibers, and few, coarse, reddish- brown, glistening fibers. The latter also pene- trate the tube wall and appear as faint ridges on the otherwise smooth interior surface. About 60—80 coarse fibers are present in each inter- node. Though essentially transverse, they are 190 Fig. 2. Galathealinum mexicanum Adegoke sp. nov. Paratype, ucmp 12155. Anterior region of para- type showing well-defined segments, coarse transverse fibers, and some encrusting calcareous annelid tubes, X 1.7. rather irregularly oriented, and a few are even confluent. The interior wall is a glossy, dark- brown color, and is devoid of the feltlike layer. Both the extreme anterior and posterior ends of the tubes are unknown. dimensions: Holotype : length 147.5 mm, di- ameter of wider anterior end 2.5 mm, diameter of smaller posterior end 2.2 mm, average length of segments 4.0 mm, thickness of wall 0.1 mm, thickness of coarse fibers 15-22p. Paratype: length 77.5 mm, diameter of wider anterior end 2.4 mm, diameter of posterior end 1.96 mm, average length of segments 4.5 mm, thickness of wall about 0.1 mm, thickness of coarse fibers 15— 22| x. holotype: Univ. Calif. Mus. Paleo, no. 32882, locality B-7469. paratype: ucmp no. 12155, from type local- ity. PACIFIC SCIENCE, Vol. XXI, April 1967 Fig. 3- Galathealinum mexicanum Adegoke sp. nov. Anterior end of holotype showing a few seg- ments with coarse fibers and remnants of segmental frills at the nodes, X 5. Fig. 4. Galathealinum mexicanum Adegoke sp. nov. Part of holotype enlarged to show irregular coarse fibers, X 5. occurrence: ucmp locality B-7469. Few tubes and some echinoids dredged by the Ver- million Sea Expedition S.I.O., from the Gulf of Tehuantepec, Mexico. Latitude 14°28/N to 14° 29'N. Longitude 93° 09'W to 93° 10'W. Depth 1,935-1,974 fathoms. Field no. P-128- 58. remarks: Galathealinum mexicanum sp. nov. resembles other described species of Gala - thealinum in the possession of a dark-brown, segmented tube covered externally by a thin, fri- able, feltlike layer. Its dimensions are closest to Pogonophora from Tehuantepec — Adegoke 191 Fig. 5. Galathealinum mexicanum Adegoke sp. nov. Part of paratype enlarged to show coarse fibers and details of constricted segmental junctions, X 5. Fig. 7. Galathealinum mexicanum Adegoke sp. nov. Part of paratype enlarged to show details of segmental junction and remnants of segmental frills at nodes. Fig. 6. Galathealinum mexicanum Adegoke sp. nov. Part of paratype enlarged to show three encrust- ing annelid tubes, X 5. those of G. hrachiosum Ivanov, whose diameter ranges from 2.0 to 2.6 mm, from which it ma y be readily distinguished by the longer segments (about 4.0 mm long as against 1 mm in G. hrachiosum', see Ivanov, 1963: Fig. El62). Moreover, the coarse fibers in the new species measure 15-22p, whereas they are 7-1 2p thick in G. hrachiosum. From G. hruuni Kirkegaard, the new species may be distinguished by its larger dimensions, relatively weaker segmental frills, and thicker, coarser fibers (2-4^ thick in G. hrunni ). The new species may be readily distinguished from G. arcticum Southward by its larger diam- eter (1.33-1.95 mm in the latter) ; larger length-diameter ratio of each segment, which is about 2 in the new species and 1 in G. arcticum ; and the thicker coarse fibers, which are only l-2p thick in G. arcticum. The longest fragment represented in the col- lection is 147.5 mm long. As the extreme ante- rior and posterior ends of this tube are not represented, and because of the rather minor variation in taper between the two ends (ante- rior diameter 2.5 mm, posterior diameter 2.2 192 PACIFIC SCIENCE, Vol. XXI, April 1967 mm), it is here suggested that the total actual length of the tube may be several times the length of this fragment. Segmental funnel-like frills, a common char- acteristic of the genus, are only weakly repre- sented here (Figs. 3, 7). According to Ivanov (1963:412), these frills are soft and pliant in G. brachiosum and consist entirely of the exter- nal fibrous layer. It is, therefore, easily conceiv- able that these frills, originally present on this new species, became shrunken and inconspic- uous because of the poor conditions of preserva- tion of the tubes. The marked concentration of irregularly oriented coarse fibers in the vicinity of each segmental junction (see Figs. 3, 4) sup- ports this contention. Pogonophoran tubes are generally straight. Most of the tubes of this species in the collec- tion studied are also straight (see Fig. 2). A few, however, are curved (see holotype, Fig. 1). This curvature is considered to be a shrink- age phenomenon, as a result of drying. REFERENCES Cutler, E. B. 1965. Pogonophora from the eastern tropical Pacific, including two new species of SibogUnum. Pacif. Sci. 19(4): 422-42 6, 13 figs. Hartman, O. 1961. New Pogonophora from the eastern Pacific Ocean. Pacif. Sci. 15(4): 542-546. and J. L. Barnard, i960. The b-nthic fauna of the deep basins off southern Cali- fornia. Allan Hancock Pacif. Exped. 22:217- 284. Ivanov, A. V. 1961. New pogonophores from the eastern part of the Pacific Ocean. I. Gala- thealinum brachiosum sp. n. Zool. Zh. 40: 1378-1384. 1962. New pogonophores from the eastern part of the Pacific Ocean. II. Hepta- brachia ctenophora sp. n., and H. canadensis sp. n. Zool. Zh. 41:893-900. 1963. Pogonophora. (Translated and edited by D. B. Carlisle.) Consultants Bureau, New York. 479 pp. Kirkegaard, J. B. 1956. Pogonophora, Gala- thealinum bruuni n. gen., n. sp., a new repre- sentative of the class. Galathea Rept. 2, 79- 83. 1956a. Pogonophora. First records from the eastern Pacific. Ibid. 2:183-186. 1961. Pogonophora. III. The genus Lamellisabella. Ibid. 4:7-10. Southward, E. C. 1962. A new species of Galathealinum (Pogonophora) from the Ca- nadian Arctic. Canadian J. Zool. 40:385-389. Studies in the Calcium and Phosphorus Metabolism of the Crab, Podophthalmus vigil (Fabricius)1 Bryant T. Sather2 ABSTRACT: By employing modifications of the molt classification by Drach (1939) and Hiatt (1948), it was discovered in laboratory-maintained crabs ( Pod- ophthalmus vigil) that a partial desiccation occurred during proecdysis followed by a rehydration at the A stages. The inorganic and organic content of the carapace, mid-gut gland, gills, and muscles were followed during the molt cycle. The carapace had the greatest in- organic fluctuations. The mid-gut gland and muscle tended to increase in both organic and inorganic matter during premolt, suggesting that these organs may serve as reservoirs for these components. The calcium and total phosphorus constituents of these organs and of the blood were determined at the various molt stages. Fluctuations in the amounts of these two elements were observed in all sampled tissues. The storage of calcium in the mid-gut gland and muscles during premolt is discussed. Phosphorus was found to be stored in the digestive gland during postecdysis but not in proecdysis. The muscle also tended to store phosphorus during premolt. As P. vigil becomes older, i.e., larger, it is unable to resorb from the exoskeleton the same quantity of calcium, but it is able to recalcify the new exoskeleton to the same extent as does a smaller crab. Calcification and hard tissue formation occurs in many forms of life. It is found in bacteria (Ennever, 1963; Rizzo, et al., 1963; Greenfield, 1963), algae (e.g., Porolithon and Halamita ), protozoans (Isenberg, et al., 1963; Be and Eric- son, 1963), coelenterates, echinoderms, molluscs, arthropods, and vertebrates. Generally, the func- tion of calcification is to give form, support, and protection, and to contribute in ionic homeo- stasis (Urist, 1962), but in some instances calcification can be considered a pathological condition. The calcium complex deposited may be in three forms — calcite, aragonite, and apatite. The latter is a calcium phosphate [Ca10(PO4)6(OH)2] and the others are cal- cium carbonate complexes. Very little phos- phorus is found in calcite and aragonite, which 1 A contribution of the Pacific Biomedical Research Center and Contribution No. 259 of the Hawaii Institute of Marine Biology, University of Hawaii. Manuscript received February 14, 1966. 2 Pacific Biomedical Research Center, University of Hawaii, Honolulu, Hawaii. Present address: Depart- ment of Zoology, North Carolina State University, Raleigh, North Carolina. are generally restricted to the lower phyla. The amount of strontium and magnesium, the crys- tal structure, and the density of the calcium carbonate determine the difference between ara- gonite and calcite. The latter has little strontium and magnesium present in its hexagonal, less dense crystalline structure. Apatite is found in vertebrate bone, dentine, cementum, and enamel. Regardless of the crystal structure and the phylo- genetic group in which it occurs, the process of calcification can be considered to be basically the same (Travis, i960, 1963), although the function may be specifically adapted for dif- ferent requirements. In crustaceans, molting is necessary in the apparent growth process. Thus, considerable quantities of calcium and organic constituents have to be resorbed from the exoskeleton prior to ecdysis, but total resorption is limited to certain areas, i.e., the endophragmal skeleton and the ecdysial sutures. After resorption (via the blood) of these constituents, the organism is confronted with an abnormally high concen- tration of these substances in its internal fluids and the animal must either store or excrete this 193 194 PACIFIC SCIENCE, Vol. XXI, April 1967 excess. If the availability of the resorbed con- stituent is sparse, the majority of this element is usually stored. Crayfish generally store some of the resorbed calcium and phosphorus as small buttons, called gastroliths, in the lin- ing of the stomach. Gastrolith formation and dissolution have been followed throughout the ecdysis cycle by Damboviceanu (1932), Numanoi (1937), Keyer (1942), Scudamore (1942, 1947), Travis (1955 b, I960, 1963), and McWhinnie (1962). Marine crustaceans generally store some of the resorbed calcium and phosphorus in the mid-gut gland. Paul and Sharpe (1916) have reported that this process occurs in Cancer pagurus. This also occurs in Carcinus maenas (von Schonborn, 1912 ; Robert- son, 1937), in Maia squinado (Drach, 1939), in Hemigrapsus nudis (Kincaid and Scheer, 1952) and in the lobster, Panulirus argus (Travis, 1955^). Miyawaki and Sasaki (1961) found the same in the fresh water crayfish, Procam- barus. Calcium is present in relatively high con- centrations in sea water and therefore this ele- ment may not be a limiting factor in molting, and so not much of it may need to be stored during proecdysis of a marine crab. The con- centration of phosphate in Hawaiian waters, however, is small (Sather, 1966), and there- fore it would seem to be necessary for the ani- mal to conserve this element to a greater extent than calcium. After ecdysis is completed, the organism would use the resorbed and stored materials for calcification of the new exoskele- ton. The amount of inorganic material stored, however, is not sufficient to account for the total amount found in the intermolt crustacean. Therefore, the animal must actively concentrate the elements from the environment. The molt cycle of crustaceans has been the subject of a great number of investigations. Apart from descriptions of morphological changes, the mineral metabolism has been stud- ied to a certain extent, particularly changes in calcium and phosphorus content (Travis, 1954, 1955^, 1963). But such changes have been in- vestigated only at random periods in the molt cycle, and only in certain tissues and organs (glands). Some emphasis has been placed on the effect of hormonal influences (eyestalk hor- mones, etc.) on the alterations (Carlisle, 1954; McWhinney, 1962). No data have been avail- able on calcium and phosphorus metabolism throughout the entire molt cycle of a crab, nor has anything been known of the concentrations and distribution of these elements in the animal at times of calcification and decalcification, pe- riods of major importance in the cycle. There- fore, these studies were undertaken on the physiological processes which occur in the molt cycle of the crab, Podophthalmus vigil. MATERIALS AND METHODS In the period from March 1961 to October 1963, approximately 1,450 specimens of P. vigil were collected from Kaneohe Bay, Oahu, Hawaii and transported to the University of Hawaii Marine Laboratory. The animals were sexed, staged, tagged, and placed in aquaria with a continuous supply of fresh sea water. Modifications of the classification schemes of Drach (1939) and Hiatt (1948) were em- ployed to determine the molt stages of P. vigil. A descriptive analysis of the molt scheme was presented by Sather (1966). Crabs in the same stage were placed in a specific aquarium. The animals were usually fed pieces of frozen fish twice a week, but occasionally fresh crab muscle or frozen beef liver was substituted. When a crab reached a desired stage, it was removed from the aquarium and carefully dried with tissue paper. A 1 ml blood sample was taken from the heart by making a small hole with a dental drill in the carapace immediately posterior of the cardiac and mesobranchial su- ture and inserting a No. 21 -gauge hypodermic needle fitted to a syringe into the exposed peri- cardium. The crab was then killed and rinsed with distilled water. The gills, mid-gut gland, muscle, and carapace were dissected free, and these, together with the "remainder,” were placed into separate tared crucibles. After weigh- ing, the crucibles were placed in a drying oven for 12 hours at 114°C. After weighing, they were dry-ashed at 550°C for 24 hours. The fresh, dry, and ashed weights were recorded and the water, organic, and inorganic contents were calculated. Aliquots of the ashed tissues were taken for the determinations of calcium and phosphorus. The blood samples were stored for later chemical analysis. The exuviae were treated in the same manner except that the fresh Calcium and Phosphorus Metabolism of P. vigil — Sather 195 weights were not determined because it was not possible to dry thoroughly the gills and endo- phragmal skeleton. The flame spectrophotometric analysis of Geyer and Bowie (1961) was used to deter- mine the calcium content of the ashed samples. The blood calcium was determined using the method of Ferro and Ham (1957a, 1957 b). The method of Bernhardt, Chess, and Roy (1961) was used to determine the phosphorus (P.205) content in both the ashed and blood samples. All flame spectrophotometric deter- minations were carried out on a Beckman DU spectrophotometer equipped with a hydrogen- oxygen burner and a photomultiplier. Blank samples were carried throughout the analysis. The hydration, organic, inorganic, calcium, and phosphorus data were transformed to arc- sin values and the latter were statistically ana- lyzed to determine whether interactions between the various parameters were present. The pa- rameters were also subjected to the D-test of Hartley to ascertain the differences among the means (Snedecor, 1959). RESULTS Table 1 contains the results of the statistical interaction analysis. The law of probability values indicate that interactions of hydration, organic, inorganic, calcium, and phosphorus contents had occurred, which illustrates that the chemical parameters of the organs did not uni- formly fluctuate from one molt stage to another. The interactions demonstrate that the compo- nents were being accumulated by the various TABLE 1 Interaction Analysis of Percentage, Composition of Five Components in Sampled Organs of P. vigil % composition no F-VALUE PROBABILITY Hydration + 16.36 55.33 < 0.01 Organic 15.78 7.56 < 0.01 Inorganic 15.98 27.86 < 0.01 Calcium 15.04 2.86 < 0.01 Phosphorus 11.30 5.86 < 0.01 no = average number in each class. + = calculated real values. fi, {2 (degrees of freedom) rr 20 and 400, respectively. organs during different stages, which suggests that the constituents were being transferred be- tween organs at different times. The results of the statistical comparisons of per cent hydration, inorganic and organic, and calcium and phosphorus among the means are incorporated in Tables 2, 3, and 4, respectively. The concentrations of the components in the various organs, throughout the molt cycle, are listed in decreasing order. In Table 2 the ap- pearance of a superscript number in a molt period signifies that the content of the organ at that period is greater than those with a lesser superscript and without a superscript. The con- tents during molt periods having equal super- scripts are not statistically different from each other. For example, the hydration of the mid- gut gland (Table 2) during the C3_4 period is significantly greater than that during Q.2, B4.2, Ai_2, D-l.2, and D3.4. The amounts of mid-gut gland water during the C^, B4_2, and A-^ periods are greater than those during D4_2 and D3_4; but, the amounts at the C4.2, B4.2, and A1-2 periods do not significantly differ from each other. The ordinates of Figures 1-7 are expressed as per cent content, which is not the most ideal TABLE 2 Comparison Among the Means: Per Cent Hydration of Four Organs of P. vigil Throughout the Molt Cycle ORGAN AND CONCENTRATION IN DECREASING ORDER mid-gut CARAPACE gill GLAND MUSCLE Ai-25* C 3 '"-'1-2 C3.45 c3-44 td ta C3-43 C!-22 Cl-24 C3-4 Bl-23 Bl-22 »l-22 Cl-2 Al-21 Ai-22 Ai-22 ^3-4 Dl-2 Dl-2 £*3-4 D4_2 D3-4 D3-4 D1.2 * Explanation of superscript numbers: 5 = Significantly greater content than those in the last 5 stages. 4 = Significantly greater content than those in the last 4 stages. 3 : Significantly greater content than those in the last 3 stages. 2 = Significantly greater content than those in the last 2 stages. 1 - Significantly greater content than those in nonsuper- scripted stages. 196 PACIFIC SCIENCE, Vol. XXI, April 1967 TABLE 3 Comparison Among the Means: Organic and Inorganic Contents of Four Organs of P. vigil Throughout the Molt Cycle ORGAN AND CONCENTRATION IN DECREASING ORDER CARAPACE GILL MID-GUT GLAND MUSCLE % ORGANIC D3-4 D3-43* Dl-24 Dl-24 Ci_2 Al-23 D -± ^3-4 D3-44 ^3-4 Di-23 A^I \-22 Bl-2 Bi_2 Cl-21 Bl-22 Ai_2 Cl-2 B1-21 Cl-2 Dl-2 C3-4 C3-4 C3-4 % INORGANIC Dl-25 Dl-24 ^3-4° C3-4 D3-42 D3.4 Dl-22 0,-2 ^-3 -4 2 C3-4 Ai.21 Cl-2 Cl-22 Cl-2 Bi-21 D3-4 B1-2X Bl-2 C3-4 Bi-2 A1.2 Ai-2 Ci-2 Ai_2 * For explanation of superscript numbers see legend for Table 2. TABLE 4 STAGE Fig. 1. Changes (mean ± S.E.) in the watei content of five organs of P. vigil during the molt cycle. Comparison Among the Means: Calcium and Phosphorus Contents of Five Organs Throughout the Molt Cycle ORGAN AND CONCENTRATION IN DECREASING ORDER CARAPACE GILL MID-GUT GLAND MUSCLE BLOOD % CALCIUM C3-42 c 1 '-'3-4 D3-4X D3-41 Dl-25 D3-42 D„ A o-4 Di-21 C3-4 Bl-2 Dl-21 Ci-21 C3-4 D1 2 Cl-2 Cl-21 Ai-,1 A1-2 Cl-2 C3-4 Bi-2 Dl-21 Bi_2 Ai-2 A1-2 Ai2 Bi-2 Cl-2 B1-2 D3.4 % PHOSPHORUS Ai-25 D3-4 Bi-21 Di-21 Di-24 Bl-22 C3-4 Ci-21 A1.21 D3.41 Cl-21 Cl-2 A1-21 C1-2 Cl-2 D3-4 Bi-2 C3-41 Bl-2 Ai-2 C3-4 A1-2 Di-21 D3 4 Bi 2 Di-2 D1-2 D3-4 C3-4 C3-4 * For explanation of superscript numbers see legend for Table 2. index. Alterations in one component, e.g., or- ganic content, may affect the per cent composi- tion of another component, i.e., inorganic con- tent. Therefore, losses or gains in one particular constituent may only reflect losses or gains in another. Alterations in per cent composition were chosen because the data appearing in most of the literature were expressed in these terms and, thus, this index made comparisons more accessible. More expressive indices would be: mg or mEq/mg N, or mg or mEq/gm water. The latter ratio is more valid when comparisons of equilibria are desired (Robertson, I960). Figure 1 illustrates the alterations in water content of the carapace, mid-gut gland, muscle, gills, and "remainder” during the molt cycle. It is clear that the organs become somewhat de- hydrated during proecdysis and rapidly rehy- drated during ecdysis. The organic and inorganic contents of the four organs are plotted in Figures 2 and 3, respectively. The comparable data for the "re- mainder” were not determined because this por- Calcium and Phosphorus Metabolism of P. vigil — Sather 197 Fig. 2. Changes (mean ± S.E.) in the organic content of four tissues of P. vigil during the molt cycle. tion was composed primarily of exoskeleton, and so the values would probably approximate those of the carapace. It is apparent that the organic contents of the mid-gut gland, gill, and muscle increased during the proecdysial stages. The greatest inorganic fluctuation was found in the carapace. Only minor alterations were found in the other tissues. The calcium and phosphorus composition of the carapace, mid-gut gland, gills, muscle, and blood were determined. The results, based on dry weight, are plotted in Figures 4-8. In Fig- ures 4-7, the data are plotted as changes in per cent dry weight. The values for the blood (Fig. 8) are presented as mM/liter. The organic and inorganic composition of the exuviae were also determined. The results are illustrated in Figure 9. Also contained in this figure are the calcium and phosphorus con- tents of the exuviae, expressed as per cent com- position. The organic, inorganic, and calcium contents of the entire exuvia were compared with those found in the exuvial carapace; this information is summarized in Table 1. The carapace con- tained less organic material and less calcium than the entire exuvia. The mount of inorganic Fig. 3. Changes (mean ± S.E.) in the inorganic content of four tissues of P. vigil during the molt cycle. Fig. 4. Calcium and phosphorus content (mean ± S.E.) of the carapace of P. vigil during the molt cycle. 198 PACIFIC SCIENCE, Vol. XXI, April 1967 STAGE Fig. 5. Calcium and phosphorus content (mean ± S.E.) of the gills of P. vigil during the molt cycle. Fig. 7. Calcium and phosphorus content (mean ± S.E.) of the mid-gut gland of P. vigil during the molt cycle. o UJ 5 Fig. 6. Calcium and phosphorus content (mean ± S.E.) of the muscles of P. vigil during the molt cycle. Fig. 8. Calcium and phosphorus content (mean ± S.E.) of the blood of P. vigil during the molt cycle. Calcium and Phosphorus Metabolism of P. vigil — Sather 199 Fig. 9. Percent composition of the exuvia of P. vigil (values based on dry weight). matter in the carapace was greater than that in the entire exuvia. To determine whether larger crabs were able to resorb the same amount of calcium as smaller crabs, the amount of calcium in the exuviae was plotted against exuvial carapace width. (The data were then statistically analyzed for regres- sion and the slope was fitted by the least squares method.) Figure 10 clearly indicates that as the crabs increased in size the amount of resorbed calcium decreased, and the regression analysis showed that the calculated slope was 0.082 (P <0.001). DISCUSSION AND CONCLUSIONS Weight Changes During the Molt Cycle Changes in weight of Crustacea during ecdy- sis are due to absorption of water (Baumberger and Olmsted, 1928; Drach, 1939; Needham, 1946; Guyselman, 1953; Travis, 1954). The alterations of body weight and water content of P. vigil have been reported elsewhere (Sather, Fig. 10. Regression of calcium content of exuviae on exuvial carapace widths of P. vigil, b = 0.082 (P < 0.001). Upper and lower curves represent the 95% confidence limits. 200 PACIFIC SCIENCE, VoL XXI, April 1967 1966). In brief, during proecdysis the crabs have a tendency to lose weight, which can be attributed to a loss of water. Between D3 and D4 the mean weight gain was about 8%. The weight change between D4 and A4 was an in- significant loss of 1.5%. However, the crabs gained 18.8% between A4 and A2. No signi- ficant weight alterations were noted between A2 and B4, B4 and B2, and B2 and Q. Between stages C4 and C2 the weight gain was 4.4%. No significant changes were found during the remainder of the molt cycle. However, the over- all weight gain between two successive inter- molt stages was approximately 34%. The total water content of the crabs during the ecdysis cycle was also followed. During the premolt stages, the crabs tend to become de- hydrated ; the water lost at D4_2 and D3_4 was calculated to be 8.1% and 7.7% below the C4 water content of 70%. The postmolt water con- tent was increased from about 62% (D3_4) to 77% ( A1j2) and 75% (B4_2), but these water content changes were statistically not signi- ficantly different from the intermolt value of 70.3%. Water Content of Four Organs and (( Remain- der” Throughout the Molt Cycle The amount of water in the various tissues of a crustacean during the molt cycle has not been previously reported. The water content of the carapace, mid-gut gland, gills, muscles, and "remainder” in P. vigil is illustrated in Figure 1. (The values are the means db S.E.) Analysis of variance on the arcsin transformed values illus- trated that interaction was present, showing that the water content of the tissues did not vary uniformly throughout the molt stages. The greatest fluctuations in water content were found in the carapace during ecdysis. A decrease of about 5% was noted during the proecdysial stages, but this was not statistically significant. The gain in water content between D3_4 and A4_2 was a significant gain of about 37%. Although the extracellular fluid volumes were not determined, this gain could possibly reflect a greater extracellular fluid volume. Dur- ing the B4_2 and C4.2 stages, the water content was decreased to 20.54%, which was caused by the incorporation of calcium salts. The alterations in gill hydration are repre- sented as the top curve in Figure 1. During the premolt stages, the gills lost 4.57% of their water content — a significant decrease. After ecdysis the gill water content was increased to about 87.64%, which was found to be a sig- nificant gain. The hydration at B4.2 was in- creased to a significant 89.42%. The per cent hydration of the gills during the Cx_2 period was not significantly different from that at the C3_4 duration (90.20%). Robertson (i960) has demonstrated that the gills of Carcinus maenas were the site of water and ion absorption but that the antennal glands were the sites for the loss of the water and the ions. Because the urine of P . vigil was not sam- pled, it is not possible to exclude these glands and the gills as the sites of water flux. The decrease in water content of the muscles during the premolt stages from a value of 85% to 7 6% was found to be significant, as was the increase to 82% at the A4_2 periods. The fur- ther increases during postecdysis were not sig- nificant. The same type of pattern is seen to occur in the mid-gut gland. The intermolt water content was found to be 85.16%. The reduced hydra- tion during the premolt stages to about 68% was a significant drop. During the A4.2 dura- tion, the water content was increased to 78.24%. The subsequent changes observed during post- molt did not differ significantly. Excluding the carapace and the remainder, the increase of about 1 1 % in the mid-gut gland during ecdysis was the greatest alteration. Rob- ertson (I960) reported that in C. maenas the water content of the mid-gut gland and its fluid increased during the early postmolt stages. This was attributed to absorption of water via the fore-gut. The results reported here for P. vigil \ are consistent with those reported by Robertson (I960) and also with the findings of Drach (1939) for Mai a squinado and Cancer pa gurus, j The fluctuations of the "remainder” during the molt cycle are also illustrated in Figure 1. It is quite obvious that this portion also lost some j water during proecdysis. The intermolt water content was calculated to be 68.69%, and post- ecdysial values of about 65% were significantly j different from the former. After ecdysis the water content rose to a significant high of 83.18%. No statistical difference was found Calcium and Phosphorus Metabolism of P. vigil — Sather 201 between the A1.2 and B±_2 values. During the last two postecdysial stages, the water content decreased to 77.78%. Because the "remainder” was largely composed of exoskeleton, this curve follows the same general pattern as the cara- pace. The increase in hydration following ec- dysis was probably due to a greater extracellular fluid volume of the exoskeletal tissues. During calcification, the extracellular water was un- doubtedly replaced by calcium salts. The water content of the whole blood during the premolt stages of P. vigil was not deter- mined. Water is not absorbed during the first three proecdysial stages (Sather, 1966). Travis (1954, 1955 b) demonstrated that the premolt water uptake by Panulirus argus was limited to 15 minutes just prior to ecdysis. However, as can be seen in Figure 1, P. vigil becomes desic- cated during proecdysis. The water content of the intermolt blood was found to be 94.14%. The value for the Ax.2 stages was 93.98% ; at 91.46%; and at the period, 93.88%. Because the greatest absorption of water occurs during the period immediately following ecdysis, it would be expected that the blood at the A1_2 duration would be somewhat diluted (to be ex- plained below). Travis (1955 b) found that the blood calcium and phosphorus concentra- tions of P. argus were decreased following molt. Robertson (I960) reported that in C. maenas the concentrations of blood constituents were reduced during the early postmolt stages. Both authors attributed their findings to the uptake and retention of water. Proecdysial and early postecdysial P. vigil exhibit significant alterations in water content. This phenomenon offers a fine thesis problem on the osmoregulatory mechanisms and modes employed by the crab during this "stressed” duration. Organic Content of Four Organs Throughout the Molt Cycle The organic content of the carapace, gills, muscle, and mid-gut gland are plotted in Figure 2. Analysis of variance illustrated that the organic content of the organs did not uni- formly fluctuate. A great variability is seen in the organic con- tent of the carapace, but none of the alterations was found to be significant from the intermolt content of 35.76%. The organic composition of the mid-gut gland increased from an intermolt value of about 10% to approximately 25% at the pre- molt stages. This accounts for the hyperglycemia reported for proecdysial Panulirus japonicus by Scheer and Scheer (1951). After ecdysis, the organic value dropped to 16.15%, which can be similarly attributed to the loss of reserves utilized for the active process of molting and for the tanning of the pigmented layer, espe- cially as the crabs do not feed until the late B stage. The value found at the C±_2 period was not significantly different from that of the pre- vious period, but it was greater than that of the C3_4 stages. During the early C stages, P. vigil becomes voracious, obviously to compensate for the loss which occurred during ecdysis. The alterations in the organic content of the muscle were similar to those of the mid-gut gland. From an intermolt level of 11.66%, the organic content increased to about 20% during proecdysis. The significant increase in organic material during these periods probably is due largely to the accumulation of glycogen to serve as an energy source for the epidermal cells which form the proecdysial tissues and the molt- ing fluid. It is quite possible that the muscle organic constituents would be at a high level just prior to ecdysis, as the active process of exuviation would require a great amount of energy. In addition, crabs in the early D3 stage did feed and the resulting energy would hive to be stored either in the mid-gut gland or in some other tissue. Undoubtedly, the former has a saturation level and the most likely tissue, therefore, would be the muscle because of its adequate blood supply. The lowest curve in Figure 2 illustrates the changes in the organic content of the gills. The minimum concentration of organic materials occurred in the intermolt period (6.51%) and the highest (10.45%) was found to occur in the D3_4 stages. The premolt gains were sig- nificantly greater than the intermolt value. During the postecdysial stages, the gill organic content decreased to the original intermolt con- tent. One possible explanation for the observed increase in gill organic content at D3.4 would be 202 PACIFIC SCIENCE, Vol. XXI, April 1967 that a supply of energy would be required to regulate the ionic ratio of the internal to ex- ternal media. Unfortunately, the glycogen con- tent of the gills was not determined. Consider- ing the decreased water content of P. vigil during late premolt, it appears that the gills may be possible sites of water and/or ion efflux. This hypothesis is not in agreement with the report by Robertson (I960), which stated that the antennal glands of C. maenas were the sites of water and ionic efflux. One possible method for determining the site of water flux in P. vigil would be the use of tritiated water. Radio- isotopes representing the extracellular ions would also verify the sites of ionic flux. Because the organic content of premolt blood was not determined, it was not possible to fol- low the changes throughout the ecdysis cycle. The organic constituent during the A4_2 period was 2.93% of the whole blood. The content in stages B4_2 was 4.02%; in the C4_2 stages, 2.86%; and in the C3_4 stages, 2.64%. The ob- served increase during the B period may be due to the commencement of food ingestion and the subsequent loss (during the C stages) by the distribution to the mid-gut gland and exo- skeleton. Inorganic Content of Four Organs Throughout the Molt Cycle The fluctuations in inorganic content are illus- trated in Figure 3. Analysis of variance indi- cated that interaction was again present. As could be expected, the carapace varied much more in inorganic content than did the other sampled tissues. The intermolt value was determined to be 42.12%. The significant in- crease to 52.01% during the D4.2 stages pos- sibly can be attributed to the formation and tanning of the new epicuticle. However, the inorganic constituent in the proecdysial tissues is not dominantly calcium (see Fig. 4). The finding is in agreement with those of other investigators (Drach, 1936, 1939; Krishnan, 1950; Travis, 1955^, I960, 1963). During the D3_4 stages when resorption was nearly com- pleted, the inorganic content was decreased sig- nificantly to 43.80%. It becomes apparent that the entire carapace is not an area where maxi- mum resorption occurs (see introductory re- marks). Following ecdysis the inorganic content of the carapace was approximately 20%, reflect- ing the amount found in the exocuticle. During the Ax_2 periods, the exocuticle was impregnated with calcium salts, with concomitant formation of the principal layer. During the B-^ and stages, periods of major calcification, the in- organic composition was increased to 27.31% and 41.28%; both were significant gains. The amount during the early C periods approxi- mated that of the intermolt content. The mid-gut gland’s alterations are plotted also in Figure 3. The increased values during premolt (D^ = 5.98%, D3.4 = 7.74%) were statistically different from each other and also from the intermolt value of 4.29%. This was probably due to the storage of some constituents absorbed from the exoskeleton. Similar storage has been reported by Paul and Sharpe (1916) for Cancer pagurus ; by von Schonborn (1912) and Robertson (1937, I960) for Carcinus maenas; by Drach (1939) for Maia squinado ; by Kincaid and Scheer (1952) for Hemigrapsus nudus; and by Travis (1955^) for Panulirus argus. The reduction in inorganic content dur- ing the postmolt stages can be attributed to the redistribution of the stored elements to the hardening exoskeleton. The gill inorganic content was found to vary slightly. The observed value of 4.15% at the D4.2 duration was found to differ significantly from those of the other durations. The greater inorganic content was due to the formation of the proecdysial tissues, which similarly occurred in the carapace, i.e., the epicuticle and exo- cuticle. As expected, the mean per cent inorganic composition of muscle did not vary significantly between stages (P > 0.05). The average con- tent of the inorganic materials was 2.44%, the range being 2.14-2.69%. The intermolt inorganic composition of the whole blood was 3.22%. During the A4.2 pe- riod it was 3.09%, and at the B4.2 it was 4.52%. The former value reflected the absorp- tion of sea water, and the latter value was prob- ably due to the mobilization of calcium, after it was actively transported by the gills to the exoskeleton. The blood inorganic content at the Ci-2 period was approximately equal to the intermolt value, indicating that the sclerotiniza- tion process was nearly completed. Calcium and Phosphorus Metabolism of P. vigil — Sather 203 Calcium and Phosphorus Contents of Five Or- gans Throughout the Molt Cycle The calcium and phosphorus contents of the carapace, gills, mid-gut gland, muscle, and blood were determined. The values, based on per cent dry weight, are plotted in Figures 4-8. The blood data are given in mM/liter. Inter- action analyses on the organs’ calcium and phos- phorus contents were positive (P < 0.01). The intermolt carapace calcium content (Fig. 4) was 49.51% and the phosphorus con- tent was only 5.42%. Both of these values are large in comparison with those reported by Prenant (1928) for five species of crabs in temperate waters. His data (per cent calcium and phosphorus, respectively) were: Carcinus maenas, 30 and 2; Maia squinado, 31 and 2; Portunus puher, 36 and 2; Cancer pagurus, 3 6 and 0.8; and Xantho floridus, 38 and 0.4. However, as noted by Vinogradov (1953), the majority of these values were only relative. Hayes, Singer, and Armstrong (1962) reported that the carapace calcium of the lobster, Ho ma- ms vulgaris, was 25.2% and the calcium con- tent of the claw was 23.8%. The phosphate contents of these two anatomical areas were reported to be 1.33% and 2.05%. The lower calcium content of temperate species may be a genetic difference or may be caused by a lesser availability of this environmental element. Un- fortunately, this latter possibility cannot be checked because calcium data at the collection sites were not available. The environmental calcium content certainly influences the amount absorbed by an organism, as well as the amount retained. It is known that the total calcium content of fresh water crustaceans is less than that in marine species. The calcium alterations of the carapace during the molt cycle are illustrated in Figure 4. Preceding molt, the calcium content varied only slightly during the D periods. After ecdysis (A^) the calcium content was dimin- ished to 26.04% and reflected the amounts in the epicuticle and pigmented layers. Be- tween Bj.g and Q.2, which was the major duration of calcification, the calcium content was increased significantly to 41.12%, which was similar to that of intermolt. The phosphorus changes of the carapace are also illustrated in Figure 4. The content at C3.4 was calculated to be 5.42%, which is much greater than that found in Panulirus argus by Travis (1957). In P. argus, in late stage C, about 3% of the total integument was com- posed of Ca3(P04)2, which is approximately O. 2% of the total phosphorus. In P. vigil, a small insignificant increase was observed during the last proecdysis stages. At stages A^, the phosphorus content was increased to 26.76% of the dry weight, which was about the same as the calcium concentration. This localization could have been due to the mobilization of phosphorus by the blood. Travis (1957) has demonstrated that the postecdysial integument stains heavily for alkaline phosphatase. Thus, a much greater amount of phosphorus would be present than during the other stages. It is thought that alkaline phosphatase liberates phosphates which combine with calcium to form the calcium phosphate complex. The high phosphate content during the A stages causes one to ponder over its significance, because the major anionic constituent of the intermolt in- tegument is carbonate and not phosphate. In P. vigil during the first C periods, the phos- phorus content decreased significantly to 11.34%. This reduction can be attributed to the increased deposition of calcium salts. The fluctuations in gill calcium are plotted in Figure 5. No significant differences were found between the intermolt value of 7.88% and the first premolt values. However, the decreased value of 0.30% at B^ did differ significantly from the other values. Because the period is the initial duration of greatest calcification, a high gill permeability, caused by calcium, would be greatly detrimental for ex- traction of calcium from the medium. Robertson (I960) demonstrated that a great influx of calcium occurred during postmolt in Carcinus maenas. Unfortunately, the amount of calcium in the gills was not measured. It appears, then, that in P. vigil the mechanism to increase the movement of calcium into the animal serves to reduce the gill calcium content, allowing cal- cium to enter across the gills at a more rapid diffusion rate, the blood then mobilizing the element to the integument. The phosphorus content of the gills during the molt cycle is illustrated in Figure 5. The 204 PACIFIC SCIENCE, Vol. XXI, April 1967 intermolt value was calculated to be 19.80% of the dry weight. Analysis of variance illus- trated that there were no significant differences among the means. However, there is a sug- gestion that the gills may store phosphorus during the Dx_2 stages. This suggestion is reinforced by the gill organic content at D4_2 (Figure 2). The phosphorus content may be indicative of an energy-requiring process for early postecdysial absorption of sea water. Figure 6 demonstrates the alterations of calcium and phosphorus content of the muscle throughout the molt cycle. The calcium con- centration increased from 7.70% at the C3_4 period to 10.72% at the D3_4 stage. Following ecdysis the calcium content was reduced to 4.14%, and at the B4_2 period it was further reduced to a significant 2.72%. During the period when the majority of calcification oc- curred (between B4.2 and C4.2), the muscle calcium was raised to 6.28%. During the process of exuviation, i.e., the time when active musclar contractions occur to facilitate withdrawal of the crab from the old exoskeleton, a large amount of energy would be required. As can be seen in Figure 2, the organic content of this organ also increased. Another requirement would be an ample supply of calcium to expedite muscular contractions. Calcium and organic reserves may be localized in this organ to insure proper exuviation. This phenomenon would then favor the survival of the crab during the process of ecdysis. The muscles, in addition to the mid-gut gland, may also serve as a place for calcium storage. The 6.58% decrease in calcium between the D3.4 and A4.2 periods may be due to the mobilization of the element to the exocuticle; during the first postecdysial stages, the latter is impregnated with and concomitantly hard- ened by calcium salts (Travis, I960, 1963). The muscle phosphorus content was greater than that of calcium; phosphorus is the most important element in muscle contraction. The values at D4.2 (25.98%) and A4.2 (24.82%) were significantly different from the intermolt content of 13.14%, but the former values were not statistically different from each other. The loss of phosphorus during D3„4 may have been due, in part, to the mobilization to the gills and carapace, but 2.67% cannot be accounted for. The reasons discussed in the above paragraph seem to be applicable to phosphorus as well as to calcium. However, the suggestion that pro- ecdysial storage of phosphorus in the muscles occurred is indeed very weak. On examination of Figure 6, it can be seen that phosphorus and calcium are controlled differently. After ecdysis the amount of calcium in the muscles is diminished, but the phos- phorus content remains relatively constant. The phosphorus and calcium fluctuations in the digestive gland are seen in Figure 7. As found in the alterations in carapace calcium and phosphorus (Fig. 4), the curves tend to be reciprocal to each other. But the mid-gut gland phosphorus content was always greater than that of calcium. Significant differences among the means did exist (P < 0.01). The intermolt calcium content of 9.31% did not differ significantly from that at D3_4 (13.05%). However, the gain suggests that some calcium is stored in the mid-gut gland during premolt. After ecdysis some of the cal- cium (5% in P. vigil ) may be used for cal- cification of the exocuticle. These observations are consistent with the reports of Paul and Sharpe (1916), von Schonborn (1912), Drach (1939), Kincaid and Scheer (1952), and Travis (1955^), but are inconsistent with that of Robertson (I960), who reported that in C. maenas the mid-gut gland secretion during postmolt (stages A and B) had about 16% more calcium than it did during intermolt. The phosphorus content decreased from 19.76% in C3_4 to 15.28% in the late D stages and increased to 21.04% and 25.50% during the At_2 and B4.2 stages, respectively. The postecdysial gain may have been due to mobili- zation from the gills (Fig. 5), which lost ap- proximately 10% during postmolt. It is obvious that this gland in P. vigil does not store phos- phorus during the premolt periods, but it does appear that the gland becomes a phosphorus reservoir after ecdysis. This finding is incon- sistent with the reports by Travis (1955 b, 1957) that phosphorus was stored in the mid- gut gland of P. argus prior to ecdysis, but that following molt, the phosphorus content rapidly decreased. The latter conclusion was based primarily on histochemical observations and Calcium and Phosphorus Metabolism of P. vigil — Sather 205 blood analysis, and no chemical analysis of the mid-gut gland was undertaken. A question arises after examining all of the phosphorus curves. Because the crabs were not feeding during the A and early B stages, where did the phosphorus originate? From Figure 5 it is seen that the phosphorus content of the gills is drastically increased during the D3_4 stages. Following ecdysis the phosphorus con- tent of the organ was reduced by approximately 9%. The phosphorus content of the mid-gut gland from D3_4 to At_2 was increased by about 6%. Therefore, possibly the gills, rather than the mid-gut gland, serve as a phosphorus reservoir. Another possible source for the ac- cumulation of phosphorus could be the water that was imbibed immediately following ecdy- sis. However, this is not likely because a pilot experiment demonstrated that phosphorus is not accumulated during and following ecdysis. The application of the radioisotoype P32 could be very useful in resolving this question. Figure 8 illustrates the calcium and phos- phorus contents of the blood during the molt cycle. The data are plotted on a volume basis, as is shown on the ordinate. The blood calcium and phosphorus levels tend to be parallel throughout the molt cycle. During the D4.2 periods, the blood calcium was significantly increased to 35.09 mM/liter from the C3_4 content of 21.58. A significant decrease to 17.68 mM/liter was observed at the D3.4 stages. In Panulirus argus , Travis (1955 b) also noted a premolt blood calcium increase, followed by a decrease in the late premolt stages. The loss was attributed to dilution when the lobster took in water. During the A period, the lobster’s blood calcium was at the intermolt value. The content was slightly increased at the B stages and, following this interval, i.e., during the C period, the concentration was de- creased below the intermolt value. In late premolt Carcinus maenas, Robertson (I960) also noted a blood calcium increase of about 21%. Within 24 hours after molting (Stage A), the blood calcium content was reduced by approximately 25%. In 2 to 14 days following ecdysis, the blood calcium was further reduced to approximately 31% of the intermolt value. The water content of P. vigil decreases during proecdysis (Sather, 1966). Also, as evidenced from inspection of the changes in water content of sampled organs (Fig. 1), de- hydration definitely occurred during the pre- molt stages. The first decrease in water content was found during the D4_2 period. This would account for the rise in the blood calcium at this interval. The observed reduction at the D3_4 stages is not due to the uptake of water. On a volume basis, the amount lost was calculated at 17.41%. However, on a dry weight basis, this loss was only 2.04%. The blood must have lost some calcium to other organs or the external medium. Thus, the calcium may have been distributed to the muscles and/or the mid- gut gland. The large standard errors (db 1.2%) for the latter two organs do not permit an accurate estimate of the quantity accumu- lated by each organ. Following ecdysis, no significant changes in the blood calcium were observed. As seen in Figure 1, the greatest increase in water content occurred at this time. It should be recalled that the water taken in was sea water, including the elements present in the medium. This has been verified by the studies of Robertson (I960). Thus, a great calcium dilution would not be expected. Also, as seen in Figures 6 and 7, the mid-gut gland and muscle lost some calcium which could have been accumulated by the blood. At period B^o the blood calcium was increased to 23.23 mM/liter. This could have been due to absorption of calcium, via the gills, from the environment. Figure 5 illus- trates that at this interval the gill calcium was drastically reduced, increasing the efficiency of extracting calcium from the external medium. Thus, except for the effect of dehydration at the early proecdysial stages, the calcium content of P. vigil blood remains more stable than in other investigated crustaceans. This fact may be due to the nearly chemically constant en- vironment of the crab. Except for one month, the environmental salinity and calcium was not less than 34 0/00 and 300 mg/liter, respec- tively (Sather, 1966). The total phosphorus fluctuations of the blood during the molt cycle are also depicted in Figure 8. The significant increase of blood phosphorus to the D^ interval of 25.52 mMI liter can be attributed to the desiccation of the animal. The reduction found at D3.4 (17.19 206 PACIFIC SCIENCE, Vol. XXI, April 1967 mM/liter) possibly reflects the relocation of phosphorus to the gills (Fig. 5). At stage A^, the value was slightly greater than the intermolt value. This decrease can be assigned to dilution, i.e., by the uptake of water from the environ- ment. In sea water, phosphorus is present in much smaller quantities than is calcium. The annual average phosphate content of sea water was less than 1 pg/ liter (Sather, 1966). Also, the results of a preliminary experiment illus- trated that the phosphorus content of the ex- ternal medium was increased when containing molting and postmolt crabs. The leveling-off of the blood phosphate at Bx.2 and the increase at Q.o was probably due to the resumption of feeding. The majority of P. vigil began to feed at B2 and only occasionally when they were in the Bx stage. Ho mar us am eric anus (Hollett, 1943), Panu- lirus argus (Travis, 1955^), and Carcinus maenas (Robertson, I960) also increased their blood phosphorus during the premolt stages. Travis (1955 b) found that after ecdysis, the phosphorus content steadily decreased. This was attributed to a depletion of phosphorus by cal- cification of the exoskeleton concomitant with a reduction in the stored mid-gut gland phos- phorus. Robertson (I960) reported that within 24 hours after ecdysis the blood phosphorus of C. maenas was slightly less than the inter- molt value. Within 2 to 14 days after molting the value was increased to about 22% above that of the intermolt level. The report of Dril- hon (1935) is not consistent with the above reports, in that the phosphorus content of the blood of premolt and postmolt Maia s quin ado was not altered. Composition of the Exuviae The discarded exoskeleton or exuvia of P. vigil is not consumed by the crab as it is in the insects. Robertson (1937) reported that the exuvia of Carcinus comprised about 46.2% of the total dry weight. Lafon (1948) stated that the value was 47.5% of the dry weight. These calculations were not made on the ex- uviae of P. vigil, but the percentages of organic, inorganic, calcium, and phosphorus contents were determined and these data are given in Figure 9. The per cent composition was based on the dry weight. The histogram illustrates that about 81% of the entire exuvia was com- posed of inorganic material and only approxi- mately 37% of the inorganic content was due to calcium. Odum (1957) reported that the calcium content of the exuviated chela of Uca pugnax was 27.7% of the dry weight, which is somewhat consistent with the calcium content of the exuvia of P. vigil. Knowing the amount of calcium in the exuvia (30%) and assuming that the intermolt carapace, which contained about 50% calcium, is representative of the entire exoskeleton, it is possible to calculate the quantity of calcium resorbed, which is about 20%. The amount of calcium stored in the mid-gut gland and the muscle was approximately 7%. Thus, the quan- tity stored and resorbed closely approximates the amount of calcium (26%) present in the early postmolt carapace. The increase in amount (23%) between A^ and C3.4 undoubtedly is acquired from the environment. The phosphorus content of the exuvia was found to be only approximately 5 X 10“ 7% of the dry weight. Unfortunately, comparable phosphorus data for other species have not been reported. Employing the above mathematical deductions, it would seem possible to account for the phosphorus budget. However, such a process produces a deficit of about seven magni- tudes. It is possible that the reproductive system and the gastrointestinal tract, which were not sampled, may have been highly selective for the storage of this element. The great resorp- tion of phosphorus is obvious and it must have been stored, because the results of an experi- ment showed that proecdysial and postecdysial crabs lose very little phosphate (2.6-5. 1 pg P04) to the environment. The organic content of the exuvia was found to be only 18.7%. This organic material has been reported to be composed of lipo-protein, chitin, mucopolysaccharides, and proteins (Tra- vis, 1955^, 1957, 1963; Dennell, I960). The amount of organic material resorbed was about 17%, which is consistent with the amounts stored in the mid-gut gland and muscle. Neto (1943) reported that the calcium con- tent of the carapace of Uca maracoani decreased as its breadth increased. The calcium content of the exuviae of P. vigil was compared with the width of the cast exoskeletons. Figure 10 Calcium and Phosphorus Metabolism of P . vigil — Sather 207 clearly demonstrates that less calcium is re- sorbed from the exoskeleton as the crab in- creases in size. The calculated slope was found to be 0.082, which indicates an increase of approximately 82 jxg Ca/mg ash for each centimeter increase in breadth. The slope dif- fered significantly from 0 (P < 0.001). A similar analysis was performed, comparing the amount of carapace calcium with the wet weight of the C4 crabs. The results, which are not illustrated in this report, demonstrated that regression did not occur and that the calculated slope was 0.0009. Thus it appears that less cal- cium is resorbed by large crabs and, therefore, more calcium appears in the exuvia. This phe- nomenon illustrates the effects of ageing on one physiological process of an invertebrate. It may be possible that the enzymatic activities responsible for crustacean decalcification are decreased with age and those required for recalcification are not influenced by senescence. A study of the alkaline phosphatase and car- bonic anhydrase activities of the epidermal cells may verify the above observations. ACKNOWLEDGMENTS The author is grateful to Dr. Sidney J. Townsley of the University of Hawaii for his suggestions, encouragement, and guidance dur- ing the study. I wish to thank also Dr. Pieter B. van Weel and Dr. Terence A. Rogers, of the University of Hawaii, for their criticisms and assistance during the preparation of the manuscript. This investigation was financed, in part, by the U. S. Atomic Energy Commission, Contract No. AT( 04-3) -235. Portions of the manuscript were taken from my Ph.D. dis- sertation for the University of Hawaii. REFERENCES Baumberger, J. D., and J. M. D. Olmsted. 1928. Changes in osmotic pressure and water content of crabs during the molt cycle. Physiol. Zool. 1:545. Be, A. W. H., and D. B. Ericson. 1963. As- pects of calcification in planktonic Foramini- fera (Sarcodina). In: Comparative biology of calcified tissue. Ann. N. Y. Acad. Sci. 109(1) :65-81. Bernhardt, D. N., W. B. Chess, and D. Roy. 1961. Flame photometric determination of phosphorus in the presence of sodium, po- tassium and other cations. Anal. Chem. 33:395. Carlisle, D. B. 1954. On the hormonal in- hibition of moulting in decapod Crustacea. J. Mar. Biol. Assoc. U. K. 33:61. Damboviceanu, A. 1932. Composition chi- mique et physicochimique du liquide cavitaire chez les crustaces decapodes (Physiologie de la calcification). Arch. Roum. Path. Exptl. Microbiol. 5:239. Dennell, R. i960. Integument and exo- skeleton. Chap. 14, pp. 449-472. In: T. H. Waterman, ed., Physiology of Crustacea, Vol. 2. Academic Press, New York. Drach, P. 1936. L’eau absorbee au cours de 1’ exuviation, donnee fondementale pour 1’ etude physiologique de la mue. Definitions et determinations quantitatives. C. R. Acad. Sci. 202:1817. 1939. Mue et cycle d’intermue chez les crustaces decapodes. Ann. Inst. Oceanogr. 19:103. and M. Lafon. 1942. Etude biochim- ique sur le squelette tegumentaire des de- capodes brachyoures (variation au cours du cycle d’intermue). Arch. Zool. Exptl. 82:100. Drilhon, A. 1935. Etude biochimique de la mue chez les crustaces brachyoures ( Maia squinado ). Ann. Physiol. Physicochim. Biol. 11:301. Ennever, J. 1963. Microbiologic calcification. In: Comparative biology of calcified tissue. Ann. N. Y. Acad. Sci. 109(1) :4-13. Ferro, P. V., and A. B. Ham. 1957,2. A simple spectrophotometric method for the determin- ation of calcium. Am. J. Clin. Pathol. 28:208. 1957A A simple spectrophoto- metric method for the determination of cal- cium. II. A semimicro method with reduced precipitation time. Am. J. Clin. Pathol. 28:689. Geyer, R. P., and E. J. Bowie. 1961. The direct determination of tissue calcium by flame photometry. Anal. Biochem. 2:360. Greenfield, L. J. 1963. Metabolism and con- centration of calcium and magnesium and precipitation of calcium carbonate by a marine bacterium. In: Comparative biology 208 PACIFIC SCIENCE, Vol. XXI, April 1967 of calcified tissue. Ann. N. Y. Acad. Sci. 109(1) :23-45. Guyselman, J. B. 1953. An analysis of the molting process in the fiddler crab, Uca pugilator. Biol. Bull. 104:115. Hayes, D. K., L. Singer, and W. D. Arm- strong. 1962. Calcium homeostatic mech- anisms and uptake of radioisotopes in the lobster. Am. J. Physiol. 202(2) : 3B3- Hiatt, R. W. 1948. The biology of the lined shore crab, Pachygrapsus crassipes Randall. Pacific Sci. 2:135. Hollett, A. 1943. Relation between moult cycle and phosphorus content of blood and muscle. J. Fish. Res. Bd. Canada 6:152. Isenberg, H. D., L. S. La vine, M. L. Moss, D. Kupferstein, and P. E. Lear. 1963. Calcification in a marine coccolithophorid. In: Comparative biology of calcified tissue. Ann. N. Y. Acad. Sci. 109(1) : 49-64. Kincaid, F. D., and B. T. Scheer. 1962. Hormonal control of metabolism in crusta- ceans. IV. Regulation of tissue composition of Hemigrapsus nudus to intermolt cycle and sinus gland. Physiol. Zool. 25:372. Krishnan, G. 1950. Sinus gland and tyrosinase activity in Carcinus maenas. Nature 165:364. Kyer, D. S. 1942. The influence of the sinus glands on gastrolith formation in the cray- fish. Biol. Bull. 82:68. Lafon, M. 1948. Nouvelles recherches bio- chimiques et physiologiques sur le squelette tegumentaire des crustaces. Bull. Inst. Ocean- ogr., Monaco 939:1. McWhinnie, M. A. 1962. Gastrolith growth and calcium shifts in the freshwater crayfish, Orconectes virilis. Comp. Biochem. Physiol. 7:1. Miyawaki, M., and N. Sasaki. 1961. A pre- liminary note on uptake of Ca45 by the hepatopancreas of crayfish, Procambarus clarkii. Kumamoto J. Sci., Ser. B, Sec. 2, 5:170. Needham, A. E. 1946. Ecdysis and growth in Crustacea. Nature 158:667. Neto, B. M. 1943. Calcium content of the carapace of Uca maracoani (Latreille). Anal. Soc. Biol. Pernambuco 4:31. Numanoi, H. 1939. Behavior of blood calcium in the formation of gastroliths in some de- capod Crustacea. Japan. J. Zool. 8:357. Odum, H. T. 1957. Biogeochemical deposition of strontium. Inst. Mar. Sci. 4(2): 39. Paul, J. H. and J. S. Sharpe. 1916. Studies on calcium metabolism. I. The deposition of lime salts in the integument of decapod Crustacea. J. Physiol. 50:183. Prenant, M. 1927. Les formes mineralogiques du calcaire chez les etres vivants, et le prob- leme de leur determinisme. Biol. Rev. Cam- bridge Phil. Soc. 2:365. Rizzo, A. A., D. B. Scott, and H. A. Bladen. Calcification of oral bacteria. In: Compara- tive biology of calcified tissue. Ann. N. Y. Acad. Sci. 109(1) :l4-22. Robertson, J. D. 1937. Some features of cal- cium metabolism of the shore crab {Car emus maenas Pennant). Proc. Roy. Soc. London B, 124:162. I960. Ionic regulation in the crab Car emus maenas (L.) in relation to the molt cycle. Comp. Biochem. Physiol. 1:183. Sather, B. T. 1966. Observations on the molt cycle and growth of the crab, Podophthalmus vigil (Fabricius). Crustacena 11(2): 185- 197. Scheer, B. T., and M. A. R. Scheer. 1951. Blood sugar in spiny lobsters. Part I of the hormonal regulation of metabolism in crusta- ceans. Physiol. Comp, et Oecol. 2:198. Schonborn, E. G. von. 1912. Weitere Unter- suchungen fiber den Stoffwechsel der Krus- tazeen. Zeitschr. f. Biol. 57:543. Scudamore, H. H. 1942. Hormonal regulation of molting and some related phenomena in the crayfish, Cambarus immunia. Anat. Record 84:514. — 1947. The influence of the sinus glands upon molting and associated changes in the crayfish. Physiol. Zool. 20:187. Snedecor, G. W. 1959. Statistical Methods Applied to Experiments in Agriculture and Biology. Iowa State College Press, Ames. 534 pp. Travis, D. F. 1954. The molting cycle in the spiny lobster, Panulirus argus Latreille. I. Molting and growth in laboratory maintained individuals. Biol. Bull. 107:433. 1955 a. Ibid. II. Pre-ecdysial histological and histochemical changes in the hepato- pancreas and integumental tissues. Biol. Bulk 108:88. Calcium and Phosphorus Metabolism of P. vigil — Sather 209 1955 b. Ibid. III. Physiological changes which occur in the blood and urine during the normal molt cycle. Biol. Bull. 109:484. 1957. Ibid. IV. Post-ecdysial changes in the hepatopancreas and integumental tis- sues. Biol. Bull. 113:451. I960. Matrix and mineral deposition in skeletal structures of the decapod Crustacea (Phylum Arthropoda), pp. 57-116. In: R. F. Sognnaes, ed., Calcification in Biological Systems. Am. Assoc. Adv. Science. Publ. No. 64. 1963. Structural features of mineraliza- tion from tissue to macromolecular levels of organization in the decapod Crustacea. In: Comparative biology of calcified tissue. Ann. N. Y. Acad. Sci. 109(1 ) :177-245. Urist, M. R. 1962. The bone-body fluid con- tinuum: calcium and phosphorus in the skeleton and blood of extinct and living vertebrates. Perspect. Biol. Med. 6:75. Vinogradov, A. P. 1953. The Elementary Chemical Composition of Marine Organisms. Sears Found. Mar. Res., Memoir II, pp. 390-462. On Some Gas trocoty line (Monogenoidean) Parasites of Indian Clupeoid Fishes? Including Three New Genera R. VlSWANATHAN UNNITHAN1 ABSTRACT: Seven species of monogenetic trematodes, including the two geno- types, Engraulicola forcepopenis George, 1961 and Engraults cohtna thru socles (Tripathi, 1959), are recorded. All seven of these atypical gastrocotylines belong to the subfamily Gastrocotylinae s.s. and are parasitic on clupeoid fishes. Four species in the present collection, viz., Engraulicola micropbaryngella sp. n., Engraulixenus malabaricus gen. et sp. n., Engrauliphila grex gen. et sp. n., and Engraults cobina triaptella sp. n., were collected from fishes of the family En- graulidae, while an entirely new type, Pellonicola elongata gen. et sp. n., was obtained from Clupeidae. The tendency to unilateral inhibition of the clamp rows is incomplete in all these atypical gastrocotylines, and all are characterised primarily by their clamp structure. Diagnostic characters, with special reference to the hap tor (its adhesive units or clamps and anchors), the male terminalia, vaginal complex, and other salient features which appear for each species. Some gastrocotylib worms have been found on the gills of clupeoid fishes at Mandapam Camp. Their clamp structure shows them to be allied to Gastrocotyle and Pseudaxine . The tendency to develop a unilateral haptor is another common feature. But of the 32 known species of Gastrocotylidae, and 8 new species described by me (in press), all 40 are parasitic on scombroid fishes (including Carangidae). Indeed, it is usual to find these more highly evolved Monogenoidea on the more highly evolved fishes, while the simpler clupeoids are typically parasitized by Mazocraeidae. The oc- casional excursions across the phylogenetic trees of the hosts for each of these families of parasites have been discussed by Sproston (in press). The present collection adds support to her hypothesis, since it contains four new forms from Engraulidae and one entirely new type from Clupeidae. Collection and treatment of specimens was performed as described in a previous paper (Unnithan, 1957:28-29). 1 Indian Ocean Biological Centre, Ernakulam-6, India. Formerly research student in the Marine Bio- logical Laboratory of the University of Kerala, Tri- vandrum, and at the Central Marine Fisheries Re- search Institute, Mandapam Camp, where most of this work was done. Manuscript received March 8, 1966. to be taxonomically important, are given All of these atypical clupeoid parasites belong to the subfamily Gastrocotylinae sensu stricto , hitherto containing only Gastrocotyle v. Ben. et Hesse, 1863, Chauhanea Ramalin- gam, 1953, and Y amaguticotyla Price, 1959= They are characterized primarily by their damp structure (Unnithan, 1961b). There are paired braces in the posterior region of the clamp capsule, as in all Gastrocotyloldea (as distinct from Microcotyloidea), but the clamps them- selves are bilaterally symmetrical; and, unlike the subfamily Priceinae, for instance, to which Pseudaxine belongs, there are no complex dorsal shields to other sclerites developed in the dorsal wall of the capsule in association with the median spring, nor are there riblike thickenings in the capsule walls. While in Gastrocotyle spp. the ventral arm of the spring is often doubly bifid, its ends sometimes form articulated stmts to the jaw rami. This condition is not found in the other genera. In all of them (with the doubtful exception of Chauhanea and Y amagu- ticotyla), the dorsal arm of the spring bears a forked appendix associated with noncuticula- rized ligaments, presumably a transitional condi- tion to that in Priceinae. The tendency to unilateral inhibition of the clamp rows is complete in Gastrocotyle spp. (also in Pseudaxine in Priceinae and in some 210 Gastrocotyiine Parasites of Indian Clupeoids — Unnithan 211 Axinidae), but it is incomplete in all the atypical clupeoid parasites described in the present study. The partition of genera is based on criteria which have appeared to be valid for genera of other subfamilies of the higher Monogenoidea, namely: (1) The relative encroachment of the clamp rows up the body proper (as distinct from a tail-like haptor quite distinct from the body, for example in Chauhanea ), which determines the zone of pivoting in all the possible feeding attitudes of the attached worm. If this pivoting zone occurs in the thicker gonad zones, then the torque set up will tend to a somatic asym- metry (e.g., Rngraulis cobind) ; if, on the other hand, it is near or beyond the end of the testes zone (as in Engraulicola , Pellonicola, and Pseudaxine) , then the highly contractile body may be irregular in outline at any one moment, but very little permanent strain would be registered, and there is no structural asymmetry in the body proper. Nor is this asymmetry present in forms where the pivoting zone occurs anterior to the gonad zone: the fore- body alone in these forms is free to bend, and again no true somatic asymmetry is developed (e.g., Gastrocotyle, Engraulixenus, and En- graultphila) . (2) The degree of suppression of one side of the haptor. It is considered that there is a greater difference between complete suppression and the inhibition of all but one, or of all but two or three clamps, than between inhibi- tion of only about half the clamps. (3) The form of the anchors, as has been shown by Llewellyn (1957: Figs. 22 and 23 for Gastrocotyle and Pseudaxine , and similarly in 1959: Figs. 8 and 9), is regarded as a generic character. The useful differences are the relative lengths of the handle (main root) and hook, of the spur (secondary root), and the approximate segment of a circle represented by the sickle-like hook. These characters are fundamental, since they are developed in the postlarval stages of the onomiracidium, and persist throughout life unless anchors are shed. The persistence of other larval anchors may be a specific character. (4) Additional sclerites associated with the penis, e.g., the peculiar forceps on the penis head. The only possible analogue is found in Heterapta Unnithan, 1961 (Heteromicrocotyli- dae), where the two spines appear to have a much deeper origin and are straight and much longer, probably functioning as vagino-dilators. (5) The occurrence of a single median dorsal vagina, or of paired vaginae opening laterally, at various levels, and their separate confluence into the lateral vitelline ducts, or the intervention of a median duct and, in some cases, the direct course of this to the ootype (as is more usual with an unpaired vagina) . (6) The relative size of the oral pouches and the pharynx (expressed as percentage of mean diameters), and the absolute size range of the latter and its shape. In view of the possibility of a wide array of related forms being discovered on tropical clupeoids in the future, I hesitate to give formal generic and specific definitions for these new forms, but prefer rather to list their diagnostic characters, with particular reference to the six criteria listed above, and other salient characters which appear taxonomically important in each case. Engraulicola micro pharyngella sp. nov. Figs. 1-6 Four specimens of this new gastrocotyiine were collected from the gills of two mature female Anchoviella commersonii (Lacepe.) ex- amined at Trivandrum on August 14, 1957. All of the specimens except the one whole mount figured were broken while staining. The description is based mainly on this well pre- pared specimen (Fig. 1), but comparative measurements on the broken ones are also included. Body proper essentially bilaterally symmetri- cal, anterior and posterior ends narrower ; "neck” long and slender, about one-sixth of the total worm; haptor triangular with a short terminal lappet (Fig. 1). Total length 1.3-1. 9 mm and maximum width 0.42-0.64 mm im- mediately in front of the ovary. Mouth subterminal and crescentic ; oral pouches thin-walled, spherical, 16-20 p, in diam- eter, opening into the buccal cavity; pharynx (unusually) smaller than oral pouches, spheri- cal and thin-walled, 14 p in diameter (ratio of oral pouch to pharynx = 130%) ; oesophagus 212 PACIFIC SCIENCE, Vol. XXI, April 1967 Figs. 1-6. Engraulicola micropharyngella sp. nov. 1, Complete worm, dorsal view; 2, clamp, dorsal view; 3a and 3b, proximal and distal anchors; 4, penis with the corona of spines and the forcepiform process; 3, penis of another specimen; 6, ootype and ovarian region, dorsal view. Gastrocoty line - Parasites of Indian Clupeoids — XJnnithan 213 long, narrow, unbranched, and bifurcating into the intestinal crura at the level of the male genital pore; crura with long irregular outer and short simple inner branches; both crura extend into the haptor and terminate at dif- ferent levels, one extending farther than the other. In the haptoral region the crura are un- branched and the ends are not confluent. In the specimen examined the haptor bears a long row of 27 clamps on one side, and a single clamp on the other. It is inclined to the body axis at 45°-60°, and is about 35% of the body length. All the clamps are of the same structure, though they vary in size from 16 X 20 |i to 28 X pi. The anteriormost clamp of the long (left) row is the smallest and the single clamp on the right side is more or less of the same size. It is reasonable to expect amphitypy, the long uninhibited row being on the left or right according to the location of the worm on the gill; further col- lection may show this relationship. Clamps typically gastrocotyloid in structure: dorsal arm of median spring short and narrow with a thin V-shaped cuticular projection from its distal end; ventral arm of median spring long and narrow, distally bifurcated; base of clamp with a thin, narrow, heavily cuticularized hinge ligament on each side, connecting the median spring with the base or region of articulation of the jaw sclerites; dorsal and ventral jaw sclerites of the two sides sym- metrical; dorsal arm of ventral jaw reaches to the level of the bifurcation of the dorsal arm of the median spring ; oblique sclerites ("braces”) long and narrow, with their distal ends touching in the median line, where their backwardly bent tips form an articulating sur- face, just within the dorsal jaw (Fig. 2). Terminal lappet small and cylindrical, 21 pi broad and 36 p long, demarcated from the rest of the haptor by a small constriction and armed with two pairs of symmetrically placed anchors : anterior pair typically sickle-shaped, 24 \i long, the hook being about three-eighths of a circle. The spur root is short and bent posteriorly to- ward the point of the sickle, while the handle is less than half the total length (75% of the sickle) (Fig. 3a). It is interesting to note that the spur pointing posteriorly toward the point is also characteristic of the large anchor of Gastrocotyle (Llewellyn, 1957: Figs. 12 and 22). The posterior pair is exceptionally small, with sharply reflexed hooked ends; their over- all length is 8 pi (Fig. 3b). The testes are all postovarian; there are about 27 small spheroidal follicles in 2-3 irregular files extending more or less to the hind end of the shorter intestinal crus or halfway down the haptor. The vas deferens runs forward, curving to the right around the ovarian zone beyond which it becomes median and opens into the base of the penis, apparently without a vesicular dilatation. The atrium masculinus is in the zone of the intestinal bifurcation. The cuplike penis has a thick muscular wall, its rim armed with a corona of 8-10 hooked spines with their tips converging. From the centre of the penis cup on the penis head is a forceps-like, lightly cuticularized structure 13-16 pi long, much longer than the penis hooks, and projecting slightly beyond the penis corona (Figs. 4 and 5) . No collar was seen like that described by George (1961) for the genotype. The ovary is an inverted U, its field 210 X 63 ji, situated in front of the testes; its outer longer limb is narrow and the distal inner limb is thicker and contains larger ova. The oviduct arises from the distal end of the ovary, runs backward, and opens into the ootype through a short narrow basal loop (Fig. 6) ; the uterus arises from the ootype close to the oviduct, runs forward along the median line, and opens near the male genital opening. Eggs spindle- shaped, 40 X 24 pi. Vitellaria massive, extending from the level of intestinal bifurcation to the tips of each intestinal crus and covering the crural branches, not confluent across the median line; vitelline follicles spherical, 8-10 pi in diameter. The transverse vitelline duct lies at the level of the first third of the ovarian zone, and the median vitelline duct tapers slowly until it reaches the ootype. The ootype is surrounded by few scattered Mehlis gland cells. The genito-intestinal canal curves toward the right side and opens into the right crus in the midoviduct zone. The median dorsal vagina is unarmed and is situated in the angle of the intestinal bifurca- tion, immediately behind the muscular un- armed rim of the atrium masculinus and sur- 214 PACIFIC SCIENCE, Vol. XXI, April 1967 rounded by small spherical gland cells. The vaginal duct is narrow and runs backward dorsally along the median line parallel to the vas deferens and opens into the vitelline ampulla, independently of the vitelline duct (Fig. 6). relationships of Engraulicola micropharyn- gella SP. nov.:2 1. Engraulicola is characterized by the gen- eral shape of the body, which resembles a "treed” riding boot, the handle of the boot tree being represented by the slender neck, the main clamp row the sole of the boot (with the metahaptor as the heel), and with the single clamp of the inhibited row suggesting the toe cap of the boot. In other genera a toelike projection is not developed. 1. In E. forcepopenis George, 1961 the foot and toe are nearly at right angles to the body proper and the testes scarcely enter the foot, the zone of pivoting being behind the testis zone. In E. micropharyn- gella sp. nov. the foot is more tapered to the toe and is only 45°-60° to the body axis, and anteriorly a few single- file testes enter the foot and are included in the zone of pivoting. The heel is a little thicker, but no haptoral wing with special gut branches is developed. ii. The haptor is less than 50% of the axial length in the genotype, but only about 35% in the new species. 2. In the three larger individuals of E. force- popenis bearing shelled eggs, the haptoral fringe had 44, 39, and 33 clamps, and the two smaller individuals (total length 1.3— 1.5 mm) had only 21 and 25 clamps. In the unbroken specimen of E. micropharyngella (1.3 mm long), there were 27 clamps; the broken in- dividuals (1.4— 1.8 mm long) had 28, 29, and 32 clamps. In all individuals of both species there is a single clamp on the toe cap. i. The clamp is wider than long in the former species (length/width == 55- 60%), and in the latter is relatively not quite so wide (70%). In both, the soli- 2 Generic characters are indicated by Arabic nu- merals, specific characters by small Roman numerals. tary clamp is nearly as long as wide, but smaller than those of the other side. The mean diagonal of the solitary clamp in the genotype is 33.2 p (calculated from the mean of the square root of the product of diameters, yj\ X w), and that of the new species is only 18 p, i.e., about 54% of the size of the single clamp in the genotype. ii. The appendix of the dorsal arm of the median spring in E. forcepopenis is shown as Y-shaped with a short stem. In the present one it is V-shaped with a minute base only, and the posterior ends of the braces are bent back as opposable knobs. 3. Of the two pairs of persistent anchors, the anterior is typically sickle-shaped in both; in E. forcepopenis the sickle is about half a circle and the hook nearly equal, but in E. micropharyngella the sickle is only about three- | eighths of a circle and the length of the handle is only about 75% that of the sickle. The total length of the anterior anchors in the former is i 29 p, and 24 p in the latter. The posterior simple, hooked anchors are much smaller in j the new species (13.6 p and 8 p, respectively). 4. The forceps on the penis head are prac- tically identical in both form and size, but the | collar observed in the genotype, projecting | ventrally from the atrium masculinus, has not been seen in the present material, where the rim of the atrium is a simple flat muscular ring. i. In E. forcepopenis, though it is larger, there are barely half the number of testic- ular follicles that are present in the new species. E. micropharyngella has no paro- varian follicles, while one or two are found in E. forcepopenis. ii. The spines of the genital corona of E. forcepopenis are invariably 12, but only 8-10 are present in E. micropharyngella. 5. While both species have a single median vagina with a duct direct to the ootype, in E. forcepopenis the vulva is halfway between the male genital pore and the vitelline ducts, and in E. micropharyngella it is strikingly farther forward, lying immediately behind the male genital pore. Gastrocotyline Parasites of Indian Clupeoids — Unnithan 215 6. The relative size of oral pouches (mean diagonal from -\/l X w) ancJ pharynx in the genotype is 50—77%, the pharynx as usual being ovoid and larger than the ovoid oral pouches. But in E. micro pharyngella, while the oral pouches are nearly round the ovoid pharynx is minute, the former 130% of the latter. Hence this most obvious specific char- acter is indicated in the name. In view of these important characters I agree with George (1961) in his creation of En- graulicola, with E. forcepopenis as the type. The above description was written before the paper by K. C. George was available to me, but we had previously agreed on the nomencla- ture of his material (described some years earlier than mine) from the same geographical region (South Malabar coast). Engraulixenus malabaricus gen. et sp. nov. Figs. 7-11 Several specimens of this new species of en- graulid parasite were obtained from the gills of T hr is socles malabaricus (Bloch) examined at Trivandrum on July 26, 1955. Four fishes were examined and all were infected by the new parasite as well as by a large number of two different species of Mazocraeidae. Out of the 52 Monogenoidea obtained, 8 specimens be- longed to the present species. This worm is foot-shaped, with a distinct heel and a long slender forebody, the total length being 1.71-2.43 mm and the maximum width 0.45-0.5 mm (Fig. 7). Mouth subterminal, without especially glan- dular or muscular lips ; oral pouches spherical, 24-28 p in diameter ; pharynx median, very large, elongated ovoid, 64 X 36 p-80 X 38 p; oral pouches not more than 40% of pharynx (by mean diagonals) ; oesophagus narrow, 0.13-0.19 mm long, bifurcating into the in- testinal crura behind the male genital pore; crura exceptional, with 2-3 cross connections bridging across the median line and without much outer branching, the dilated ends (un- branched in posterior third) extending to dif- ferent levels. At the anterior region of the long clamp row, the crus of that side has a few wider lateral branches, forming the base of what is probably a metahaptoral wing. An oblique con- Figs. 7—11. Engraulixenus malabaricus gen. et sp. nov. 7, Complete worm, ventral view ; 8, clamp of the long row, dorsal view; 9, proximal anchors; 10, distal anchors; 11, penis with the corona of hooks and for- cepiform process. nection also occurs between the distal ends of the crura, behind the testes zone, in most of the specimens (Fig. 7). The haptor occupies about 50% of the total length; a fleshy flange adherent to the body (yet with an increasing tendency to diverge from it) forms the wing referred to above. The long clamp row has 42-49 clamps, each 32 X 48 p-32 X 66 p; the short row has two nearly sessile clamps 24 X 28 [h and 28 X 36 p ; lappet 64 X 28 p, armed with two pairs of anchors. The anterior pair (Fig. 9), 32 p long, have a shape different from that in Gas- 216 trocotyle and Engraulkola : the hook is barely a quarter of a circle, with the handle consider- ably longer, while the stout spur is at right angles to the handle. The posterior anchors (Fig. 10) are also unusual in having an in- cipient spur behind the short hook; the total length of these anchors is 16 p. The clamp structure shows slight variations from that of Engraulkola : there is a marked gradation of size toward the middle of the long row. The clamps are much wider than long, except for the first and next anterior clamps, which, like the two remnant primaries of the in- hibited side, are more nearly squarish (Fig. 8). The 20-39 testes are oval, 28 X 40 p-36 X 48 p, arranged in 4-5 files in the intercrural field behind the ovary but with a few parovarian testes on the left side. The narrow vas deferens arises from the postovarian testes, extends for- ward on the left side of the body, parallel to the vitelline duct and enlarges into a seminal vesicle near the anterior end of the median vitelline duct. From the anterior margin of the seminal vesicle, the vas deferens continues for- ward and opens into the base of the penis, some distance anterior to the intestinal bifurcation. The penis is muscular and armed with a corona of 12 recurved hooks around its bulb-like base, and there is a forceps-like double spine within (Fig. 11). The forceps spines appear rather shorter than in Engraulkola. Male genital pore is strengthened by a rim of radial muscle fibres but is without a projecting collar, and is situ- ated at about 0.27 mm from the anterior end of the body. The inverted U -shaped ovary occupies a field in the middle of the body’s length and it is about one-tenth as long as the latter, the ova as usual becoming bigger toward the oviduct. The oviduct arises from the distal end of the ovary and enlarges into a small sphincter-like ovijector which continues through the ootype region and opens into the fertilization chamber near the vitelline ampulla. The uterus can be traced forward from the anterior margin of the ootype, parallel to the vas deferens, and it opens into the unarmed uterine pore situated immediately in front of the male pore. The paired vaginal pores are unarmed and submarginal and lie in front of the ovarian zone at two-thirds the distance from the male PACIFIC SCIENCE, Vol. XXI, April 1967 terminalia to the anterior end of the ovary. The vaginal ducts are S-shaped and in the specimens examined were distended with sperm; they run backward and unite in the zone of transverse vitelline ducts to enter the wide median vitelline duct which extends backward, narrows poste- riorly, and opens into the ootype, in the small vitelline ampulla. Thus, there is no true median vaginal duct. The vitellaria extend from the region of in- testinal bifurcation to the distal ends of the crura and are not confluent across the median line even in the region of the crural bridges; the spherical follicles are 8-12 p in diameter. The transverse vitelline ducts meet along with the lateral vaginal ducts immediately anterior to the ovarian zone, to form the median vitel- line duct which also functions as a vaginal duct. The genito-intestinal canal connects the base of the ootype with the right intestinal crus, passing sharply obliquely forward across the proximal region of the ovary; its union with the right crus is in the midovarian zone (i.e., more an- terior than is usual). relationships OF Engraulixenus malabari- CUS GEN. ET SP. NOV.: 1. Engraulixenus has an elongated tapering body, slender anteriorly, with a long foot- shaped hindbody tapering backward and which has an unusually prominent heel with a spurlike extension; this region is the typical haptoral wing (perhaps a metahaptor: see Unnithan, 1967^), which receives short wide branches from the adjacent intestinal crus which branches more or less profusely in the anterior part of the wing; these gut branches carry with them vitelline follicles. The haptor extends slightly obliquely at only 25° to the body axis for at least 50% of its length; thus the zone of pivot- ing of the attached worm is between the wide testis zone and the ovarian zone. The torque strains set up have not greatly disturbed the symmetry of the body proper but doubtless account for the haptoral wing and subjacent lateral field. The arc of feeding exploration is evidently extensive because of the slender con- tractile forebody. Thus, Engraulixenus is less symmetrical than Gastrocotyle. 2. The inhibited clamp row retains two of its primary clamps in all individuals. Gastrocotyline Parasites of Indian Clupeoids- -Unnithan 217 i. The long clamp row in mature worms, 1.7-2. 4 mm long, bears 42-49 nearly sessile units, closely set, the posterior edge of one touching the anterior edge of the next one. ii. The larger clamps are at least as wide as long and are the widest of any described in the present study. iii. The dorsal appendix on the spring is a stalked stout U -shaped piece with parallel arms not divergent as in most of its rela- tives. iv. The ventral arm of the spring is not truly bifurcated and is slender throughout. v. The braces are bent posteriorly for mutual articulation. 3. Of the two pairs of persistent anchors, the anterior are characteristically shaped, with the handle markedly longer than the hook, which is barely one-quarter of a circle, and with a stepped conical spur at a right angle to the handle. The anterior anchors are more slender and shorter than in Engraulicola, and the pos- terior anchors have hooks which recurve for only one-third the length of the blade, and have an incipient spur and a stout handle. They are just half as long as the anterior pair. 4. The forceps on the penis head are similar to those in Engraulicola , but they may be rela- tively stouter and shorter; there is no collar projecting from the rim of the atrium mascu- linus. i. The penis spines are sigmoid and 12 in number. ii. The 20-39 testes are massive in 4-5 files anteriorly. iii. A vesicula seminalis is present in front of the ovarian zone. 5. In the paired vaginae the vulvae are supramarginal and the lateral vaginae join the transverse vitelline ducts near their confluence, so that there is no true median vaginal duct. i. The vulvae are situated at two-thirds the distance from the male genital pore to the anterior end of the ovary. The vitel- line ducts are usually long and oblique and they become confluent into the me- dian vitelline duct distinctly anterior to the ovarian zone. 6. The pharynx is exceptionally elongated and ovoid, the longest (80 p) in the whole group; the mean diagonal of the spherical oral pouches is 40% or less that of the pharynx. 7. Exceptional intercrural bridges occur twice or thrice in the forebody, and often in the post- testicular zone there is an oblique bridge. This is the most obvious generic feature, but it is perhaps less important than are the preceding criteria taken together. i. The ends of the crura are subequal and markedly inflated, the longer being on the inhibited side, and reaching to opposite the sixth or seventh clamp from the pos- terior end. The specific name is derived from the hoT and locality, the South Malabar coast. Engrauliphila grex gen. et sp. nov. Figs. 12-17 Specimens of this new gastrocotyline genus were found swarming on the gills of Thrissocles dussumieri (Val.) examined at the Southern Indian Marine Biological Laboratory at Trivan- drum on October 5, 1955 and at Ayirumthengu on September 8, 1955. Those from Trivandrum had a multiple infection including a relative, Engrauliscobina triaptella sp. n., while those from Ayiramthengu were infected with the present species only. Numerous specimens were collected from a single fish, a minimum of at least 50 being very common. However, many fishes of the same host species examined at Vizhinjom (another marine biological station, 8 miles south of Trivandrum) on August 19, 1954 were not infected by E. grex. There the characteristic parasite was Engrauliscobina triap- tella. This worm is foot-shaped with a spurred heel, but the "leg” comprising the anterior half is only about half as wide as the stout hind- body, which tapers evenly to the terminal lap- pet; total length 1.3-1. 8 mm and maximum width 0.25-0.3 mm, including the haptoral wing or spur of the heel (Fig. 12). The subterminal mouth is wide, with an an- terior circlet of scattered sticky cells. The spher- ical oral pouches are 24-32 p in diameter, with thin walls but with long muscle fibres extending backward ; the relatively long ovoid pharynx is 218 PACIFIC SCIENCE, Vol. XXI, April 1967 40 X 60 p-44 X 72 [i, with thick walls and radial muscle fibres strengthening it; the mean diagonals of the oral pouches are about 50% of the diagonal of the pharynx. The narrow oesophagus is 0.12-0.18 mm long and bifur- cates into the intestinal crura just behind the male terminalia. These crura lack both intercrural branches and bridges, and have very few outer branches; they terminate posteriorly at different levels in the posterior third of the hindbody, where their ends are slightly inflated. The crus on the inhibited side of the haptor is the longer, reaching to opposite the fifth clamp or so from the posterior end of the longer row, well in front of the two clamps of its own side. The intestinal crus on the side of the body Figs. 12-17. Engrauliphila grex gen. et sp. nov. 12, Complete worm, ventral view; 13, clamp, ventral view; 14, anchors; 13, hind end of haptor with the anchored lappet; 16a and 16b, male genital pore with the armed penis; 17, ootype and ovarian region, ventral view. Gastrocotyline Parasites of Indian Clupeoids — Unnithan 219 bearing the main row of clamps has more ex- tensive outer branches, particularly in the hap- toral wing, which are accompanied by vitel- laria. This winglike expansion of the haptor is similar to that in Engraulicola micropharyngella described above, but although the present worms are on the whole smaller, the haptor is stouter. All the clamps have relatively long muscular stalks (about as long as the width of the clamps) which project sideways in close file. As in the previous species, the clamps increase in size toward the middle of the row, where there is a slight irregularity. This may indicate the end of the euhaptor and the beginning of the metahaptor, which tends to grow with in- creasing independence of the body proper, its anterior part being free from it— as the meta- haptoral wing (Fig. 12). The long row makes an angle of only about 30° with the body axis. There are always two remnant clamps on the inhibited side, smaller (20 >< 28 p-28 X 36 p) and resembling their opposite primaries. The long row is more than half the length of the worm, with 40-48 clamps, each 28 X 40 p— 28 X 60 p. The terminal lappet is trapezoidal, 0.04 X 0.02 mm, with two pairs of anchors. The an- terior anchors are 28 p long and slender; the hook is less strongly curved (only one-eighth to a quarter of a circle), with a short stepped spur, projecting at right angles at the end of the slender handle which is more than half the total length (Figs. 14 and 15). Thus there is a strong resemblance to the anterior anchors of Engraulicola micropharyngella. The posterior anchors generally resemble those of the last species, but the simple hook is barely one-third as long and is less curved (about half a circle), while in the last species the hook was more than half a circle (cf. Figs. 14 and 10). The clamp structure is quite distinctive in detail (Fig. 13) ; the larger clamps are as much as twice as wide as they are long. In sharp con- trast to the two previous species, the ventral arm of the median spring is broad and widely bifurcated in a thick pointed fork. The dorsal arm carries a V-shaped stout appendix with : arms widely diverging. The braces are not bent i at the posterior ends to form articulating facets, as in the preceding species. The testicular zone is entirely flanked by the haptor flange, which extends over the hind end of the ovary also, so that the pivoting axis is in a thicker part of the worm and the torque here would account for the broad haptoral wing. The testicular zone is not involved in the torque and the 2-4 files are rather regular and compact (Fig. 12). There are 15-31 testes, ovoid or spheroidal, with one or two parovarian testes. The vas deferens is long and wide, arising from the median anterior testicular zone and extend- ing forward as a zigzag duct to open into the base of the penis. The anterior extremity of the vas deferens, before it joins the penis, is straight, forming the ejaculatory duct; this has a poste- rior dilatation between the horns of the vaginae functioning as the seminal vesicle (Fig. 17); the penis is muscular, ventral, situated about 0.18-0.2 mm from the anterior end, and bear- ing a corona of 10 sharp hooks, but its tip lacks forceps spines. The atrium masculinus is un- armed and circular, with a muscular rim and a thick ring of radial muscle fibres but no project- ing membranous collar (Figs. 16a and 16b). The ovary takes the form of an inverted U with a long narrow proximal (outer) limb and a wide short distal (inner) limb, situated in the middle third of the body, in front of the tes- ticular zone. The thin and narrow oviduct de- scends from the distal end of the ovary and opens into the fertilization chamber, through the well-developed, spindle-shaped ovijector (Fig. 17). The ootype is surrounded by closely packed Mehlis gland cells. The uterus ascends from the ootype to open into the unarmed uterine pore in front of the atrium masculinus. Eggs were seen in only one of the specimens. The vitellaria extend from the intestinal bi- furcation to almost the distal ends of the crura ; they are not confluent across the median line, and their spherical follicles are 8-10 p in di- ameter. The transverse vitelline ducts meet to form the median duct at the anterior quarter of the ovarian zone, as in E. triaptella , but here they are joined by the lateral vaginal ducts. The median vitelline duct is broad anteriorly and tapers posteriorly to open into the vitelline am- pulla, which is feebly demarcated in most of the specimens. The genito-intestinal canal is well differentiated, arising from the ootype close to the ovijector, and running obliquely into the right crus. 220 The two dorsal vaginal pores are unarmed, one in each midlateral field in front of the ovary in the anterior part of the middle third of the body, well in front of the transverse vitelline ducts. The lateral vaginal ducts are packed with sperm cells and twisted in S -shaped sinuous ducts which run backward to unite with the transverse vitelline ducts near their junction with the median vitelline duct. relationships of Engrauliphila grex gen. et sp. nov.: 1. Engrauliphila has a haptor-body relation similar to that in Engraulixenus. i. The stouter body is provided with a somewhat thicker haptoral flange and the clamps have more muscular and longer stalks. ii. In comparable-sized worms, the haptor is more extensive, reaching into the ovarian zone, so that the zone of pivoting is in a thicker region and the resulting torque would account for the relatively more massive (metahaptoral) wing. iii. The length of the clamp row is more than 50% that of the relaxed worm. 2. The inhibited clamp row retains only two primary clamps in all specimens. i. The long clamp row bears 40-48 trans- versely elongated stalked clamps in close file. ii. The dorsal appendix on the spring is stout and V-shaped. iii. The ventral arm of the spring is broad, splayed, and bifurcate. iv. The braces do not have bent articular ends. 3. The two pairs of persistent anchors are of distinctive shape, both less curved than in En- graulixenus and entirely unlike those of Gastro- cotyle and Engraulicola. 4. The penis head is devoid of forceps and a collar is lacking round the atrium. i. The corona consists of 10 sharp divergent spines, but no sigmoid spines. ii. The 15-31 testes are in 2-4 compact files, with 1 or 2 parovarial. iii. There is a vesicula seminalis in the pre- ovarian zone. PACIFIC SCIENCE, Vol. XXI, April 1967 5. The vaginae are paired and lie in mid- lateral fields on the dorsal side; there is no median vaginal duct, since they join the trans- verse vitelline ducts near their confluence, as in Engrauliphila ; the vulvae are not supramar- ginal and are situated much nearer the ovary. i. The vulvae are less than an ovary length in front of the ovarian zone. ii. The transverse vitelline ducts are situated at the level of the anterior quarter of the ovarian zone. 6. The pharynx is elongated ovoid, and the oral pouches are spheroidal and much smaller. i. The mean diagonal of the latter is about 50% that of the pharynx. ii. The crura lack inner branches and there are no intercrural bridges at all. iii. The unequal ends of the crura are only slightly dilated. The assemblage of differences in the generic criteria taken together are in sufficient contrast to those of Engrauliphila for Engraulixenus to j be recognized as distinct. The specific name j grex refers to the exceptionally high infestation rate on Thris socles dussumieri. Engraulisco hina triaptella sp. nov. Figs. 18-25 Specimens of this second species of Engrau- liscobina Unnithan, 1967^ were obtained from ! the gills of Thrissocles dussumieri (Val.) ex- | amined at Vizhinjom and Trivandrum on Au- gust 19, 1954 and October 5, 1955, respectively, j Two fishes examined at Vizhinjom were infected by four individuals (two on each fish), while several of the T. dussumieri examined at Tri- vandrum were found to be parasitized by one specimen of E. triaptella along with a large number of Engrauliphila grex. Such multiple infection was not observed on the several Thris- f socles dussumieri examined at Ayiramthengu on September 8, 1955, which were infected only by Engrauliphila grex. The essential asymmetry of these worms is shown typically in one of the longer but some- what contracted specimens (Fig. 18) ; a younger one is shown extended in Figure 19, in which the bulging of the shorter side is marked. The characteristic shape is triangular, as it is in the Gastrocotyline Parasites of Indian Clupeoids — Unnithan 221 genotype Engrauliscobina thrissocles (Tripathi). There is no demarcation of the haptoral region from the body proper, since the clamp row flanks the side of the body opposite the an- terior region of the testicular zone or extends partly into the ovarian zone providing one of the shorter sides of the triangle. The length of these typical specimens is 2.4 and 2.9 mm, and their maximum width (between the gonad zones but excluding the haptoral wing) is 0.4 and 0.7 mm, respectively, giving a width-to-length ratio of 16.5% :24% — the latter being more typical. The worms are strongly flattened dorso- ventrally and highly extensible and contractile; in extension they are able to flex the body over a wide arc based on the fixed haptor, the axis of pivoting being in the thickest and widest zone; the resulting stresses would account for the convexity on the short side above the long clamp row. The shape of the whole body is like that of a scraper (particularly so in the geno- type) ; the handle of the scraper is here more abruptly demarcated, forming the neck which is about one-fifth the total length of the body axis, as it is in the genotype ; and here again the body axis is bent at a small angle between the ovarian and testicular zones, even in the contracted spec- imen (cf. my Fig. 18 and Tripathi’s [1959] Fig. 5 6a). The subterminal mouth has scattered gland cells only on the anterior lip. The oral pouches are longitudinally ovoid, 24 X 20 p, and have long muscle fibres extending posteriorly from their thick outer walls. The pharnyx is oval, Figs. 18-25. Engrauliscobina iriaptella sp. nov. 18, Complete worm, ventral view; 19, another complete worm with three clamps in the short row; 20, hind end of the haptor with the pair of incipient clamps of the worm with only two clamps in the short row; 21, first clamp nearest the lappet on the short row, dorsal view; 22, one of the middle clamps of the long row, dorsal view; 23, male genital pore with the penis head; 24, corona of spines of the male genital pore; 23, ootype and ovarian region, ventral view. 222 thick-walled, and only slightly larger than the oral pouches, 24 X 40 p, the mean diagonals of the pouches being as much as 70% that of the pharynx. The oesophagus is 0.1 6 mm long and unbranched, and bifurcates into the intestinal crura immediately behind the male genital pore. At the posterior end of the neck region, the crura have numerous complex outer branches and a few simple inner branches ; they are particularly extensive on the side bearing the clamp row. The crural ends are close together near the tip of the body, but they are not con- fluent across the median line, nor were any intercrural bridges seen. The haptor, represented by the unilateral clamp-bearing flange on the hindbody and in- cluding the lappet, is slightly less than half the total length of the slightly contracted worm (Fig. 18), but in the extended condition it is only 36%. The long clamp row (usually on the left) is 0.825-1.35 mm long, bearing 27-35 almost sessile clamps. The inhibited side of the haptor is represented in all specimens by 3 rem- nant clamps, and together they make a row only 0.10-0.12 mm long. Occasionally (as shown in Fig. 18), there may appear to be only 2 rem- nant clamps in the short row (usually the right side in my collections), but in Figure 20 it will be seen that the first two primary clamps are relatively minute and lie close together near the median line between the second pair of primary clamps — virtually the end clamp of each row. In most specimens the three clamps of the in- hibited side are subequal and in a linear series, as in Figure 19. The terminal lappet is short, narrow, and cylindrical, 0.82 X 0.25 mm-0.98 X 0.48 mm, armed with two pairs of symmetrically arranged anchors. The anterior anchors have their hooks only slightly sickle-shaped (one-quarter to one- third of a circle), with a knoblike spur at the top of the slender handle, which may be less than one-half the total length of the anchors (40 p). The posterior anchors are almost C- shaped, with a reflexed strongly rounded hook but a very short handle (total length 12-16 p). Both pairs of anchors are similar to those fig- ured by Tripathi (1959), but while the anterior pair is like those of Engrauliphila the posterior pair resembles only those of Ent granite ola micro- pharyngella. PACIFIC SCIENCE, Vol. XXI, April 1967 The clamps in the long row are 44 X 52 p- 35 X 75 p, and in the short row 20 X 28 30 X 48 p. Thus, the primary clamps of both rows are smaller and only slightly wider than long, but those in the long row are graded as usual, the largest being on either side of the middle region, and more than twice as wide as long. The braces are situated across this longer diameter; they are nearly straight with bent ends lacking the bent articular facet, though they do meet medially near the level of the divergent V-like arms of the dorsal appendix. The ventral arm of the spring is very slender, with or with- out a very slight enlargement at its end, but this extremity is always with a minute notch, never truly bifurcated (see Figs. 21 and 22). There are 9-12 irregularly oval testes, 60 X 75 p-75 X 150 p, arranged in two files and not in a single mass as depicted for the geno- type. The row on the side opposite the clamps is completely postovarian, while that nearer the clamp row extends forward to the middle of the ovary along the outer edge of the median vitel- line duct; these parovarian follicles were not found in E. thris socles. The vas deferens orig- inates from the anteriormost testes of the par- ovarian file, runs forward parallel to the uterus, and opens into a zigzag ejaculatory duct which in turn opens into the penis. No vesicula sem- inalis was seen. The muscular penis is small and conical, without forceps, and opens into the circular atrium masculinus, which is surrounded by a muscular ring 20-24 p in diameter, sit- uated at about 0.16-0.19 mm from the anterior end of the body. The penis itself bears around its widest diameter a corona of 12 sharp conical spines pointing vertically from the ventral sur- face; the spines are nearly straight. Figures 23 and 24, drawn from ventral and dorsal aspects, are intended to demonstrate the entire absence of forceps on the penis head. The ovary is in the form of an elongated in- verted U, with the distal (inner) arm wider and containing a number of large ova, and the proximal end slightly swollen and overlapped by the anterior testes of the postovarian file. The short and narrow oviduct expands slightly to form a muscular ovijector, before opening into the vitelline ampulla (Fig. 25). The wide median uterus arises from the outer margin of the fertilization chamber in the ootype, runs Gastrocotyline Parasites of Indian Clupeoids — Unnithan 223 forward parallel to the common vitelline duct, and opens into the unarmed uterine pore in front of the male genital pore. Two or three eggs with polar filaments were observed in most of the specimens, but usually were too col- lapsed for reliable measurements. The vitellaria occupy wide lateral fields ex- tending from the zone of intestinal bifurcation to the hind end of each crus; follicles are spherical, 20-25 \i in diameter, not confluent across the median line. The transverse vitelline ducts are broad and lie at the level of the an- terior quarter of the ovary ; the median vitelline duct is long and wide, and originates behind the anterior third of the ovary; it narrows as it passes backward, and it opens into the swollen vitelline ampulla in the ootype region (Fig. 25). The single median dorsal vaginal pore is cir- cular, 20 p in diameter, unarmed, but sur- rounded by a group of small spherical gland cells. It is situated a short distance behind the intestinal bifurcation (midway between the male genital pore and the transverse vitelline ducts). It is in this zone that asymmetry is particularly striking: on the side of the clamp row (usually at the left) there is, at least in nonextended worms, a marked hump on the profile before the indentation at the base of the neck (Fig. 18), and a low furrow from the vulva on the dorsal side leads obliquely to the indentation. This is the anterior limit of the lateral branch- ing of the crus and attendant vitellaria of that side. On the opposite side the profile is nearly straight from the neck zone to the lappet, and the vitellaria extend farther anteriorly along with short external crural branches to the bi- furcation on that side. This notch opposite the vulva may facilitate a finer hold during copu- lation in these worms, where the torque from the oblique attachment must be considerable. The median narrow vaginal duct runs backward dorsal to the uterus, between the oviduct and the median vitelline duct, to open directly into the fertilization chamber. It is quite independent of the vitelline duct. The genito-intestinal canal originates from the base of the ootype, runs parallel to the ovijector, and opens into the in- testinal crus. Two ill-defined excretory pores, one on each margin, are noticeable, midway between the male genital pore and the pharynx. relationships of Engrauliscobina triaptella sp. nov.: 1. In the more or less contracted state, E. triaptella is a triangle with the long clamp row as its shortest side; the inhibited haptor side of the worm is only slightly convex. In the gen- erally similar genotype, E. thris socles (Tripathi, 1959), the body is a much narrower triangle; in both there is a narrow neck, about one-fifth of the total length. The haptoral row embraces more of the body in the genotype, including the hind region of the ovarian zone, but it is more restricted in E. triaptella, being barely included in the ovarian zone. Hence, the torque in the latter species is less, and the (meta-) haptoral wing is not so extensive, in order to balance these stresses, as it is in the genotype. A further con- sequence of the torque is visible in E. triaptella in the vaginal zone, marked by a hump on the profile on the attached side of the worm and an inhibition of lateral crural branches and vitellaria anterior to the hump and neckbase, the opposite side being unaffected. In fact, the asymmetry in this species is more marked than in any other gastrocotylid and approaches that in some Opisthogynidae and Protomicrocotyl- idae. i. The clamp flange is about 36%-48% of the total length in E. triaptella (the greater the contraction the greater the pro- portion, of course), while in the genotype it is about 50%. ii. The angle made by the haptoral axis with that of the body proper is about 45°, com- pared with nearly 60° in the genotype. iii. The body torque produces a permanent slight bend in the axis of the ovarian to the testicular zones. 2. There are remnant clamps on the inhib- ited side of the haptor, only 2 in the genotype, but 3 in E. triaptella (hence its name). The un- inhibited row in mature worms bears at least 30 nearly sessile clamps. i. There are up to 35 clamps in the new species, but 40-42 in E. thrissocles. ii. The secondary clamps (but not the squar- ish primaries) are at least twice as wide as long in E. triaptella. iii. The ventral arm of the median spring is 224 PACIFIC SCIENCE, Vol. XXI, April 1967 slender and tapers to a minutely notched tip. iv. The dorsal appendix on the spring is a simple V shape. v. The braces are slender and nearly straight and lack the bent articular distal facet in E. triaptella. There is no description or figure of the secondary clamps for E. thrissocles. 3. The two pairs of persistent anchors on the lappet are of characteristic shape in both species. The genotype retains a pair of minute larval hooks at the tip of the lappet. i. The anterior anchors are slender, with feebly curved hooks (about one-third to one-half a circle), and with a knoblike spur and the handle barely half or less of the total length (40 p). ii. The posterior anchors are C -shaped or sharply reflexed simple hooks with negli- gible handles (root). 4. The penis head is absolutely devoid of forceps, the corona hooks are straight. i. There are 10 penis hooks in E. triaptella but 12 in E. thrissocles. ii. There are 9 to 12 irregularly shaped, rather large testes in two files, with some of the files (nearer the clamps) paro- varial. In E. thrissocles there are perhaps no parovarial testes. iii. There is no intercalary vesicula seminalis on the vas deferens. 5. The single median dorsal vagina leads to a median duct independent of the vitelline ducts. i. The zone of the vulva is midway between the male genital pore and the ovary. ii. The vitelline ducts are horizontally trans- verse at the first quarter of the ovarian zone. 6. The moderately ovoid pharynx is only about twice as long as the oral pouches. i. The oral pouches’ mean diameter is about 70% of that of the pharynx (about 56% in the genotype). ii. The lateral crural branches are extensive in the haptoral wing, but there are no intercrural bridges and the crura extend without dilation to near the posterior tip of the body in both species. The two species are closely related yet clearly distinct and occur on different species of Thris- socles, the genotype occurring in the northern Bay of Bengal (Puri) and E. triaptella in the South Arabian Sea and at two stations in south- ern Kerala. Pellonicola elongata gen. et sp. nov. Figs. 26-33 Several specimens of this new gastrocotyline were obtained from the gills of Pellona (llisha) hrachysoma (Blkr.) examined at Trivandrum in 1955. A single fish 17 cm long was examined on July 5, and 4 parasites were found on its outer gills; a single female fish 17 cm long, examined on August 27, had 4 parasites. From 5 fishes 18 cm in average length, examined on July 27, 14 specimens were obtained. The 4 specimens collected on July 5 proved to be the longest among the total of 22 specimens. The long, narrow, ribbon-like body (2.25- 4.5 mm long, maximum width 0.25-0.48 mm, or one-tenth its total length), tapers gently for- ward to a nearly straight neck little more than one-tenth of the total length and itself one-fifth to one-quarter as long as wide. The middle third of the body is expanded 25% to accom- modate the ovarian zone and reaches its maxi- mum at the proximal ovarian zone; behind this the testicular zone is almost straight-edged and is as wide as the preovarian zone (Fig. 26). The haptor is relatively far shorter than in any other gastrocotyline yet known, barely one-fifth the total length (0.525 X 0-15 mm-1.0 X 0-3 mm), one clamp row being twice as long as the other, with subsequal clamps ending in a short telescopic lappet. The subterminal funnel-shaped mouth is sur- rounded by a highly muscular spherical-to-oval organ having a deep circular lip with a ventral notch, and a chamber measuring 45-50 p. Be- hind it are the ovoid oral pouches, 20 X 28 p- 24 X 36 p, and these have thick walls and long muscle fibres extending down the length of the neck (Fig. 27). The pharynx is a large ovoid structure, 34 X 38 p-36 X 40 p. The mean diagonal of the oral pouches is about 61-78% Gastrocotyline Parasites of Indian Clupeoids — Unnithan 225 Figs. 26-33. Pellonicola elongata gen. et sp. nov. 26, Complete worm, dorsal view; 27, anterior region, dorsal view; 28, terminal lappet with anchors; 29, anchors; 30, clamp, dorsal view; 31, median spring of the clamp of another worm, dorsal view ; 32, male terminalia with the uterus, ventral view ; 33, middle of the body, dorsal view. 226 PACIFIC SCIENCE, Vol. XXI, April 1967 of that of the pharynx. The oesophagus is long, narrow, and unbranched, and bifurcates into the intestinal crura well behind the male pore. The crura have relatively few regular outer branches, extend backward to the level of the proximal end of the short row of clamps, and terminate independently, but close to each other, without inflations. The crus adjacent to the longer clamp row sends out wide oblique branches to each of the anterior clamps. The haptor in Pellonicola , in contrast to that in all the previous genera, is distinct from the body proper and constitutes a posterior tail. Its right and left haptoral frills are parallel, but one is only half as long as the other: in my specimens, the left frill is always the longer, 0.45-0.93 mm; the right frill range is 0.225- 0.456 mm and it bears 8-10 clamps, 36 X 20 p-44 X 36 p, while the left frill has 17-22 clamps, 36 X 24 P-48 X32 p. The terminal lappet apparently is unique in its telescopic arrangement and consists of an oblique plaque capable of being retracted as a whole into the hollow end of the haptor (Fig. 28) ; it is 60 X 28 p-68 X 36 p, and is armed with the usual two pairs of anchors (Figs. 28 and 29). The anterior pair is dis- tinctively shaped like a sickle, 40-44 p long, with the handle about equal in length to the hook, but the spur root is more than half as long as the handle and projects at right angles from it. The tip of the spur is slightly curved but does not actually pivot backward, as in Engraulzcola micropharyngella (Fig. 3), which otherwise it most resembles in this series (par- ticularly on account of its sickle hook, which is slightly more than a half circle, although the handle is more slender and relatively longer). The posterior anchors are bent in a deep C -shape and their over-all length is only 16-20 p with virtually no handle, as in Engrau- liscobina triaptella (Fig. 20, compare with Fig. 29) . The placing of the anchors is invariable; the anterior pair is always directed outward and the posterior pair inward, their hooks nearly touching, in all genera. The clamps are, exceptionally, slightly longer than wide, but their structure is very different in detail from any of those previously discribed : the median spring is highly modified, the ven- tral arm being vase-shaped in outline and ap- parently split longitudinally, with each half bowed outward ; the tip is bifurcated. The short arm of the median spring, which has radiating tendonous striae at its distal end, has an ap- pendix which in some specimens is typically U-shaped (Fig. 31), but in others is shaped like half a Maltese cross (Fig. 30). The braces (oblique sclerites) are stout and slightly wavy, and, although their inner ends touch in the middle line, there is no sharp bend here to form the familiar articular facet. The dorsal arm of the ventral jaw sclerite is unusually small, often appearing as a mere knob at the region of articulation of the dorsal and ventral jaws (compare Figs. 28 and 30). There are 34-52 testes arranged in 2-3 rather regular files. On the side of the uninhibited clamp row there is a single file of a few testes extending parovarially up to the level of the transverse vitelline duct on the left side (Fig. 33) ; the anterior testes are largest (48 X 68 p), while the posterior ones are smaller (16 X 20 The vas deferens is long and zigzag, arising from between the anterior testes and running forward; in front of the ovary it widens into a large vesicular duct to about the level of the lateral vaginae, behind the in- testinal bifurcation. This part probably func- tions as a seminal vesicle. Thence the vas deferens continues forward, to open into the base of the penis. The muscular cuplike penis is armed equatorially with a corona of 10-12 ! recurved spines, 10-12 p long (Fig. 32), but the penis head within is devoid of forceps. The median ventral atrium masculinus has a rim of radial muscle fibres and is situated at about 0.12-0.27 mm from the anterior end of the body, in front of the intestinal bifurcation. As usual, the ovary is in the form of an in- verted U and is situated in the middle third of the body, slightly shifted to the right side; the proximal region is oval and lies immediately in front of the testicular zone on the right side; it is long and narrow, and the distal limb is short (two-thirds the length of the longer Fmb) and contains large ova, each 12-14 p in diam- eter. The short and narrow oviduct descends from the distal end of the ovary and enlarges into a well-demarcated ovijector at about the middle of its length and thence continues back- ward to open into the median vitelline duct, Gastrocotyline Parasites of Indian Clupeoids — Unnithan 227 within the ootype region (Fig. 33). The wide median uterus, with cuticularized walls, ascends from the distal margin of the ootype, extends forward beyond the vaginal region, and opens into the unarmed, ventral, uterine pore, imme- diately in front of the male pore (atrium mas- culinus) (Fig. 32). In one of the specimens a collapsed egg, with a body 120 \i long, was ob- served at about the middle of the uterus. The vitellaria extend from behind the zone of the male terminalia to the anterior level of the short row of clamps, surrounding the crura and their branches but not confluent across the median line at the hind end; the vitelline fol- licles are spherical, 14-1 6 p in diameter. The transverse vitelline ducts are slightly oblique and are situated near the posterior end of the anterior third of the ovarian zone ; at their con- fluence they receive the median vaginal canal. The median vitelline duct extends backward parallel to the ovary and opens into the vitelline ampulla. The genito-intestinal canal is very narrow, irregularly wavy, and arises from the ootype. It runs obliquely forward to open into the right intestinal crus, near the proximal limb of the ovary. The vaginal pores are unarmed, just supra- marginal on each side of the body at about 0.25-0.52 mm from the anterior end, with the right pore usually slightly anterior to the left. The two vaginal ducts from the base of the lateral vaginal pores unite obliquely as a V across the median line to form a long median zigzag vaginal duct which runs backward, dorsal to the uterus, and opens into the median vitel- line duct at the junction of the transverse vitel- line ducts, hence indirectly to the ootype region. In most of the specimens traces of vitelline mat- ter were observed extending forward beyond the level of transverse vitelline ducts, into the median vaginal duct. Spindle-shaped egg s, with a filament at each pole, were seen in worms 4.5 mm long. relationships of Pellonicola elongata gen. et sp. nov.: 1. The elongated ribbon-like body with the distinct but short caudal haptor is outstanding in Gastrocotylinae. The body proper is free from the attachment zone, so the zone of pivot- ing in the extreme feeding attitudes is behind the testicular zone; but, because of the inequal- ity of the attachment basis, the stresses will be slightly greater on one side than on the other, and so the profile in contracted worms is not as symmetrical as in Microcotyle , for instance. This slight asymmetry is most obvious in the vaginal zone at the neckbase, as it is in the most asymmetrical species of Engraulis cobina. i. The body axis makes an angle of up to 60° with the haptor axis. ii. The ratio of body width to length is only 1:10, and of haptor to body length about 1:57 2. The unilateral inhibition is far less than that in any other genus of the asymmetrical gastrocotylines. In this respect it is comparable with Scomberocotyle Hargis, 1956, but in that genus a metahaptoral wing, or a secondary stimulation of the secondary clamp replication, accounts for the larger number of clamps on one side (Unnithan, 1967^). It is possible that further observation on P. elongata may show that a similar growth relation exists here and that the anterior moiety of the long clamp row does represent a metahaptor, the posterior moiety being the euhaptor with regular paired clamps in the opposite row. If this is so, the formative region for the long row would be near its middle, and that of the short row at its anterior limit, as is usual for the euhaptor. i. The long row has a total of 17-22 clamps, while the shorter has 8-10. ii. All clamps are slightly longer than wide. Perhaps this is a generic character. iii. The ventral arm of the median spring is of a unique vase shape and is split and bowed, joining distally in a very shod bifurcation. iv. The appendix on the short dorsal arm is sometimes cruciform, indicating an incip- ient cuticularization of the lateral liga- ments. v. The dorsal arms of the ventral jaw rami are remarkably reduced. vi. The braces across the middle of the cap- sule are stout but lack distally bent artic- ular facets. 3. Two pairs of persistent anchors are pres- ent on a telescopic lappet. Anterior anchors are 228 PACIFIC SCIENCE, Vol. XXI, April 1967 sickle-shaped, at least half a circle, with a nearly equal handle and a curved (not bent) spur. i. Anterior anchors 40-44 p, posterior an- chors 16-20 p bent in a deep C -shape. 4. No additional sclerites are formed within the penis corona. i. 10-12 curved penis hooks, and atrium well in front of intestinal bifurcation. ii. 34-52 testes in 3 compact files with 6-8 parovarian testes. iii. Intercalary vesicula seminalis anterior to ovarian zone. 5. There are two paired lateral vaginae with the vulvae supramarginal, as they are in Engrau- lixenus , but very far forward (one-fifth of dis- tance from male pore to anterior border of ovary). The lateral ducts run obliquely back in a Y, to form a median vaginal duct (unusual with paired vaginae) which does not run di- rectly to the ootype, but is confluent with the median vitelline duct. i. The horizontal transverse vitelline ducts are at the level of the mid third of the ovarian zone, which is rather more pos- terior than in related species. 6. The ovoid pharynx is not particularly long (as it is in Engmulixenus and Engrauliphila) , but the oral pouches are more powerful than in other genera. i. The mean diagonal of the oral pouches is 66-78% of that of the pharynx in P. elongata. ii. There are distinct simple branches of the right crus to each of the more anterior (unpaired) clamps of the long row. iii. The crura are nearly equal and extend to the body tip without dilatations. 7. There are no intercrural bridges, but there is an exceptional circumoral, sucker-like, circular lip in Pellonicola that is not present in allied genera. The above characters show that Pellonicola is a somewhat aberrant gastrocotyline, and that it is the first ever found on a member of the Clupeidae. ACKNOWLEDGMENTS It is a pleasure to record here my indebted- ness to Miss Nora G. Sproston for guidance and criticism. My sincere thanks are also due to Dr. N. K. Panikkar who provided me with facilities at the Central Marine Fisheries Re- search Institute, Mandapam Camp, and Dr. C. C. John for permission to work in the Marine Biological Laboratory, Trivandrum. REFERENCES Bychowsky, B. E. 1957. Monogenetic Trem- atodes, their Systematics and Phylogeny. [In Russian] Akad. Nauk USSR, pp. 1-509. George, K. C. 1961. On a new gastrocotylid trematode, Engraulicola forcepopenis gen. et sp. nov. on white-bait, from southern India. J. Mar. Biol. Assoc. India 2(2) :208-215. Hargis, W. J., Jr. 1956. Monogenetic trema- todes of Gulf of Mexico fishes. Part XII. The family Gastrocotylidae Price, 1953. Bull. Mar. Sci. Gulf and Caribbean 6:28-43. Llewellyn, J. 1957. The larvae of some monogenetic trematode parasites of Plymouth fishes. J. Mar. Biol. Assoc. U. K. 36:243- 259. 1959. The larval development of two species of gastrocotylid trematode parasites from the gills of Trac burns trachurus. J. Mar. Biol. Assoc. U. K. 38:461-467. Price, E. W. 1959. A proposed reclassifica- tion of the gastrocotyloid Monogenea. J. Parasitol. 45, 4(2) :22-23. Ramalingham, K. 1953. A new genus of trematode ( Chauhanea ) from the gills of Sphyraena acutipinnis Day. J. Zool. Soc. India 5(1) : 59-63. Sproston, N. G. 1945. A note on the compara- tive anatomy of the clamps in the super- family Diclidophoroidea (Trematoda: Mono- genea). Parasitology 36:191-194. 1946. A synopsis of the monogenet'c trematodes. Trans. Zool. Soc. London 25(4): 185-600. 1965. Host ecology correlated with speciation in some monoxenous parasites. (In press.) Tripathi, Y. R. 1957 (issued 1959). Mono- Gastrocotyline Parasites of Indian Clupeoids — Unnithan 229 genetic trematodes from fishes of India. Ind. J. Helminthol. 9(1-2) : 1-149. Unnithan, R. V. 1957. On the functional morphology of a new fauna of Monogenea on fishes from Trivandrum and environs. Part I. Axinidae fam. nov. Bull. Cent. Res. Inst. Univ. Kerala, Trivandrum 5 (2) Ser. C, Nat. Sci.:27— 122. — 1961. Ibid. Part III. Heteromicro- cotylidae (Microcotyloidea) . Ind. J. Hel- minthol. 13(2) : 1 1 2 — 145 . 1962. Ibid. Part II. The family Opis- thogynidae fam. nov. (Gastrocotyloidea) and Abortipedinae sub fam. nov. (Protomicro- cotyloidea). Parasitology 52:315-351. 1967<2. Ibid. Part IV. Microcotylidae s.s. and their repartition in subsidiary taxa. (In press.) I96lb. Patterns of secondary growth and a revision of the systematics in Micro- cotyloidea and Gastrocotylidae (Monoge- noidea). (In press.) The Growth of Arachnoides placenta (L.) (Echinoidea) Judith Hines and Ron Kenny1 ABSTRACT: Arachnoides placenta (L.) increases in diameter 7 mm during the first year, 4 mm during the second year, and 4 mm in the third year after meta- morphosis. The relationship between diameter and weight of the test is approximately of cubic form. The spawning period in north Queensland probably is June to July, and the time of metamorphosis probably September. The growth characteristics are compared with other echinoid species. The "sand dollar,” Arachnoides placenta (L.), has been recorded from a wide geo- graphical area ranging from the Andaman Islands through the Philippines and along the northern and northeastern coasts of Australia (Clark, 1946). The distribution of the species on the Queensland coast extends from Thurs- day Island (Clark, 1921) to Mackay (Endean, 1953, 1956). Clark (1938) recorded some observations on the habits of this "sand dollar” from Darwin, but no data on the growth of the species have been traced. METHODS AND RESULTS The animals used in this study were taken at Lucinda (18°31'S, 1 46° 19'E) in the mid- dle of the range of A. placenta along the Queensland coast. Lucinda beach is a gently sloping sandy strand with many offshore shoals. The spring tidal range is 8 ft. At low tide Arachnoides is distributed from approximately mean sea level to below low water of spring tides, lying on, or a little under, the sand surface. Collections were made at daytime spring low tides on 10 occasions from March 1961 to February 1962. In collecting, all specimens within a narrow strip extending from the last high water mark to below low tide were taken. In general, animals were collected by hand, but random digging and sieving along the length 1 Zoology Department, University College of Townsville, Townsville, Queensland, Australia. Manu- script received May 8, 1966. of the traverse ensured sampling of the popula- tion down to a diameter of approximately 6 mm. Measurements of diameter were read to the nearest millimeter and weights of oven-dried whole animals to the nearest milligram. The temperature data were recorded at Townsville Harbour breakwater (19° 15'S, 146° 50'E) and varied from a June mean of 20.5°C to a January mean of 31.5°C (Fig. 3). The records of diameters were arranged in 1-ram class intervals (Table 1), and the data were smoothed prior to plotting as a series of monthly histograms (Fig. 2). From these histograms the modes of frequency distribution of diameter were extracted and drawn as a growth curve (Fig. 3). These results show Arachnoides increasing in diameter from 11 to 18 mm during the Novem- ber to February period of the first year after metamorphosis. From March to October there is little increase in size, but a second season of active growth commences in October and in the ensuing three to four months the diameter enlarges from 18 to 22 mm. In the second year there is a further seven- to eight-month period with no obvious growth, followed by another growing season when the animal increases from 22 to 26 mm in diameter during approximately four months. The rate of growth during the active growing period decreases with age, being 2.3 mm per month in the first growing season, 2 mm per month in the second, and 1.3 rnm per month in the third. 230 Growth of Arachnoides placenta — Hines and Kenny 231 Fig. 1. Selected examples of A. placenta. The period of growth occurs during the part of the year when sea water temperatures are rising (Fig. 3). The relationship between diameter and weight of test for 125 selected specimens is shown in Figure 4. The equation for the rela- tionship is of the form log W = 3.061 (log d) — 1.519 where W is the weight of the dried test and d the diameter of the animal. DISCUSSION Extrapolation from the known data would suggest that metamorphosis occurs during Sep- tember. If it is assumed that the length of larval life is similar to that for another tropical echinoid, Tripneustes esculentus, which is two months (Lewis, 1958), it may be deduced that Arachnoides placenta spawns in June or July. Lewis (1958) lists July and August as the spawning period for T. esculentus ; and Hyman (1955), quoting Mortenson, notes that DAellita sexiesperforata spawns in March and April in the West Indies. It would appear that the West Indian species spawn during the early summer for their locality, while A. placenta spawns during the Queensland winter at sea temperatures between 20 °C and 22 °C. The occurrence of some Arachnoides in- dividuals of diameters up to 36 mm suggests a life span of up to five years, assuming the rate of growth for older animals to be similar to that calculated for specimens up to the third year. Crozier (1920) suggested four to five years as the normal life span for the similar but larger Atlantic species Mellita sexiesper- forata at Bermuda, and Moore (1934) esti- mated four to eight years as the duration of life for Echinus esculentus in Britain. Lewis (1958), working with Tripneustes at Barbados, was not able to determine a life span because of commercial fishing for the species. The recording of a specimen of A. placenta of 96 mm diameter from Lindeman Island (Clark, 1946) suggests the possibility of a situation similar to that described by Moore 232 PACIFIC SCIENCE, Vol XXI, April 1967 TABLE 1 Frequency Distribution of Diameter in Millimeters of A. placenta (Original Data) DIAMETER (in mm) NUMBER OF SPECIMENS MAR. 25 MAY 21 JUNE 15 JULY 17 AUG. 15 OCT. 5 CCT. 22 NOV. 21 DEC. 12 JAN. 16 FEB. 12 7 5 8 1 1 1 7 9 4 1 1 10 1 1 2 2 3 3 11 1 1 11 7 5 1 1 7 12 1 1 4 11 8 8 1 2 13 6 3 6 25 13 15 1 6 14 3 9 2 33 15 25 4 15 10 10 8 51 30 35 3 7 6 16 2 13 5 60 35 31 3 1 3 13 10 17 17 17 10 65 60 36 6 10 2 12 9 18 26 22 30 63 55 40 8 20 9 29 22 19 34 22 25 55 45 27 21 24 20 33 25 20 33 22 30 43 68 27 23 53 37 24 11 21 35 22 31 45 38 16 33 50 50 31 19 22 41 25 24 47 36 12 35 45 49 28 24 23 31 26 29 33 24 11 31 43 45 25 17 24 33 23 18 34 24 13 22 24 45 18 10 25 26 25 14 47 15 9 21 25 28 14 24 26 21 17 11 34 13 4 11 20 11 26 23 27 8 15 10 20 9 2 11 11 14 23 9 28 17 15 16 14 6 5 10 13 12 8 5 29 7 10 16 13 10 2 7 7 12 5 4 30 4 6 5 5 2 2 2 2 4 2 3 31 3 4 1 2 1 4 2 1 1 4 3 32 6 5 2 2 4 3 2 3 4 33 2 1 5 4 2 1 2 3 34 1 1 1 1 2 35 1 1 1 36 2 1 Growth of Arachnoides placenta — Hines and Kenny 233 ff I I « I I I /5““* Fig. 3. Growth curve for A. placenta related to sea water temperatures at Townsville. (1934), where size and test thickness distin- guished an "inshore race’’ of Echinus escu- lentus. Direct comparison of the growth rate of A. placenta with the results of growth studies on other echinoids is difficult because of marked species differences in size and shape. In Table 2 the increase in size per annum is expressed as a percentage of the diameter for the preceding year for selected species. The data for Mellita are possibly those most relevant, inasmuch as this species is similar in shape to Arachnoides in contrast to the spherical form of the other species listed. TABLE 2 Comparison of Growth Rates of Selected Echinoids PERCENTAGE INCREASE IN DTAMETER YEAR Echinus ( Moore, 1935) Psam- mechinus (Bull, 1938) Strongy- locentrotus (Norway) (Greig, 1928* ) 2 36 30 150 3 12 12 60 4 10 3 33 5 14 23 25 6 5 20 Strongy- locentroius (Maine) Mellita (Swan, ( Crozier, 1958) 1920) Arachnoides 176 30 22 68 14 19 25 43 Quoted by Hyman (1955). wr in 6 MS 234 PACIFIC SCIENCE, VoL XXI, April 1967 Fig. 4. Diameter and weight relationship for A. placenta. Growth of Arachnoides placenta — Hines and Kenny 235 It would appear that the percentage increase in diameter during the second year is less for Arachnoides than for other echinoids. Those species inhabiting cooler waters apparently grow more rapidly than do the tropical A. placenta, but by the end of the third year Mellita, a warm water echinoid, and Arach- noides present a similar picture. The diameter-to-weight relationship for A. placenta approximates the expected cubic form and is similar to that expressed by Swan (1958) for Strongylocentrotus droehachiensis. However, the wide range of weight values for any one diameter makes it difficult to reduce the relationship to a series of separate equations as Swan (1958) has done for the Maine species. The authors wish to thank the members of the technical staff of the University College of Townsville who assisted in the collection of specimens for this study. REFERENCES Bull, H. O. 1938. The growth of Psam- mechinus miliaris (Gmelin) under aquarium conditions. Rept. Dove Mar. Lab., ser. 3, 6:39-42. Clark, H. L. 1921. The echinoderm fauna of Torres Strait ; its composition and origin. Carnegie Inst. Publ. 214, Dept. Mar. Biol., vol. 10:1-224. 1938. Echinoderms from Australia. Mem. Mus. Comp. Zool. Harvard 55:1-597. 1946. The echinoderm fauna of Aus- tralia; its composition and origin. Carnegie Inst. Publ. 566:1-567. Washington, D.C. Crozier, W. J. 1920. Notes on the bionomics of Mellita. Am. Naturalist 54:435-442. Endean, R. 1953. Queensland faunistic records. 3. Echinodermata (excluding Cri- noidea). Pap. Dept. Zool. Univ. Qld. 1:51— 60. 1956. Queensland faunistic records. 4. Further records of Echinodermata (excluding Crinoidea). Pap. Dept. Zool. Univ. Qld. 1:121-140. Hyman, L. H. 1955. The Invertebrates, vol. 4. Echinodermata. McGraw-Hill, New York. Lewis, J. B. 1958. The biology of the tropical sea urchin Tripneustes esculentus Leske, in Barbados, British West Indies. Canadian J. Zool. 36:607-621. Moore, H. B. 1934. A comparison of the biology of Echinus esculentus in different habitats, part I. J. Mar. Biol. Assoc. U. K. 19:169-185. Moore, H. B. 1935. Ibid., part II. J. Mar. Biol. Assoc. U. K. 20:109-128. Queensland Government. 1966. Tide tables for the coast of Queensland. Dept, of Harbours and Marine, Brisbane. Swan, E. F. 1958. Growth and variation in sea urchins of York, Maine. J. Mar. Res. 17:505-522. A New Siphonophora, Vogtia kuruae n. sp.1 Angeles Alvarino The genus Vogtia Kolliker is represented by four species: Vogtia pentacantha Kolliker, 1853; V. spinosa Keferstein and Ehlers, 1861; V. serrata (Moser), 1925; and V. glabra Bige- low, 1918. 1 Contribution from Scripps Institution of Oceanog- raphy, University of California, San Diego. These studies have been conducted under the Marine Life Research Program, the Scripps Institution’s com- ponent of the California Cooperative Oceanic Fish- eries Investigations; and supported by the National Science Foundation (NSF G-19417, GB-2861). Manuscript received December 14, 1965. The most useful diagnostic feature of the nectophores of the first three species has been described as being angular, prismatic, and pen- tagonal (Table 1). The last species, V. glabra , has rounded nectophores which are rather sim- ilar to those of Hippopodius hip p opus. Bigelow and Sears (1937) described the first three spe- cies above as "the three angular belled species." Actually that characteristic is most conspicuous in Vogtia kuruae n. sp. Holotype: usnm Cata- logue Number 52609; Paratype: usnm Cata- logue Number 52610.) [G. 1. Young nectophore of Vogtia kuruae n. sp. 236 Vogtia kuruae n. sp. — Alvarino 237 TABLE 1 General Characteristics of the Nectophores of Four Species of Vogtia * V. pentacantha KOLLIKER V. spinosa KEFERSTEIN AND EHLERS V. serrata (moser) V. kuruae n. sp. Pentagonal, prismatic, Pentagonal, prismatic ; Angular, prismatic; Prismatic, star-shaped ; ridges with promi- facets and ridges ridges serrated, facets both ridges and nences; facets with conical gelat- smooth. facets smooth, with- smooth. inous prominences. out serrations or Ventral channel joins Ventral channel joins conical prominences. the dorsal one at to the dorsal Ventral channel joins Ventral channel joins about the first 1/4 channel at about the the dorsal one at the dorsal one at of the nectosac. first 1/8 of the about the first 1/3 about the middle of nectosac. of the nectosac. the nectosac. * Vogtia glabra Bigelow is not included in this comparison because it has rounded nectophores. The nectophores of V. kuruae n. sp. present an outline like a three-pointed star. Three isosceles triangles are arranged surrounding the nectosac in such a way that the imaginary bases or smallest side of the triangles circumscribe the nectosac (Figs. 1 and 2). These nectophores thus display the most exaggerated angular shape of all the previously described species, where the three-pointed shape is already incipient. In this species both edges and facets are completely smooth, without protuberances, spines, or serra- tions. On the dorsal part of the nectophores appears the nectosac, a shallow cavity outlined as a quite Fig. 2. Adult nectophore of Vogtia kuruae n. sp. 238 PACIFIC SCIENCE, Vol. XXI, April 1967 TABLE 2 Distribution of Vogtia kuruae N. sp. EXPEDITION OR CRUISE POSITION DEPTH IN METERS STATION Pacific Ocean Downwind 46° 35'S, 113°12'W 2010-0 20a 1957 23°39'S, 118°12'W 514-0 37 Monsoon* 1960-1961 49°26'S, 132°18'E 49°21'S, 132039"E 1878-0 13 Shellback 9°52.5'S, 81°32'W 313-0 125 1952 12°59'S, 85°01'W 311-0 144 8°07/S, 84°58'W 176-0 149 4°05'S, 85°00'W 298-0 153 Tethys 1960 21°33'N, 123°02'W 21°21'N, 123°12'W 1500-0 4 18°44'N, 124°24'N 18°16'N, 124°24'W 2586-0 5 7°47'N, 129°37'W 7°26'N, 129°34.5'W 3114-0 9 10o09'N, 147°08'W 10°35'N, 147°29.6'W 868-0 19 26°13.9'N, 141°34.5'W 26°22.1'N, 141°06.9'W 3000-0 28 29°01.2'N, 132°09'W 29°11.6'N, 131°41.5'W 3000-0 31 30°47.6'N, 125°25'W 30°59'N, 124°53.8'W 868-0 33 Transpac* * 47°35.7'N, 167°44.8'E 510-340 49C 1953 same same 680-510 1015-0 49D 44°06'N, 161°39'E 653-490 59D 44o09'N, 152°56.8'E 675-435 66D Troll*** 17°59'N, 134°24'E 200-0 21 1955 15°56'N, 132°27'E 200-0 22 15°17'N, 124°17'E 200-0 33 20°43'N, 123°29'E 200-0 35A 29°54'N, 132°45'E 200-0 4lA 28°28'N, 135°52'E 200-0 43A CalcoFi 32°50'N, 120°42'W 140-0 87.65 Cruise 5804 31°27'N, 121°57.5'W 420-0 90.90 29o40'N, 120°52'W 618-0 100.90 Naga 6°23'N, 102°11'E 176-0 60-324 1959-1961 9°54'N, 110°34'E 630-0 60-525 Indian Ocean Monsoon 1960-1961 18°49'S, 88°05'E 18°4l'S, 87°51'E 1643-0 6 33°19'S, 72°34'E 33°38'S, 72°31'E 1878-0 9 36°35'S, 95°28'E 36°32'S, 95°52'E 2000-0 11 Vogtia kuruae n. sp. — Alvarino 239 TABLE 2 ( Continued ) EXPEDITION OR DEPTH IN CRUISE POSITION METERS STATION Atlantic Ocean Lusiad 00°56'N, 11°29'W 2300-0 79 1963 01°25'N, 11°43'W 18°58'S, 10°15'W 18°30'S, 2000-0 55 19°13'S, 13°44'W 18° 58S', 13°37'W 2000-0 52 30°09'S, 04°42'W 30°07'S, 05°15'W 3500-0 24 32°30'S, 09°04'E 32°24,S, 08°25'E 3400-0 14 33°47'S, 15°48'E 33°46'S, 15°29'E 2000-0 11 * This species did not appear in the one-meter net oblique tows taken from various depths (356-200 m) to the surface. The records included correspond to mid-water trawls. ** In the small number of stratified samples obtained during this expedition, the species occurred in only a few, and always at depths below 300 m. *** It is interesting to note that the species occurred in the upper 300 m in the tropical regions, or in zones of upwelling in subtropical waters (CalcoFi records). This emergence of the populations in the tropical regions is not apparently related to either temperature or salinity; but it might be associated with the oxygen concentration, or indirectly with the inorganic- organic phosphate-phosphorus (see Reid, 1965: Figs. 2-5). perfect circle. In the nectosac the four radial channels follow nearly direct courses, as in Hip- popodius hippopus. There is a crescent ventral sinus, which appears mostly in an M shape, a distinctive characteristic of the species, but in most of the nectophores it is not clearly seen. Sometimes the middle pyramid of the necto- phores is more enlarged (Fig. 1) than the others (as in Bigelow, 1931: Fig. 190). This might be related to the age of the nectophore. The five loose nectophores collected at Cocos (4°56'N, 84°35'W), provisionally referred to V. s errata Moser by Bigelow (1931), probably belong to the present species, "because they entirely lack the large conical gelatinous spines so characteristic of V. spinosa and of V. penta- canthd' and because of "their peculiarly elon- gated outline with pyramidal apex, much more prominent than in any Vogtia previously de- scribed." Likewise, the nectophores determined as be- longing to V. pentacantha (Bigelow, 1913) later corrected to V. serrata (Bigelow and Sears, 1937) might be V. kuruae , especially those shown in Bigelow’s Plate 5, Figure 9- Bigelow (1913) stated, "In pentacantha the surfaces of the facets are smooth at all ages,” and later he added, "But in the present species the older nectophores have no spines at all. The ridges, like the facets are perfectly smooth, though in the very youngest nectophores the margins of the facets are always ? more or less irregular, and I found one in which they are distinctly spinous." It could be that Bigelow’s (1913) material included both V. pentacantha or V . ser- rata and the present species, because his Figure 9 in Plate 5 is rather different from the others, and similar to V. kuruae. In V. kuruae n. sp. I found that both young and old nectophores present smooth ridges and facets, a characteristic which does not correspond to any of the existing described species. The nectophores of the four species previ- ously described differ in details of form, as is clearly shown when comparing the present fig- ures of V. kuruae with the published descrip- tions of the other species. See Bigelow, 1911: 210, pi. 15, figs. 5-13; 1913:66, pi. 5, figs. 7-8; 1918:405, 406, 407, pi. 4, figs. 1-7; 1931:537, 538; Browne, 1926:61; Chun, 1897:35, pi. 1, figs. 11-14; Haeckel, 1888:177, 182, 364, pi. 29, figs. 9-14; Keferstein and Ehlers, 1861:24, pi. 5, figs. 16-17; Kolliker, 1853:31, pk 8, figs. 1-8; Leloup, 1933:17, 18, 19; 1934:6; Moser, 1925:420, pi. 27, figs. 6-8, pi. 28, figs. 8-9; Totton, 1932:331. 240 PACIFIC SCIENCE, Vol. XXI, April 1967 DISTRIBUTION OF V . kuYUde N. SP. Loose or interlocked nectophores of this spe- cies have been found in plankton samples col- lected and analyzed during a number of expedi- tions as shown in Table 2. The fact that V . kuruae is more abundant in deep tows suggests that it is characteristically a deep water species. Data on the bathymetric distribution of this species were taken off California (30°30'N, 120°00'W). The stratified samples were col- lected with the bongo or bmoc open-closing net (McGowan and Brown, 1966) at various depths during August 27 and 30, and Septem- ber 1, 2, 3, and 5, 1965. During these series of collections, V. kuruae did not occur in samples collected in the upper 300 m, nor in the samples from below 1030 m. It did occur in samples taken at 460-410, 500-420, 620-530, 775-685, and 1030-860 m. REFERENCES Bigelow, H. B. 1911. Reports on the scientific results of the expedition to the eastern tropical Pacific, 1904-1905. XXIII. The Siphono- phorae. Mem. Mus. Comp. Zool. Harvard 38(2) :173-401. 1913. Medusae and Siphonophorae col- lected by the U. S. Fisheries steamer "Alba- tross” in the N.W. Pacific, 1906. Proc. U. S. Natl. Mus. 44 (1946) : 1-1 19. 1918. Some Medusae and Siphono- phorae from the western Atlantic. Bull. Mus. Comp. Zool. Harvard 62(8) : 365-442. 1931. Siphonophorae from the Arcturus Oceanographic Expedition. Zoologica, Sci. Contrib. N. Y. Zool. Soc. 8(11) : 525-592. and Mary Sears. 1937. Siphonophorae. Rept. Danish Oceanogr. Expedition, 1908- 1910, to the Mediterranean and adjacent seas, 2. Biology, H 2:1-144. Browne, E. T. 1926. Siphonophorae from the Indian Ocean. Collected on H.M.S. "Sealark” Expedition. Trans. Linn. Soc., ser. 2, Zool. 19:55-86. Chun, C. 1897. Die Siphonophoren der Plank- ton Expedition. Ergeb. der Plankton Exped. der Humbolt-Stiftung 2K.b:l-126. Haeckel, E. 1888. The Siphonophorae. Rept. Sci. Res. Voy. H.M.S. "Challenger,” Zool- ogy 28:1-380. Keferstein, W., and E. Ehlers. 1861. Zool- ogische Beitrage gesammelt in Winter 1859- 60 in Neapel und Messina. Beobachtungen liber die Siphonophoren von Neapel und Messina. Liepzig. 34 pp. Kolliicer, A. 1853. Die Schwimmpolypen oder Siphonophoren von Messina. Leipzig. 96 pp. Leloup, E. 1933. Siphonophores calycophor- ides provenant des campagnes du Prince Albert I de Monaco. Res. Camp. Sci. Monaco 87:1-64. 1934. Siphonophores calycophorides de l’Ocean Atlantique tropical et austral. Bull. Mus. Hist. Nat. Belg. 10(6):l-87. McGowan, J. A., and D. M. Brown. 1966. A new opening-closing paired zooplankton net. Univ. Calif. Scripps Inst. Ocean. SIO Ref. 66-23:1-55. Moser, F. 1925. Die Siphonophoren der Deut- schen Siidpolar Exped., 1901-1903. Deut- chen Siidpolar Exped. 17, Zool. 9:1-541. Reid, J. L. 1965. Intermediate waters of the Pacific Ocean. Johns Hopkins Oceanographic Studies 2:1-80. Totton, A. K. 1932. Siphonophora. Great Barrier Reef Exped., 1928-1929. Sci. Repts., Brit. Mus. (Nat. Hist.) 4(10) :317-374. The Systematics of the Prickly Sculpin, Cottus asper Richardson, a Polytypic Species Part I. Synonymy, Nomenclatural History, and Distribution1 Richard J. Krejsa2 ABSTRACT: The prickly sculpin, Cottus asper, is a geographically widespread, polytypic species characteristically represented by very prickly, nonmigratory, fresh- water spawning "inland” forms, and less prickly, catadromous, brackish-water spawning "coastal” forms. Part I, the first contribution in a series on the systematics of this species, presents a synonymy complete for the period 1836-1936, with a resume of the most important citations from 1936 to 1965. A nomenclatural history of the species is given. The distributional range is listed and also presented in illustration. The prickly sculpin ranges over about 3,000 miles of Pacific North Temperate coastline and inland as far as 300 miles. The species exists in two primary modes of morphological vari- ability: one, a nonmigratory, fresh-water spawner, has extensive squamation on certain regions of the body; the other, a catadromous, brackish-water spawner, has little or none. Prickly sculpin eggs are spawned naturally in environments which are known to vary in at least one major factor, i.e., salinity. The mor- phological, behavioral, and ecological varia- tions existing within this species make it an excellent subject for systematic analysis. SYNONYMY It has been 130 years since Cottus asper was first described by Sir John Richardson. Prior to the present study, regional systematic treat- ments of this widespread species resulted in a proliferation of generic and specific taxa, all referable to C. asper. Early revisionary work by Girard in 1851 and 1852 was incomplete because of lack of specimens. Recent regional 1 The data for this paper are taken from a thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Uni- versity of British Columbia. Manuscript received March 23, 1966. 2 Institute of Fisheries, University of British Columbia, Vancouver, Canada. Present address: De- partment of Anatomy, College of Physicians and Sur- geons, Columbia University, New York. works (Robins and Miller, 1957; McAllister, 1957; McAllister and Lindsey, 1959; and Bond, 1963) have included the species as part of a geographical area or river drainage system, but there has never been a comprehensive treatment of the species throughout its entire range. In the period 1836-1936, 41 reports of C. asper were cited in the literature. In the same period, 32 additional citations occurred which were either misidentifications or synonyms properly referable to C. asper. In none of the systematic treatments published since the original descrip- tion in 1836 has there been a synonymy con- taining more than 7 citations. McAllister (1957) listed 15 citations in his unpublished M.A. thesis. The present synonymy consists of 73 citations published during the period 1836-1936, and is thought to be complete for that period. In the past 30 years, the species has been cited in- cidentally in so many fishery journals and publications that only the major systematic, or otherwise noteworthy, citations have been re- ported in the remaining synonymy. Cottus asper Richardson, 1836 Cottus asper. Richardson, 1836:295, pi. 95, fig. 1 (original description and figure; Colum- bia R. ; collected by Dr. Gairdner, probably near Fort Vancouver, Washington Territory). Storer, 1846^:260, and 1846^:8 (northwestern coast of N. America). Girard, 1850:409, and 1851^:189 (discusses propriety of present 241 242 PACIFIC SCIENCE, VoE XXI, April 1967 nomenclature). Eigenmann, 1895:118 (abun- dant in Fraser system from tidewater to 1,900 ft; Mission, Sicamous, Kamloops, and Griffin L., British Columbia; and Umatilla, Oregon). Gilbert and Evermann, 1895:201 (description; comparison with Sacramento R. form; Walla Walla R. at Wallula, and Lake Washington, Washington). Seale, 1896:854 (Lake Washington). Gilbert, 1896:418 (de- scription; stream entering Departure Bay, Van- couver Island). Jordan and Evermann, 1896: 439 (synonymy; streams of the Cascade Range, from Vancouver Island to Oregon). Gilbert, 1898:1 (Columbia R. ; notes absence in Kla- math R.). Jordan and Evermann, 1898:1944 (description; synonymy; Walla Walla; De- parture Bay; about Port Townsend; streams of the Cascade Range, from Vancouver Island to Oregon). Evermann and Meek, 1898:83 (Lake Washington). Meek, 1899:231 (Lake Souther- land, Olympic Peninsula, Washington). Jordan, 1905:445 (streams of the Pacific coast). Snyder, 1905:337 (description; affinities; habitat preference; San Franciscito, Madera, San Antonio, Guadalupe, Coyote, and Alameda creeks, all flowing into San Francisco Bay). Evermann and Goldsborough, 1907^:306 (characters; prickling descriptions; Deep Bay, Naha R., and Steelhead Cr., Loring, Alaska; Hunter Bay, Yes Bay, and McDonald L., Alaska). Evermann and Goldsborough, 1901b'. 110 (Fraser R. at Mission, Shuswap L. at Sica- mous, Thompson R. at Kamloops, and Griffin L.). Rutter, 1908:145 ( Cottopsis parvus first placed in synonymy with asper; synonymy, in part, except Uranidea semis caber (sic) centro- pleura Eigenmann and Eigenmann; summary of 9 localities in Sacramento R. system). Snyder, 1908^:269 (Russian R., California). Snyder, 1908^:184 (characters; prickling description; summary of 41 localities: from Lake Washing- ton, Columbia and Sacramento R., and river basins in between). Nichols, 1909:172 (head of Chilkoot L., Alaska). Evermann and Lati- mer, 1910:138 (4 localities in Marin Co., and 2 localities in San Francisco Bay, California; 12 localities in Olympic Peninsula, Washington). Snyder, 1913:72 (characters; Pajaro R., Cali- fornia). Snyder, 1916:381 (Papermill and Walker creeks, California). Kermode, 1917:20 (Hanceville, British Columbia [Chilcotin R. drainage]). Jordan, 1919:249 (designates Cottus asper Richardson as the orthotype of Cottopsis Girard). Bean and Weed, 1920:7 6 (mouth of Fraser R.) . Hubbs, 1921:7 (re- identification of San Luis Cr., California, speci- mens misidentified by Jordan as C. gulosus [1895:141]; range extension to Ventura R., California). Fowler, 1923:282 (Hanceville, and Shawnigan L., British Columbia; Shawni- gan L. specimen misidentified by Kermode [1909:87] as Uranidea gulosa) . Crawford, 1927:177 (streams flowing into Puget Sound). Schultz, 1929:48 (listing only). Schultz, 1930: 14 (most streams and lakes of western Wash- ington). Jordan, Evermann, and Clark, 1930: 383 (synonymy; streams of the Cascade Range, southeastern Alaska to Oregon; south to Sacra- mento R.). Kermode, 1931:19 (Cowichan L., Vancouver Island). Evermann and Clark, 1931: 56 (summary of 32 recorded localities in Cali- fornia). Taft, 1934:251 (spawning migra- tion). Schultz and DeLacy, 1936,^:128 (syn- onymy; coastal streams from Alaska to Ventura Co., California; fresh water and brackish water; review of most records from Puget Sound to Oregon, and addition of 26 more localities). Schultz and DeLacy, 1936^:213 (additional synonymy; 3 new localities). Schultz, 1936:179 (keys to species of Cottus ; coastal streams from Alaska to Ventura Co., California; fresh-water and brackish). Dymond, 1936:71 (description; throughout southwestern British Columbia, in- cluding southern Vancouver Island; 16 locali- ties listed). Sumner, 1942:1-25 (common in tidewater areas along the Oregon coast) . Hubbs and Wallis, 1948:141 (identification of "Cot- tus sp.” recorded by Dill, 1946:54). Bailey and Dimick, 1949:14 (comparison with Cottus hubbsi') . Shapovalov and Dill, 1950:387 (list- ing only). Wilimovsky, 1954:285 (southeast Alaska to California). Robeck, et al., 1954: I B-65 (Columbia R., above Trinidad, Washing- j ton; cited as "prickly sculpins Cottus sp.," these may include Cottus rhotheus in part). Lindsey, 1956:777 (Pacific Slope of N. America from Alaska to California; British Columbia main- land from Columbia, Fraser, and Skeen a sys- j terns, Stikine R. headwaters, and Peace R. [Summit L., Heart L., Angusmac Cr., and McLeod L.] ) . Robins and Miller, 1957:229 ( Cottopsis parvus again removed from syn- Systematics of Prickly Sculpin, I — Krejsa 243 onymy of C. gulosus') . Lindsey, 1957:657 (British Columbia: Columbia R. ; Fraser R.; Skeena R.; coast drainages south of Skeena; Nass R. ; Stikine R. ; Peace R.). Wilimovsky, 1958:62 (key to species in Alaska; southeast Alaska to California). Shapovalov, Dill, and Cordone, 1959:173 (listing only). Carl, Clem- ens, and Lindsey, 1959:158 (description; Pacific drainages from Chilkoot L., Alaska, to Ventura R., California. In British Columbia: lakes and rivers of the Columbia, Fraser, Dean, Skeena, Nass, and Stikine systems; coastal rivers of the mainland and Vancouver Island, and Queen Charlotte Islands; headwaters of Peace R. system from Summit L. to McLeod L.). McAllister and Lindsey, 1959:70 (de- scription; synonymy; intraspecific variation; localities as in Carl, Clemens, and Lindsey, 1959). McAllister, 1960:42 (collection in salt water, Pt. Atkinson, British Columbia). Bond, 1961:36 (key to species in Oregon; prickling variation; Columbia R. drainage). Bond, 1963: 79 (synonymy; life history observations; oxy- gen, temperature, and salinity tolerance of adults; fish associates; habitat preference; lists 35 new collection localities in coastal Oregon, 30 localities from Columbia R. drainage in Oregon, and also some lakes in the southwest corner of Rainier National Park, Washington). Bailey and Bond, 1963:19 (recognition of several species groups within Coitus in western N. America; characters and list of species in the " asper species group”). Krejsa, 1965:1-109 (synonymy; nomenclatural history; distribution; life history; morphological variation; salinity tolerance of eggs; phylogeny of C. asper and closely related species). Bohn and Hoar, 1965: 977 (salinity effects on iodine metabolism; physiological divergence of inland and coastal C. asper). Centridermichthys asper. Richardson, 1844:76 (River Oregon [= Columbia R.]). Gunther, 1860:170 (description; synonymy; fresh waters of the Oregon and Washington Territories). Lord 1866^:130 (life history; spawning behavior, in part; Puget Sound; | "streams flowing through the Sumass and Chilukweyuk prairies” [Sumas and Chilliwack R. ?], British Columbia; in part, all streams east and west of the Cascades). Cottopsis asper. Girard, 1851^:303 (intro- duction of Cottopsis gen. nov. ; synonymy; limited to River Oregon [= Columbia R.]). Girard, 1851c:185 (not seen). Girard, 1852:61 (definition of Cottopsis gen. nov., based on Richardson’s description of Cottus asper ; syn- onymy; Columbia R.). Girard, 1859:51 (de- scription; synonymy; based on 8 specimens from Astoria and Fort Dalles, Oregon, and Fort Steilacoom, Puget Sound, Washington Territory). Suckley (in Cooper and Suckley, 1859) 1859:351, and Suckley, 1860:351 (de- scription; synonymy; small fresh-water streams emptying into Puget Sound; Ft. Steilacoom; and Columbia R., 200 miles above mouth). Jordan and Jouy, 1882:5 (Puget Sound; Co- lumbia R. ; Mare Island and Sacramento R., California) . Uranidea asp era. Jordan and Gilbert, 1883:694 (description; synonymy; streams west of the Sierra Nevada and Cascade Moun- tains). Jordan, 1885: 110 (subgenus Cottopsis and a list of species therein). Trachidermus richardsonii. Heckel, 1840:162 (synonymy; Columbia R.), (not Cottus richard- soni of Agassiz, 1850). Note: Girard, 1852:62, erred in reporting the date of Heckel’ s paper as 1837, and in the spelling of Trachidermus. Cottopsis parvus. Girard, 1856^:144 (orig- inal description ; Presidio on San Francisco Bay, California). Girard, 1857:11 (descrip- tion; Presidio, and Monterery, California). Girard, 1859:54 (description; synonymy; Monterey, Presidio, Fort Reading, and Peta- luma, California). Cooper, 1868:492 (listing only). Jordan, 1877:5 (as the young of Cottop- sis asper). Centridermichthys parvus. Gunther, I860: 170 (description; synonymy; fresh waters of California). Lord, 1866^:352 (listing; "fre- quenting the same localities as . . .” Centri- dermichthys asper). Uranidea aspera var. parvus. Jordan and Gilbert, 1883:694 (Sacramento R. forms). Cottus gulosus parvus. Jordan and Ever- mann, 1898:1945, and Jordan, Evermann, and Clark, 1930:383 ( Cottopsis parvus’. Monterey, Presidio, Fort Reading, and Petaluma, Cali- fornia) . 244 PACIFIC SCIENCE, Vol. XXI, April 1967 Uranidea gulosa, in part. Jordan and Gilbert, 1883:695 (misidentifications: all specimens from "Vancouver’s Island" and probably those from "about Port Townsend,” cf. Jordan and Evermann, 1898:1944). Kermode, 1909:87 (listing only; misidentification : Shawnigan L., Vancouver Island, cf. Fowler, 1923:282). Cottus gulosus, in part. Jordan, 1895:141 (misidentification: San Luis Cr., near Avila, California, cf. Hubbs, 1921:7). Jordan and Evermann, 1898:1945 (misidentification: all specimens from San Franciscito Cr., Santa Clara Co., California). Jordan, Evermann, and Clark, 1930:383 (probable misidentifications: speci- mens from Loring and Boca de Quadra, Alaska). Evermann and Clark, 1931:57 (mis- identifications: Presidio, Monterey, Fort Read- ing, Petaluma, and San Luis Cr., California). Evermann and Clark, 1931:12, 13 (misidenti- fications: Monterey, Presidio, Fort Reading, and Petaluma). Bean and Weed, 1920:76 (ques- tionable identification: 4 specimens from Victoria, Vancouver Island, British Columbia). Wilimovsky, 1954:285 (doubts validity of southeast Alaska record). Centridermichthys gulosus. Lord, 1866£: 352 (listing; "frequenting the same localities as . . .” Centridermichthys asper). Cottus sp. Dill, 1946:54 (San Joaquin R., near Friant, California; identification as asper by Hubbs and Wallis, 1948:141). NOMENCLATURAL HISTORY The specific name asper is currently well founded in the genus Cottus, to which it was originally designated by Richardson in 1836. But, as shown in the preceding synonymy, the binomen was extremely unstable for the first 100 years after its introduction. After an initial period of uncertainty regarding its affinity to marine or to fresh-water Cottoids, three main nomenclatural difficulties are encountered: the often-repeated misidentification as Cottus gulo- sus (Girard) ; the failure to recognize that Cottopsis parvus and Cottus asper are con- specific; and the failure to recognize the specific relationship of asper to other species in the genus Cottus. The almost immediate placement of asper into Trachidermus by Heckel (1840), and then into the synonymous Centridermichthys by Richardson (1844), reflects the early opinion that asper was more closely allied to the marine Cottoids. Girard (1851 1852) recognized its affinities with the fresh-water genus Cottus, but distinguished it from that genus by erecting the genus Cottopsis, based on the presence of palatine teeth and the "skin beset with prickles, instead of being smooth and scaleless." Lacking any specimens, Girard defined Cottopsis on the basis of Richardson’s original description but, on p. 63, where he quoted Richardson’s entire discussion of prickles (p. 295), he misquoted Richardson by attributing to him the statement, "There are no scales." Girard’s lack of speci- mens proved unfortunate since soon thereafter (1856£) he named and described Cottopsis parvus from the Presidio (in San Francisco), California, comparing it not with C. asper but with Cottopsis gulosus Girard, also newly described (1856^) from the San Joaquin R., California. In his later report (1859), Girard had 8 specimens of C. asper in his possession, from the Columbia R. and Puget Sound. Obviously he again failed to recognize the con- specificity of asper and parvus, and he followed Richardson’s original description rather than comparing them with specimens of parvus, which he seems to have reserved for com- parison with gulosus. Jordan (1877) referred, in passing, to Cot- topsis parvus as the young of C. asper. Jordan and Jouy (1882), however, Fsted specimens of Cottopsis asper from Mare Island and Sacra- mento R., California, and from Puget Sound and the Columbia R. Less than a year later, Jordan and Gilbert (1883) placed asper in the genus Uranidea DeKay, subgenus Cottopsis, based on the presence of palatine teeth and the gill membranes being broadly united to the isthmus. In the same report, they referred to the Sacramento R. form of U. asp era as "var. parvus, smaller in size, paler in color and with the interorbital space concave, narrower than eye." Eigenmann (1895) used the valid name to describe specimens from the Fraser and Colum- bia rivers, as also did Gilbert and Evermann (1895), who suggested that the nominal spe- Systematics of Prickly Sculpin, I — Krejsa 245 cies was separable "at least subspecifically from the Sacramento River form." Seale (1896) and Gilbert (1896) used the valid name for north- ern specimens. But, obviously, Jordan (1895) and Jordan and Evermann (1896) still thought in terms of a distinct Californian species ( gulo - sus ) and a distinct northern species ( asper )• Jordan misidentified a specimen of asper from San Luis Cr., near Avila, Calfornia, as gulo sus. Jordan and Evermann listed the range of the nominal species from Vancouver Island to Oregon, and of gulosus, from California Coast Range streams and inland in the San Joaquin R. Gilbert (1896) referred to Cottus asper of the Columbia and Cottus gulosus of the Sacramento as "two species so extremely similar that it is difficult to distinguish them.” Jordan and Ever- mann (1898) repeated the suggestion of Gil- bert and Evermann that the nominal species is separable, at least subspecifically, from the Sacramento R. form, " Cottus gulosus .” That Jordan and Evermann perceived neither the conspecific relationship of parvus to asper nor the limits of the valid species Cottus gulo- sus becomes more fully evident on the next page (p. 1945) of their 1898 report. Their description of Cottus gulosus (Girard) is taken from misidentified specimens of C. asper col- lected in San Franciscito Cr., Santa Clara Co., California. These were large specimens "3 to 7 inches in length” and, most significantly, the count for anal rays is given as "A. 16 to 18.” Both of these characters separate asper from gulosus. Furthermore, they include Cottopsis parvus Girard, from Monterey, the Presidio, Fort Reading, and Petaluma, California, in the synonymy of gulosus. Snyder (1905) collected and correctly identi- fied Cottus asper from the same locality, San Franciscito Cr. He was probably the only one of his time to understand and explain the true relationships of asper, parvus, and gulosus. On p. 337, he stated: Recent authors have identified the common Sacra- mento form which represents the Cottus asper of the Columbia River with the Cottopsis gulosus of Girard. They have sometimes considered the Sacramento form as identical with C. asper and have placed the name gulosus in the synonymy of the latter. At other times they have considered the species as a slightly differ- entiated form worthy of recognition in nomenclature, and have used the name gulosus to designate it. The former view concerning the species is probably cor- rect. The association of the name gulosus with it, however, is without warrant. The latter belongs to a species easily distinguished from C. asper, differing notably in having a much shorter anal fin. There are usually fewer dorsal spines and rays, a more limited distribution of prickles, and an almost uni- form absence of palatine teeth. In C. asper the dor- sal has 8 to 10 spines and 19 to 22 articulated rays, the anal 16 to 18 rays, while in C. gulosus the dorsal has 7 to 9 spines, 17 to 18 rays, the anal 12 to 14 rays. Snyder then continues with a note on habitat preference: In its distribution C. asper appears to be largely confined to the lower courses of the streams, being especially abundant near tide water, while C. gulosus is found further up, where the water is clear and the current rapid. Rutter (1908) correctly placed Cottopsis parvus into the synonymy of Cottus asper, pre- sumably recognizing that the two were con- specific. However, he incorrectly synonymized Uranidea semiscabra centropleura Eigenmann and Eigenmann, which is properly referable to Cottus gulosus. Snyder was the first to consider a series of specimens of the nominal species throughout its entire range, as then known, and, in the same issue of the Bulletin of the Bureau of Fisheries in which Rutter had correctly synonymized parvus, he noted the extreme variation of prick- ling investment. While recognizing the varia- tion between streams, he also noted that the prickling variation is common among individ- uals from the same stream. Although explicitly aware of Snyder’s com- ments on asper and gulosus, Evermann and Goldsborough (1907^) identified 16 specimens of gulosus from Loring and Boca de Quadra, Alaska. The reliability of these identifications is questionable. Kermode’s listing (1909) of Uranidea gulosa from Shawnigan L., Vancouver Island is probably a misidentification, since Fowler (1923:282) listed the same specimen as Cottus asper. Snyder (1913, 1916) again recorded the occurrence of C. asper and gulosus in differing habitats of the same stream. Hubbs (1921) recognized Jordan’s earlier misidentification of gulosus from San Luis Cr., California. He also commented on the variability of prickling in C. asper from several streams. 246 PACIFIC SCIENCE, Vol. XXI, April 1967 It would seem that with the accession of Snyder’s insight into the problem, the valid name was destined for stability. However, Jordan, Evermann, and Clark (1930) repeated the earlier error of Jordan and Evermann (1898) by including Cottopsis parvus as a synonym of Cottus gulosus. In a similar man- ner, they also incorporated the error of Ever- mann and Goldsborough (1907^), previously cited, by including the misidentified specimens of asper from Loring and Boca de Quadra, Alaska, in the list of records for gulosus. In the same work, Jordan and Evermann extended the range of asper (cited in 1898 as: "streams of the Cascade Range, from Vancouver Island to Oregon”) by appending the phrase "south to Sacramento River.” Evermann and Clark (1931) also retained Cottopsis parvus in the synonymy of C. gulosus and perpetuated Jor- dan’s misidentification of the San Luis Cr. gulosus, which Hubbs had correctly re-identi- fied as asper ten years before (1921). Schultz and DeLacy’s catalogue (1936) in- cluded a comprehensive listing of Washington and Oregon localities for C. asper. However, some remain doubtful since Schultz and DeLacy frequently misidentified C. asper as gulosus and/or per plexus. They also incorrectly main- tained the presence of C. gulosus in Alaska. Robins and Miller (1957) presumably over- looked the earlier citation of Rutter (1908) and removed Cottopsis parvus from the syn- onymy of gulosus, placing it in the synonymy of asper, supposedly for the first time. McAllister and Lindsey (1959) first sug- gested the probable existence of "coastal” and "non-coastal” populations of Cottus asper on the basis of morphological and, perhaps, be- havioral differences. Bond (1961) hinted at the possibility of polytypy in Cottus asper when he stated in his key that the body is "well covered with prickles, especially in inland waters and in young in- dividuals from coastal waters.” Bond (1963) gave the most comprehensive treatment yet recorded for Cottus asper and 12 other species in the genus. His study, however, was more concerned with interspecific rather than intra- specific relationships within the genus. Bailey and Bond (1963) indicated their concern for the supraspecific relationships within the genus Cottus by their recognition of several species groups, one of which is the " asper species group.” Krejsa (1965) offered morphological, be- havioral, and distributional evidence for genetic divergence between "coastal” and "inland” populations of C. asper. Bohn and Hoar (1965) offered physiological evidence in sup- port of Krejsa’ s hypothesis. Unfortunately, their brief introductory remarks regarding the life histories and prickling patterns are some- what inaccurate, and therefore misleading, in- terpretations of Krejsa’ s unpublished thesis. These minor points will be clarified in a future publication. DISTRIBUTION Range. Pacific Slope drainage of North Am- erica: coastal streams from Seward, Alaska, to Ventura R., California; lakes and streams of the Queen Charlotte Islands and Vancouver Island; and all major Pacific drainages from the headwaters of the Stikine R. in British Columbia, to the Kern R., San Joaquin R. drainage, California. The following are ex- ceptions: Fraser R. in the area of the Rocky Mountain Trench, east of Prince George, British Columbia (area not yet collected) ; Kootenay Lake drainage of the Columbia R. in British Columbia; Upper Snake R. of the Co- lumbia R. drainage in Washington and Ore- gon; Middle Fork of the Willamette R. in Oregon, above Oakridge ; Klamath R. Basin in Oregon; and Sacramento R. drainages above Lake Shasta, California. Arctic Slope drainage of North America: headwaters of the Peace R. in British Columbia: from Summit L. to McLeod L., Crooked R. drainage; from Tacheeda L., Parsnip R. drainage; from Tchentlo L., Nation R. drainage. The present distributional range of Cottus asper is illustrated in Figure 1. REFERENCES Bailey, R. M., and C. E. Bond. 1963. Four new species of freshwater sculpins, genus Cottus, from western North America. Occas. Papers Mus. Zool. Univ. Mich. 634:1-27. — and M. F. Dimick. 1949. Cottus Systematics of Prickly Sculpin, I — Krejsa 247 Fig. 1. Distributional range of Coitus asper. 248 PACIFIC SCIENCE, Vol. XXI, April 1967 hubbsi, a new cottid fish from the Columbia River system in Washington and Idaho. Occas. Papers Mus. Zool. Univ. Mich. 513:1-18. Bean, B. A., and A. C. Weed. 1920. Notes on a collection of fishes from Vancouver Island, B. C. Trans. Roy. Soc. Canada (3) 13 (1919):69-83. Bohn, A., and W. S. Hoar. 1965. The effect of salinity on the iodine metabolism of coastal and inland prickly sculpins, Cottus asper Richardson. Canadian J. Zool. 43 (1965) :977-985. Bond, C. E. 1961. Keys to Oregon freshwater fishes. Agr. Expt. Sta., Oregon State Univ., Tech. Bull. 58:1-42. 1963. Distribution and ecology of fresh- water sculpins, genus Cottus, in Oregon. Unpublished Ph.D. thesis, Univ. Mich. 186 pp. Carl, C. G., W. A. Clemens, and C. C. Lindsey. 1959. The fresh-water fishes of British Columbia. British Columbia Prov. Mus. Handb. 5 (3rd ed. rev.) :1-192. Cooper, J. G. 1868. Fishes. In: J. G. Cooper and T. F. Cronise, The Natural Wealth of California. Chap. 7, pp. 487-498. H. H. Bancroft and Co., San Francisco. Crawford, D. R. 1927. Extension of the range of Cottus aleuticus. Copeia 1927(160) :177- 178. Dill, W. A. 1946. A preliminary report on the fishery of Millerton Lake, California. Calif. Fish and Game 32(2):49-70. Dymond, J. R. 1936. Some fresh-water fishes of British Columbia. Rept. British Columbia Comm. Fish, (for 1935):60-73. Eigenmann, C. H. 1895. Results of explora- tions in western Canada and northwestern United States. Bull. U. S. Fish Comm. 14 (for 1894) :101-132. Evermann, B. W., and H. W. Clark. 1931. A distributional list of the species of freshwater fishes known to occur in California. Calif. Div. Fish and Game, Fish Bull. 35:1-67. and E. L. Goldsborough. 1901a. The fishes of Alaska. Bull. U. S. Bur. Fish. 26 (for 1906) :219-360. - 1901b. A check list of the freshwater fishes of Canada. Proc. Biol. Soc. Washington 20:89-120. and H. B. Latimer. 1910. On a col- lection of fishes from the Olympic Peninsula, together with notes on other West Coast species. Proc. Biol. Soc. Washington 23: 131-140. and S. E. Meek. 1898. A report upon salmon investigations in the Columbia River Basin and elsewhere on the Pacific Coast in 1896. Bull. U. S. Fish Comm. 17(for 1897): 15-84. Fowler, H. W. 1923. Records of West Coast fishes. Proc. Acad. Nat. Sci. Philadelphia 75:279-301. Gilbert, C. H., 1896. The ichthyological col- lections of the U. S. Fish Commission Steamer Albatross during the years 1890 and 1891. Rept. U. S. Comm. Fish and Fish, (for 1893) , Appendix 6:393-476. 1898. Fishes of the Klamath Basin. Bull. U. S. Fish Comm. 17(for 1897):1-13. and B. W. Evermann. 1895. A report upon investigations in the Columbia River Basin, with descriptions of four new species of fishes. Bull. U. S. Fish Comm. l4(for 1894) : 169-207. Girard, C. 1850. A monograph of the fresh- water Cottus of North America. Proc. Am. Assoc. Adv. Sci. 2:409-411. — 1851 a. On the genus Cottus Auct. Proc. Boston Soc. Nat. Hist. 3(1848-1851) :183- 190. 1851 A Some additional observations on the nomenclature and classification of the fishes of the genus Cottus. Proc. Boston Soc. Nat. Hist. 3(1848-1851) :302-305. 1851c. Revision du genre Cottus des j Auteurs. Nouv. Mem. Soc. Helvet. Sci. Nat. ; 12(1851) :185. [Not seen.] 1852. Contributions to the natural his- tory of the freshwater fishes of North Am- erica. I. A monograph of the cottoids. Smith- sonian Contr. Knowledge 3 (art. 3): 1-80. 1856^. Descriptions of new fishes, col- lected by Dr. A. L. Heermann, naturalist attached to the Survey of the Pacific Railroad Route, under Lieut. R. S. Williamson, U.S.A. Proc. Acad. Nat. Sci. Philadelphia 7(1854- 1855) :129— 141. 1856A Observations upon a collection of fishes made on the Pacific coast of the United States, by Lieut. W. P. Trowbridge, Systematics of Prickly Sculpin, I — Krejsa 249 U.S.A., for the museum of the Smithsonian Institution. Proc. Acad. Nat. Sci. Philadelphia 7(1854-1855) :142-156. 1857. Report upon fishes collected on the Survey. 6 (Part IV, No. 1) :9-34. In: House of Representatives, 33rd Congr., 2nd Sess., Exec. Doc. 91, Reports of explorations and surveys for a railroad route from the Mississippi River to the Pacific Ocean. Vol. 6: Report of Lieut. Henry L. Abbot upon explorations for a railroad route from the Sacramento Valley to the Columbia River, made by Lieut. R. S. Williamson, assisted by Lieut. Henry L. Abbot. 1854—1855. Wash- ington, D.C. — 1859- Fishes. In: General report upon the zoology of the several Pacific railroad routes. U. S. Pac. R. R. Expl. and Surv. 10 (part 4) : 1-400. Also found as Part IV, 33rd Cong., 2nd Sess., Senate Exec. Doc. 78. Gunther, A. C. L. G. I860. Catalogue of the Acanthopterygian Fishes in the Collection of the British Museum, Vol. 2, pp. 1-548. Heckel, J. J. 1840. Ichthyologische Beitrage zu den Familien der Cottoiden, Scorpae- noiden, Gobioiden und Cyprinoiden. Ann. Wien. Mus. Naturgeschichte 2(1840) :143- 164. Hubbs, C. L. 1921. Notes on Cottus asper and Cottus aleuticus. Copeia 1921(9) :7— 8. and O. L. Wallis. 1948. The native fish fauna of Yosemite National Park and its preservation. Yosemite Nature Notes 27 (12) : 131-144. Jordan, D. S. 1877. Contributions to North American ichthyology, No. 2. Bull. U. S. Natl. Mus. 10:1-120. 1885. A catalogue of the fishes known to inhabit the waters of North America, north of the Tropic of Cancer, with notes on the species discovered in 1883 and 1884. In: Rept. U. S. Comm. Fish and Fish, (for 1884) :1-185. Also appeared as Appendix E, Rept. U. S. Comm. Fish. 1885 (1887): 789-973. ■ 1895. Notes on the fresh-water species of San Luis Obispo County, California. Bull. U. S. Fish Comm. l4(for 1894) :141-142. 1905. A Guide to the Study of Fishes. Vol. II, pp. 1-599. Henry Holt and Co., New York. 1919. The genera of fishes. Part II. From Agassis to Bleeker, 1833-1858, twenty- six years, with the accepted type of each. Le- land Stanford, Jr. Univ. Publ., Univ. Ser. 1919:1-284. and B. W. Evermann. 1896. A check list of the fishes and fish-like vertebrates of North and Middle America. Rept. Comm., U. S. Comm. Fish and Fish, (for 1895), Appendix 5:207-584. 1898. The fishes of North and Middle America. Part II. Bull. U. S. Natl. Mus. 47:1241-2183. — B. W. Evermann, and H. W. Clark. 1930. Check list of the fishes and fish-like vertebrates of North and Middle America north of the northern boundary of Venezuela and Colombia. Rept. U. S. Comm. Fish, (for 1928), Part 2:1-670. — and C. H. Gilbert. 1883. Synopsis of the fishes of North America. Bull. U. S. Natl. Mus. 16(1882) :1-1018. ■ and P. L. Jouy. 1882. Check list of duplicate fishes from the Pacific Coast of North America, distributed by the Smith- sonian Institution in behalf of the United States National Museum. Proc. U. S. Natl. Mus. 4 (1881) :1-18. Kermode, F. 1909. Visitor’s guide to the nat- ural history and ethnological collections in the Provincial Museum. Prov. Mus. Nat. Hist, and Ethnol., Victoria, British Columbia 1909: 1-92. 1917. Fishes. In: Rept. Prov. Mus. Nat. Hist., Prov. British Columbia (1916) : 18-22. 1931. Accessions: Ichthyology. In: Rept. Prov. Mus. Nat. Hist., Prov. British Columbia (1930) :17-23. Krejsa, R. J. 1965. The systematics of the prickly sculpin, Cottus asper : an investigation of genetic and non -genetic variation within a polytypic species. Unpublished Ph.D. thesis, Univ. British Columbia, June, 1965. 109 pp. Lindsey, C. C. 1956. Distribution and taxon- omy of fishes in the Mackenzie drainage of British Columbia. J. Fish. Res. Bd. Canada 13(6) :759-789. 1957. Possible effects of water diver- sions on fish distribution in British Columbia. J. Fish. Res. Bd. Canada 14(4) :651-668. 250 PACIFIC SCIENCE, Vol. XXI, April 1967 Lord, J. K. 1866^. The Naturalist in Vancouver Island and British Columbia. Vol. 1. Richard Bentley, London. 358 pp. 1866A The Naturalist in Vancouver Is- land and British Columbia. Vol. 2. Richard Bentley, London. 375 pp. McAllister, D. E. 1957. The systematics of the freshwater sculpins of British Columbia. Unpublished M.A. thesis, Univ. British Co- lumbia, April, 1957. 109 pp. I960. List of the marine fishes of Can- ada. Natl. Mus. Canada Bull. 168, Biol. Ser. 62:1-76. and C. C. Lindsey. 1959. Systematics of the freshwater sculpins ( Cottus ) of British Columbia. Natl. Mus. Can. Contr. Zool., Bull. 172:66-89. Meek, S. E. 1899- Notes on a collection of cold- blooded vertebrates from the Olympic Moun- tains. Field Mus. Nat. Hist. Publ. 31, Zool. Ser. (1897) 1 (12) :225-236. Nichols, J. T. 1909. A small collection of Alaskan fishes. Proc. Biol. Soc. Washington 21:171-174. Richardson, Sir J. 1836. Fauna Boreali- Amer- icana. Part 3. The Fish. Richard Bentley, London. 327 pp. 1844. Ichthyology. In: R. B. Hinds, ed., Voyage of the H.M.S. "Sulphur,” under the command of Captain Sir Edward Belcher, during the years 1836-42. [Not seen.] Robeck, C. G., C. Henderson, and R. C. Palange. 1954. Water quality studies on the Columbia River. U. S. Dept. HEW, Publ. Health Serv., R. A. Taft Sanitary Engr. Cen- ter, Cincinnati. 99 pp. + 4 appendices. Robins, C. R., and R. R. Miller. 1957. Clas- sification, variation and distribution of the sculpins, genus Cottus , inhabiting the Pacific Slope waters in California and southern Ore- gon, with a key to the species. Calif. Fish and Game 43(3) :213-233. Rutter, C. 1908. The fishes of the Sacramento- San Joaquin Basin, with a study of their dis- tribution and variation. Bull. U. S. Bur. Fish. 27 (for 1907) :103-152. Schultz, L. P. 1929. Check-list of the fresh- water fishes of Oregon and Washington. Publ. Fish., Univ. Washington Coll. Fish. 2(4) :43-50. 1930. Notes on the species of Cottus in western Washington. Copeia 1930(1) : 14-1 5. 1936. Keys to the fishes of Washington, Oregon and closely adjoining regions. Univ. Washington Publ. Biol. (Dec., 1936), 2(4) : 103-228. and A. C. DeLacy. 1936^. Fishes of the American Northwest. A catalogue of the fishes of Washington and Oregon, with dis- tributional records and a bibliography. In- stallment 3. Mid-Pacific Mag. (April-June) 1936:127-142. 1936 A Fishes of the American Northwest. Installment 4. Mid-Pacific Mag. ( July-Sept. ) 1936:211-226. Seale, A. 1896. List of the fresh-water fishes collected in the vicinity of Seattle, Washing- ton, by Edwin C. Starks. Proc. Calif. Acad. Sci. (2)5 (for 1895), part 2:852-854. Shapovalov, L., and W. A. Dill. 1950. A check list of the fresh-water and anadromous fishes of California. Calif. Fish and Game 36(4):382-391. and A. J. Cordone. 1959. A revised check list of the fresh water and anad- romous fishes of California. Calif. Fish and Game 45(3) : 159-180. Snyder, J. O. 1905. Notes on the fishes of the streams flowing into San Francisco Bay. Rept. Bur. Fish, (for 1904), App. 5:327-338. 1908^. The fauna of the Russian River, California, and its relation to that of the Sacramento. Science, N. S. 27:269-271. 1908 A The fishes of the coastal streams of Oregon and northern California. Bull. U. S. Bur. Fish. 27(for 1907) : 153-189. — 1913. The fishes of the streams tributary to Monterey Bay, California. Bull. U. S. Bur. Fish. 32(for 1912) :47-72. 1916. Fishes of the streams tributary to Tomales Bay. Bull. U. S. Bur. Fish. 34(for 1914) :375— 381. Storer, D. H. 1846^. A synopsis of the fishes of North America. Mem. Am. Acad. Arts and Sci., N. S. 2:253-550. 1846A A Synopsis of the Fishes of North America. Metcalf and Co., Cambridge. 298 pp. Suckley, G. 1859. Report upon the fishes col- lected on the Survey. Chap. II: Report upon the fishes exclusive of the Salmonidae, pp. Systematics of Prickly Sculpin, I — Krejsa 350-368. In: J. G. Cooper and G. Suckley, The Natural History of Washington Terri- tory. Part III. Bail lire Brothers, New York. 399 pp. I860. Report upon the fishes collected on the Survey. Chap. 2: Report upon the fishes exclusive of the Salmonidae. U. S. Pacif. R.R. Rxpl. and Surv. (for 1853-55), 12(2) part 3 (5) : 350-368. Also appears as: House of Representatives Exec. Doc. 56, Washington, I860, pp. 350-368. Sumner, F. H. 1942. A reconnaissance survey of Oregon coastal streams. Rept. Oregon 251 State Game Comm., pp. 1-25 (mimeo). In: C E. Bond, 1963, p. 81. 1953. Migrations of salmonids in Sand Creek, Oregon. Trans. Am. Fish. Soc. 82 (1952) : 139-1 50. Taft, A. C. 1934. California steelhead experi- ments. Trans. Am. Fish. Soc. 64(1933) : 248— 251. Wilimovsky, N. J. 1954. Fist of the fishes of Alaska. Stanford Ichthyol. Bull. 4:279-294. 1958. Provisional Keys to the Fishes of Alaska. FJ. S. Fish and Wildlife Serv., Fish. Res. Fab., Juneau. 113 pp. Chromosomes of Some Opisthobranchiate Mollusks from Eniwetok Atoll, Western Pacific1 J. B. Burch2 and R. Natarajan3 ABSTRACT: Chromosome numbers are reported for nine species of opistho- branchiate mollusks from Eniwetok Atoll, Marshall Islands, western Pacific. In the Nudibranchia, both Dendrodoris nigra (Dendrodorididae) and Herviella mietta (Favorinidae) were found to have 13 bivalents during male meiosis. In the Anaspidea, Dolabrifera dolabrifera and Stylocheilus longicauda (Aplysiidae) both had 17 bivalents during male meiosis. In the Cephalaspidea, Haminoea linda and H. musetta (Atyidae) each had 17 pairs of chromosomes during male meiosis and Lathophthalmus smaragdinus and Smaragdinella calyculata (Smaragdinellidae) had 18 pairs. In the Soleolifera, Onchidella evelinae had 18 bivalents during male meiosis. The extreme conservativeness of chromosome numbers in opisthobranchiate mol- lusks is demonstrated by that fact that all 18 nudibranchs from 10 families studied so far have the single haploid chromosome number 13, and that 18 of the 21 spe- cies of the orders Entomotaeniata, Anaspidea, Cephalaspidea, and Sacoglossa have 17 pairs of chromosomes. The haploid number 18 is here reported for the first time for nonsoleoliferan opisthobranchiate mollusks. The more advanced, mostly fresh-water, order Basommatophora, in which the haploid number 18 is the basic number, may well have been derived from a taxon within or related to this cepha- laspid superfamily (Philinacea) . In recent years, detailed investigations have been made on the chromosomes of many ba- sommatophoran and stylommatophoran snails (Burch, 1965), but relatively few species of the opisthobranchiate4 orders have been studied, 1 The field work for this investigation was sup- ported by the Division of Biology and Medicine, U. S. Atomic Energy Commission. The research was supported (in part) by grants from the National Science Foundation, Washington, D. C. (GB 787) and the National Institute of Allergy and Infectious Diseases, U. S. Public Health Service (5 Tl AI 41-08). Manuscript received March 14, 1966. 2 Museum and Department of Zoology, University of Michigan, Ann Arbor. Supported by a Public Health Service research career program award (No. 5-K3-AI-19, 451-03). 3 Museum of Zoology, University of Michigan, Ann Arbor. 4 Boettger (1955) considers the orders Opistho- branchiata and Pulmonata (together equivalent to the subclass Euthyneura) to be unnatural ones, and in- stead divides the Euthyneura into seven orders. While we do not disagree with this, it is still convenient to speak of his five lower (mainly marine) euthy- neuran orders (Nudibranchia, Soleolifera, Cephala- spidea, Sacoglossa, and Anaspidea) as "opistho- mainly because of the difficulties they present in collection and identification. Relying on the studies of various authors during the early part of the present century, Makino (1951) listed the chromosome numbers of 16 opisthobran- chiate species, but recent investigations by Inaba and our present studies indicate that the earlier records are not dependable and so are obsolete. Previous reliable reports on the chromosomes of opisthobranchiate gastropods are those of Inaba and Hirota (1954, 1958), Inaba (1959*, 1959£, 1961), Natarajan (1959, I960), Mancino and Sordi (1964^ and b), and Burch (1965). These authors give the chromosome numbers of 36 species belonging to 21 families and 7 orders (Tables 1 and 2). This is a very small number when compared with the great multitude of species currently recognized in the opisthobran- chiate orders. The present paper presents the chromosome numbers of 9 opisthobranchiate branchs” as distinguished from the more advanced Basommatophora (mainly freshwater) and Stylomma- tophora (land inhabitants). 252 Chromosomes of Eniwetok Opisthobranchiates — Burch and Natarajan 253 TABLE 1 Chromosome Numbers Previously Reported in the Opisthobranchiate Orders Notaspidea and Nudibranchia SPECIES CHROMOSOME NUMBER 2n n LOCALITY REFERENCE NOTASPIDEA Pleurobranchidae Pleurobranchaea n o vaezealandiae 24 12 Japan Inaba, 195 9a NUDIBRANCHIA Dorididae Glossodoris f estiva 13 Japan Inaba and Hirota, 1958 Glossodoris pallescens 13 Japan Inaba and Hirota, 1954; 1958 Rostanga arbutus 26 13 Japan Inaba, 195 9a Discodoris pardalis 13 Japan Inaba, 1959^ Doris verrucosa 26 13 Italy Mancino and Sordi, 1964a Dendrodorididae Dendrodoris miniata 26 13 Japan Inaba and Hirota, 1958 Dendrodoris nigra 26 13 Japan Inaba and Hirota, 1958; Triophidae Kaloplocamus ramosus 13 Japan Inaba, 1959^ Inaba and Hirota, 1958 Plocamopherus tilesii 26 13 Japan Inaba and Hirota, 1958 Goniodorididae Okenia barnardi 13 Japan Inaba and Hirota, 1958 Fimbriidae Melibe papillosa 26 13 Japan Inaba, 1959a Dotonidae Doto bella 13 Japan Inaba, 1961 Arminidae Dermatobranchus striatus 26 13 Japan Inaba and Hirota, 1958 Cuthonidae Catriona pupillae 13 Japan Inaba, 1961 Facelinidae Facelina ceylonica 13 Japan Inaba and Hirota, 1958 Facelina japonica 13 Japan Inaba and Hirota, 1958 species belonging to 6 families and 4 orders (Table 3), which were collected in shallow waters around three islands of Eniwetok Atoll in the western Pacific by the senior author and Dr. William H. Heard during early I960. MATERIAL AND METHODS The species studied in this investigation and the localities where they were collected are listed below. Duplicate specimens have been deposited in the collections of the Museum of Zoology at the University of Michigan, the University of Sao Paulo, and the University of Hawaii. NUDIBRANCHIA 1. Dendrodoris nigra (Stimpson) (Fig. 1). North end of Japtan Island, under loose pieces of dead coral. April 1, I960. 2. Herviella mietta Marcus and Burch (Fig. 2). North end of Eniwetok Island on the la- goon side, in about 10 cm of water at low tide, under submerged pieces of dead coral. April 2, I960. ANASPIDEA 3. Dolabrif era dolabrif era (Rang) (Fig. 3). Under loose pieces of coral on seaward tide flat at the north end of Parry Island. March 25, I960. 254 PACIFIC SCIENCE, Vol. XXI, April 1967 TABLE 2 Chromosome Numbers Previously Reported in the Opisthobranchiate Orders Entomotaeniata, CEPHALASPIDEA , Anaspidea, Sacoglqssa, AND SOLEOLIFERA SPECIES CHROMOSOME NUMBER 2n n LOCALITY REFERENCE ENTOMOTAENIATA Pyramidellidae Tiberia fas data 17 Japan Inaba, pers. comm. ANASPIDEA Aplysiidae Petalifera punctulata 34 17 Japan Inaba, 1959<* Notarchus leachti freeri 17 Japan Inaba, 1959 a CEPHALASPIDEA Acteonidae Cylichnatys angusta 17 Japan Inaba, pers. comm. Philinidae Philine japonica 17 Japan Inaba, 1959^ Aglajidae Aglaja gigliolii 34 17 Japan Inaba, 1959^ SACOGLOSSA Elysiidae Elysia amakusana 17 Japan Inaba, 1959^ Elysia viridis 17 Italy Mancino and Sordi, 1964& Stiligeridae Alder ia nigra 17 Japan Inaba, 1961 Hermaeopsis variopicta 17 Italy Mancino and Sordi, 1964& Placida dendritica 34 17 Italy Mancino and Sordi, 19 64& Placida viridis 34 17 Italy Mancino and Sordi, 19 64£ Stiliger vesiculosus 34 17 Italy Mancino and Sordi, 1 964b Juliidae Berthelinia Umax 17 Japan Inaba, 1961 Polybranchiidae Bosellia mimetic a 14 7 Italy Mancino and Sordi, 1964b SOLEOLIFERA Veronicellidae Veronicella floridana 16 U.S.A. Burch, 1965 Laevicaulis alte 17 India Natarajan, I960 Onchidiidae Oncidiella kurodai 17 Japan Inaba, 1961 Onchidium verraculatum 36 18 India Natarajan, 1959 4. Stylocheilus Ion gi cauda (Quoy and Gai- mard) (Fig. 4). In tide flats of Eniwetok Is- land. March 4, i960. CEPHALASPIDEA 5. Haminoea linda Marcus and Burch (Fig. 5). Parry Island, in sand, in about 2 m of water, in lagoon, about 17 m from shore. March 31, I960. 6. Haminoea musetta Marcus and Burch (Fig. 6). Middle part of Parry Island on sea- ward tide flats. April 2, I960. 7. Lathophthalmus smaragdinus (Ruppel and Leuckart) (Fig. 7). Collected at the south end of Parry Island, under loose pieces of coral on seaward tide flat. March 15, I960. 8. Smaragdinella calyculata (Broderip and Chromosomes of Eniwetok Opisthobranchiates — Burch and Natarajan 255 TABLE 3 Chromosome Numbers of Opisthobranchs Observed in This Study SPECIES CHROMOSOME NUMBER (n) NUMBER OF SPECIMENS GIVING RESULTS NUDIBRANCHIA Dendrodorididae Dendrodoris nigra 13 5 Favorinidae Herviella mietta 13 (2n= 26) 1 ANASPIDEA Aplysiidae Dolabrifera dolabrifera 17 9 Stylocheilus longicauda 17 2 CEPHALASPIDEA Atyidae Haminoea linda 17 1 H amino e a mu sett a 17 2 Smaragdinellidae Lathophthalmus smaragdinus 18 2 Smaragdinella calyculata 18 2 SOLEOLIFERA Onchidiidae Onchidella evelinae 18 2 Sowerby) (Fig. 8). In lagoon at north end of Eniwetok Island. SOLEOLIFERA 9. Onchidella evelinae Marcus and Burch (Fig. 9). In cracks in coral slabs above water line (at low tide) on the lagoon side at the north end of Eniwetok Island. April 5, I960. The materials examined consisted of ovo- testes fixed in either Newcomer’s (1953) or Carnoy’s (1887) fluids, or the fixative of San- felice (1918). The material fixed in New- comer’s or Carnoy’s fluids was stained by the acetic-orcein squash technique (La Cour, 1941), and reproductive tissues fixed in Sanfelice’s fluid were sectioned at either 8 or 10 micra and stained with Newton’s (1926) crystal violet. All observations were made on meiotic cells of spermatogenesis (except in Herviella mietta , where spermatogonial cells were also studied) with a Nikon (Nippon Kogaku) microscope equipped with a 100X (n.a. 1.25) oil immer- sion objective and 10X, 20 X, and 30 X ocu- lars. Drawings were made with the aid of a camera lucida and reproduced at a table-top magnification of 4260X- OBSERVATIONS 1. Dendrodoris nigra (Fig. 10). The five individuals of this species on which we were able to obtain satisfactory observations all had 1 3 bivalents during prophase of the first meiotic division. 2. Herviella mietta (Fig. 11). Only one specimen of this species gave satisfactory results. It had 13 bivalents during diakinesis and 26 chromosomes during metaphase in spermato- gonial cells. 3. Dolabrifera dolabrijera (Figs. 12 and 13). Nine specimens had meiotic cells that were satisfactory for chromosome number de- terminations. All had 17 bivalents during Pro- phase I and Metaphase I. The chromosomes of a cell during diakinesis from an acetic-orcein squash preparation are shown in Figure 12. Metaphase I bivalents from a cell of paraffin sectioned material are shown in Figure 13. 256 PACIFIC SCIENCE, Vol. XXI, April 1967 Chromosomes of Eniwetok Opisthobranchiates — Burch and Natarajan 257 1 8f * ft* « 12 11 •if »•*# • 16 ••I 13 15 19 18 #»* % r i * t «• * V 17 Figs. 10-20. Chromosomes of Eniwetok opisthobranchs. 10, Dendrodoris nigra. 11, Herviella mietta. 12 and 13, Dolabrifera dolabrifera. 14, Stylocheilus longicauda. 13, Haminoea musetta. 16, H. linda. 17, Lath- op hthalmus smaragdinus. 18, Smaragdinella calyculata. 19 and 20, Onchidella evelinae. Figs. 10-12, 14, 13, 17, and 20 are of male diakinesis bivalents; 13, 16, 18, and 19 are of male Metaphase I bivalents. Figs. 10-12, 14, 13, 17, 18, and 20 are from acetic-orcein squash preparations; 13, 16, and 19 are from sectioned material stained with crystal violet. Measurement line divided into micra. Figs. 1-9. Eniwetok opisthobranchs used in this study. 1, Dendrodoris nigra. 2, Herviella mietta. 3, Dolabrifera dolabrifera. 4, Stylocheilus longicauda. 3, Haminoea linda. 6, H. musetta. 7, Lathop hthalmus smaragdinus. 8, Smaragdinella calyculata. 9, Onchidella evelinae. Measurement lines are divided into millimeters. 258 PACIFIC SCIENCE, Vol. XXI, April 1967 4. Stylocheilus longicauda (Fig. 14). Satis- factory results were obtained from two speci- mens. Both had 17 bivalents during diakinesis. 5. Haminoea linda (Fig. 16). We were able to obtain cells that were satisfactory for study from only one specimen. These cells had 17 bivalents during Metaphase I. 6. Haminoea musetta (Fig. 15). The two specimens studied both had 17 bivalents in cells at the diakinesis stage. The diakinesis bivalents of one such cell are shown in Figure 15. 7. Lathophthalmus smaragdinus (Fig. 17). Two specimens had meiotic cells that were satis- factory for chromosome number determinations. All dividing cells from which accurate counts could be made had 18 bivalents. 8. Smaragdinella calyculata (Fig. 18). Satis- factory results were obtained from two speci- mens. Both had 18 bivalents during diakinesis and Metaphase I. 9. Onchidella evelinae (Figs. 19 and 20). The two individuals on which we were able to obtain satisfactory observations both had 18 bivalents during diakinesis and Metaphase I. DISCUSSION The chromosome numbers of the eight genera of opisthobranchiate mollusks presented here add to the information previously obtained by reliable authors. Of the species studied five belong to three families not studied in past re- ports, i.e., the Favorinidae (Nudibranchia) , and the Atyidae and Smaragdinellidae (Cephala- spidea) . Dendrodoris nigra (Nudibranchia, Dendro- dorididae) was studied previously by Inaba and Hirota (1958). We found the same number of chromosomes (n=13) for this species from Eniwetok as they reported it to have from Japan. Herviella mietta (Nudibranchia, Favor- inidae) from Eniwetok also had a haploid number of 13, which adds to the growing information regarding the extreme conservative- ness of chromosome numbers of most opistho- branchs. All 16 species of nudibranchs (belong- ing to nine different families) studied so far have this same haploid number, n=13. Among the Cephalaspidea three species have been studied previously (Inaba, 1959^ and per- sonal communication), each belonging to a dif- ferent family. Each of these three species had a haploid number of 17. In the present in- vestigation two species from each of two addi- tional families were studied. Haminoea linda and H. musetta (Bullacea, Atyidae) had a hap- loid chromosome number of 17, but Smarag- dinella calyculata and Lathophthalmus smarag- dinus (Philinacea, Smaragdinellidae) each had the haploid number 18. It will be interesting to see if species of the other families of the Phili- nacea (the Philinidae, Scaphandridae, Agla- jidae, Gastropteridae, and Runcinidae) also have 18 pairs of chromosomes. If so, this would separate this superfamily from all other cepha- laspideids and, additionally, from all other cyto- logically known entomotaenids and anaspideids and from most of the sacoglossans. The haploid number 18 in this group may have another significance in that it seems to strengthen Pel- seneer’s (1893) and Boettger’s (1955) views regarding the origin of the Basommatophora from the Cephalaspidea, since the haploid num- ber 18 is basic for the Basommatophora. Dolahrifera dolahrifera and Stylocheilus longicauda (Anaspidea, Aplysiidae) both had the haploid number 17, as did the two species studied from this family by Inaba (1959^). Onchidella evelinae (Soleolifera, Onchidii- dae) had a haploid number of 18, which is one bivalent more than that reported by Inaba (1961^) for O. kurodai of the same genus from Japan, but n=18 is the same number reported by Natarajan (1959) for Onchidium verracu- latum from India. Much more cytological information is desirable for the various species i belonging to this aberrant order, which is some- I times placed with the "pulmonates” (e.g., Baker, 1955). The conservativeness of chromosome number in the opisthobranchs indicates that these mol- lusks are extremely resistant to changes in chromosome numbers, regardless of major evolved morphological diversities within the 1 various groups, and that certain major divisions (i.e., the Nudibranchia and the orders with n=17 and higher chromosome numbers) have probably been separated for an extremely long geological time. In this regard, Bosellia mimet- ica is either an extremely aberrant species, or i its cytological evolution has been much more rapid than has been the evolution of its gross Chromosomes of Eniwetok Opisthobranchiates — Burch and Natarajan 259 morphology in respect to all other opistho- branchs so far studied. ACKNOWLEDGMENTS Grateful acknowledgment is made to the United States Atomic Energy Commission for supporting the study by the senior author at Eniwetok by providing travel funds, logistical support, and use of the facilities of the Eniwe- tok Marine Biological Laboratory. The coopera- tion of the U.S.A.E.C. Eniwetok Field Office and of Holmes and Narver, Inc. greatly facil- itated the field collecting. A note of gratitude is due Dr. I. Eugene Wallen, formerly with the U.S.A.E.C., Dr. Robert W. Hiatt, University of Hawaii, and Prof. Henry van der Schalie, Uni- versity of Michigan, for promoting these studies; to Dr. William H. Heard, Florida State University, for assistance while at Eniwe- tok; and to John L. Tottenham, Staff Artist, Museum of Zoology, University of Michigan, for preparing illustrations of the animals used in this study. REFERENCES Baker, H. B. 1955. Heterurethrous and aula- copod. Nautilus 68:109-112. Boettger, C. R. 1955. Die Systematik der euthyneuren Schnecken. Verh. dtsch. zool. Ges., 1954, pp. 253-280. Burch, J. B. 1965. Chromosome numbers and systematics in euthyneuran snails. Proc. First European Malacol. Congr., pp. 215-241. Carnoy, H. B. 1887. Conference donnee a la Societe beige de Microscopie. Appendice: I. Les Globules polaires de V As cans clavata. La Cellule, Rec. Cytol. Histol. Gen. 3(2/3): 227-273. Inaba, A. 1959^. Cytological studies in mol- luscs. III. A chromosome survey in the opis- thobranchiate Gastropoda. Annot. Zool. Japan 32(2):81-88. ■ 1959 b. On the chromosomes of a nudi- branch gastropod, Dendrodoris ( Dendro - dons') nigra (Stimpson). J. Sci. Hiroshima Univ. 18(9) :95— 98. 1961. Chromosomes of some opistho- branchs. [In Japanese.] Dobutsugaku Zasshi 70:24. — and R. Hirota. 1954. On the chromo- somes of Glossodoris pallescens. [In Japa- nese.] Dobutsugaku Zasshi 63:437. 1958. A chromosome survey in ten species of nudibranchs (Gastropoda, Mollusca). Japan. J. Zool. 12(2) : 157 — 162. La Cour, L. 1941. Acetic-orcein: A new stain- fixative for chromosomes. Stain Technol. 16: 169-174. Makino, S. 1956. An Atlas of the Chromosome Numbers in Animals. 2nd ed. Iowa State Col- lege Press, pp. 22-23- Mancino, G. and M. Sordi. 1 964a. Ricerche cariologiche in Doris verrucosa Cuvier, 1804 (Gasteropodi, Opisthobranchi) del litorale livornese. Atti Soc. Toscana Sci. Nat., Ser. B, 71. 14 pp. - 1964 A II corredo cromosomico di alcuni Opisthobranchi Sacoglossi del mar Tirreno. Ibid. 71. 12 pp. Natarajan, R. 1959. Chromosomes of the slug Onchidium verraculatum Cuv. (Mol- lusca: Gastropoda: Pulmonata). J. Zool. Soc. India ll(l):30-33. I960. Further cytological studies in Pulmonata (Mollusca: Gastropoda). Ibid. 12(l):69-79. Newcomer, E. H. 1953. A new cytological and histological fixing fluid. Science 118(3058): 161. Newton, W. C. F. 1926. Chromosome studies in Tulip a and some related genera. J. Linn. Soc. 47(316) :339-354. Pelseneer, P. 1893. Recherches sur divers Opisthobranches. Academic royale de Bel- gique. Pp. 1-157, pis. 1-25. Sanfelice, F. 191 8. Recherches sur le Genese des corpuscules du Molluscum contagiosum. Ann. l’lnst. Pasteur 32(8) :363-371. Observations on the Biology of the Lousefish, Phtheirichthys line at us (Menzies) Donald W. Strasburg1 The lousefish, Phtheirichthys lineatus (Men- zies), is a slender member of the Echeneidae which is often free-swimming but which also attaches to immotile objects or slow-swimming fishes. It is pantropical, uncommon, and little known biologically. This paper reports on 24 lousefish in the collections of the Bureau of Commercial Fisher- ies Biological Laboratory in Honolulu. Capture data for these specimens are presented in Table 1. Most of the other data were derived from preserved fish with the exception of one 256- mm specimen which was maintained alive in a swimming pool for several weeks. Mr. Richard D. Samuels and Mr. Richard N. Uchida captured the fish, Mr. Everet C. Jones identified many of the food items, and Mr. Tamotsu Nakata prepared the figure. All are employees of the Bureau of Commercial Fish- eries Biological Laboratory, Honolulu. COMMON NAME The American Fisheries Society (1960:48) lists the common name of Phtheirichthys line- atus as "slender suckerfish." P. lineatus is an uncommon or rare fish, however, and no one of my acquaintance applies this rather arbitrary name to it. In Honolulu, Phtheirichthys is usu- ally called "lousefish," a name used earlier by Jordan (1907:680). "Lousefish" is used throughout this report. ATTACHMENT The data in Table 1 were grouped to reflect the various objects to which lousefish attached. Most small fish (ca. 40-130 mm standard length) came from essentially motionless ob- jects: buoys, baits, and large dead fish hanging 1 Bureau of Commercial Fisheries Biological Laboratory, Honolulu, Hawaii. Manuscript received March 31, 1966. from the longline gear that caught them. One lousefish was attached to a living porcupinefish, Diodon hystrix Linnaeus, a notoriously slow swimmer. Lousefish have been reported from other slow swimmers including turtles (Men- zies, 1791:187), barracuda (Jordan and Ever- mann, 1898:2268; Schultz, 1943:256; Smith, 1950:341), and large groupers (Smith, 1950: 341 ) . It is difficult to accept Jordan and Ever- mann’s statement (1898:2268) that lousefish occur on spearfishes. Six specimens were free-swimming when col- lected, including the three largest ones which were attracted to a submerged light at night. One of the latter was kept alive in the ship’s baitwell for 10 days, during which time it was occasionally seen to attach to the tank walls for 1 or 2 seconds at a time. Later it was trans- ferred to a circular plastic swimming pool, 23 ft in diameter and 4 ft deep, which was supplied with running salt water. Here its attaching and other behavior were observed for two 1-hr periods each day for 25 days. The captive lousefish attached only under two conditions: when I entered the pool or when it had been fed to satiety. It was necessary to clean the pool’s windows and drain-strainers every few days ; when I entered the water to do this work, the fish abruptly ceased swimming and attached to the pool’s bottom or side. It usually remained attached for a few minutes and then darted toward me, especially if I was swimming. When I swam, it accompanied me | around the pool, staying about 1 ft from my bare feet, but making no effort to attach to me. ' The lousefish was fed once or twice a day. Feeding was avid to the point of greediness, jj the satiated fish having a bulging belly and cheeks, and usually being unable to dose its mouth. In this state its swimming movements t appeared to be hampered. It sank to the pool’s : bottom and did one of three things: either it swam slowly back and forth with a great deal I 260 Biology of the Lousefish — Strasburg 261 TABLE 1 Capture Data for Phtheirichthys lineatus DATE LATITUDE LONGITUDE SEX* STANDARD LENGTH (mm) HOST OR HABIT 2/ 1/57 13°44'S 110°02' W p 55.7 dip-netted beneath night-light 10/ 5/62 19°12'N 156°05' W - 256 dip-netted beneath night-light 10/ 5/62 19°12'N 156°05' W M 273 dip-netted beneath night-light 9/29/51 21° 24' N 158° 15' W F 300 dip-netted beneath night-light 10/30/58 04° 58' S 149°52' W p 32.8 caught by British Columbia trawl 11/ 1/58 05°00' S 149°58' W p 48.4 caught by British Columbia trawl 8/28/56 1 1 ° 2 3' S 134° 32' W p 46.0 in T hunnus albacares stomach 7/18/50 03°00' S 171 °22' W p 53.0 attached to longline buoy 7/20/60 11°30'N 161°21'E p 53.2 attached to longline buoy 8/26/56 13°26' S 132°16' W p 54.0 attached to longline buoy 7/18/50 03°00'S 171 °22' W p 56.1 attached to longline buoy 2/ 7/63 17°57'N 149°39' W p 51.3 attached to longline bait (squid) 2/23/63 20°4l'N 150°06' W p 55.5 attached to longline bait (squid) 2/ 7/63 17°57'N 149°39' W p 58.9 attached to longline bait (squid) 7/31/63 21°24'N 149°51' W p 61.6 attached to longline bait (squid) 2/ 6/63 18°16'N 149°46' W p 92.4 attached to longline bait (squid) 8/27/56 1 2 ° 1 6' S 133°20' W p 44.0 attached to longline bait (fish) 7/24/63 14°22'N 149°58' W - 81.3 attached to longline bait (fish) 7/10/63 14°24'N 150°11' W - 112 attached to longline bait (fish) 5/ 5/62 02°09'N 157°13' W p 48.9 attached to dead T hunnus albacares 2/ 6/63 18°16'N 149°46' W F 126 attached to dead Coryphaena equiselis 8/29/62 Kahana Bay, Oahu p 59.6 attached to living Diodon loystrix 7/18/50 03°00'S 171 °22' W p 48.6 unknown 3/15/59 14°49'N 150°12' W p 60.2 unknown * (?) = sex could not be determined; ( — ) = specimen not dissected. of lateral wriggling, or it rested on the bottom with its belly down, or it attached to the pool’s bottom or side. Each of these activities lasted as long as 15-20 minutes and appeared to represent a post-gorging rest. Sometimes the stuffed fish would regurgitate part of its meal and revert to swimming. Although bottom- resting was the most common post-feeding act, it is unlikely that this could happen in nature because substrates suitable for resting would not be available. If satiation resulted in such a tem- porary weight gain or such a loss of swimming mobility that a resting place was necessary, it is probable that a wild fish would never venture far from its attachment site. It is also probable that full meals are rare in the wild. The lousefish made no use of its fins when resting on the bottom or attached. The large caudal was collapsed about its long central rays, the dorsal and anal were folded against the body, the pelvics were depressed, and the pectorals were semiflexed. SWIMMING Except for the situations noted above, the captive lousefish swam constantly in either of two patterns. More commonly it swam back and forth across the pool, passing near the center each time. In a round trip, one passage was at the pool’s mid-depth (2 ft) and the return just above the bottom (4-ft depth). On the mid-depth trip the fish’s attitude was normal (disk up) ; on the bottom trip it was inverted (disk down). During the short vertical connecting trips, the disk faced the tank walls. The fish oriented to the walls or bottom when 6—12 inches from them. Less commonly it cir- cuited the pool’s periphery on its side, with the disk facing the pool wall. Some differences in swimming method were noted within these patterns. When the fish was swimming with the disk up, the head was es- sentially horizontal, the body axis hung about 8° below the horizontal, and the pectorals were 262 PACIFIC SCIENCE, Vol. XXI, April 1967 well extended. When it swam inverted, the head was again horizontal and the trunk axis dipped about 8° below the horizontal, but the caudal fin touched the pool bottom and the pec- toral fins were folded. When it swam on its side along the wall, the pectorals were half folded but the caudal did not touch the wall. The in- verted fish kept its head about 1^> inches above the bottom, but when swimming on its side 2 or 3 inches separated the head and the wall. In most swimming the caudal fin was well expanded to a rhomboidal or oval shape, the pelvics were folded, and the dorsal and anal fins were only partly erect. The fish’s sinuousity was impressive, as it could execute 180° turns In a circle about 6 inches in diameter. The pec- torals and sometimes the pelvics were erected when the fish changed course or its plane of swimming, and also when it had just been fed to satiation. In addition, the gorged fish occa- sionally swam at an angle of 15° to 35° from the horizontal. Swimming speeds were calculated for each of the three swimming attitudes and for the hun- gry and fed fish. Data were obtained by timing the fish as it swam between reference points on the pool’s sides. Speed data are given in Table 2. In contrast to normal and inverted swimming, side-swimming when satiated was observed only once. The values given are the extremes of a number of readings. The loss of speed after feeding was marked but of short duration. Un- fortunately, the only data showing this change relate to tail-beat frequency, not to speed. Tail beats were counted at 66/minute immediately after feeding, but suddenly increased to 78 and 84 beats/minute 11 minutes later. TABLE 2 Swimming Speed for the Lousefish RANGE IN SWIMMING STATE OF FISH SWIMMING ATTITUDE SPEED (ft/sec) Not fed for 7-12 hr normal inverted on side 1.61-2.05 1.88-2.04 1.41-2.21 Fed to satiety normal inverted on side 0.80-1.61 0.87-1.32 1.15-1.39 ° MUlNUrltl ★ SATIATED o o ★ 3 * o * ★ * °0 25 50 75 100 125 150 175 NUMBER OF TAILBE ATS/ MINUTE Fig. 1. Speed versus tail-beat frequency in the lousefish. Figure 1 shows the relation between speed and tail-beat frequency. Speed and tail-beat data were obtained simultaneously as the fish tran- sited the pool. The number of beats per minute ranged from 48 to 144. No speed estimate was obtained for the 48-beat value, which occurred 2 minutes after gorging and 1 minute before the fish sat on the bottom. RESPIRATION While the lousefish was swimming, its mouth was always open and no opercular movements were detectable. In this respect it resembled Remora remora (Linnaeus), which ceased oper- cular pumping under conditions of optimum, if artificial, water flow over the gills (Strasburg, 1957). The speed range (or the flow rate) ob- served for the hungry lousefish (1.41-2.21 ft/sec), however, was more than twice the range (0.75-0.88 ft/sec) judged optimal for R. remora (Strasburg, 1957). When gorged, the lousefish usually sat on the bottom or attached. Its mouth remained open and opercular respiratory movements were con- spicuous. The well-developed tongue was alter- nately pressed to the mouth’s roof and returned to its bed, in a rhythm keyed to the opercular movements. Sometimes the fish oriented to face the weak current at the pool’s inlet. The num- ber of respirations per minute ranged from 84 to 120 — a much lower range than for R. remora , which had from 203 to 244 respirations per Biology of the Lousefish — Strasburg minute when attached but unfed (Strasburg, 1957). Stuffing the digestive tract with food possibly inhibits the respiratory rate on some mechanical basis. Once when the lousefish fed until it bulged, and then attached to the bottom, it had a rate of 84 respirations per minute. Six minutes later, when it regurgitated four cubes of shark flesh, the rate increased to 114 respira- tions per minute. FOOD AND FEEDING The results of stomach analyses of four louse- fish were presented in an earlier paper (Stras- burg, 1959:246). It has since been possible to examine 16 more stomachs and obtain better definition of the scope of the diet, and also to observe the feeding behavior of a living speci- men. Of the 20 lousefish stomachs examined 2 were empty. The balance contained mostly planktonic crustaceans, especially small and larval forms. The crustaceans and the number of stomachs in which they occurred were as follows: hyperiid amphipods (1), unidentified amphipods (6), crab larvae (2), unidentified decapod larvae (1), stomatopod larvae (2), Euphausiacea ( 1 ) , Mysidacea ( 1 ) , Ostracoda (3), unidentified crustacean fragments (3), and the following copepods: Candacia pachydactyla (5), C. aethiopica (2), Candacia sp. (1), Scolecithrix danae (1), Euchaeta sp. (3), un- identified calanoids (3), Sapphirina sp. (1), Oncaea sp. (1), and unidentified cyclopoids (1). Also, 2 stomachs contained fish flesh and bones, and 2 had small flakes of rusty iron. None of the food species is parasitic, and it is thus unlikely that the lousefish is a cleaner, in contrast to most other echeneids (Szidat and Nani, 1951:413; Maul, 1956:14; Strasburg, 1959:246). Instead, it appears to feed rather selectively on planktonic animals, especially on such conspicuous creatures as the black-and- white copepods Candacia pachydactyla and C. aethiopica , the iridescent blue copepod Sap- phtrina, and large stomatopod larvae. Rust flakes from the fishing vessel probably were ingested because they were relatively conspic- uous. Mysids had been eaten only by the ! 59.6-mm lousefish attached to a living Diodon hystrix caught in Kahana Bay, Oahu. Mysids 263 abound in this shallow bay, and the fish had gorged on 41 of them. The captive lousefish was fed small cubes of bread, shrimp, or shark flesh once or twice a day. Presentation of food was preceded by a cue (described below) and continued until the offerings were ignored. The fish reacted to food particles from a distance of about 3 ft ; the water was very clear. It refused particles lying on the bottom, and preferred to feed on sinking food pieces near the surface. Pieces less than about 3 mm in greatest dimension were ignored in favor of larger ones (up to about 1 cm in greatest dimension). The 3-millimeter pieces were "inhaled” without noticeable jaw move- ments, whereas those in the 7- to 10-mm range were gulped with conspicuous jaw action. As noted earlier, the fish was a greedy feeder and ate until its belly and cheeks bulged. A full meal required 3 or 4 minutes to consume. When fed once a day for 3 successive days, the volume of shark flesh eaten per meal was 5.1, 5.0, and 2.6 cc. On the last day the fish also ate bread crumbs whose volume had not been measured. The lousefish was accidentally placed in the wrong pool at first, and for a time I planned to recapture it by dip net for transfer elsewhere. It was necessary to condition the fish to the dip net, for its earlier capture had made it wary. A dip net was placed in the pool before each feeding, and food was offered over the net so that the fish had to swim above the meshes to feed. On the third day of this pro- cedure (the fifth feeding) the fish swam slowly but directly to the net when it was placed in the pool. At the next feeding it dashed immediately to the net, and continued to do so whenever the net was shown. Because it proved unneces- sary to transfer the fish, this conditioning went for nought. The fish was subsequently trained to come to the feeding area when the water’s surface was slapped with the hand. COLOR AND COLOR CHANGE A description of the life colors of the louse- fish could not be found in the literature and is therefore presented here. In sunlight the dorsal surfaces of the head and body are navy blue and the ventral surfaces are white. In shade the 264 PACIFIC SCIENCE, Vol. XXI, April 1967 navy blue often appears as velvety black. Separating these contrasting colors are three lengthwise stripes along the sides. The upper stripe is narrow and light blue, the middle is broad and black, and the lower is narrow and silvery white. The light-blue stripe commences on the upper snout (where it meets its fellow), touches the upper edge of the eye, and con- tinues thence to the caudal base. The black stripe runs from the snout through the eye to the caudal base, and expands on the caudal fin so that the fin is black except for narrow white dorsal and ventral margins. The silvery- white stripe runs from the rictus to the lower edge of the eye and thence along the lower mid-side to the caudal base. The dorsal and anal fins are black basally and margined with white anteriorly. The pectorals and pelvics are plain blackish. This color was characteristic of the captive lousefish during its first 7 days in the swimming pool. Thenceforth its ventral surface gradually darkened, changing from white to gray-spotted (eighth day) to uniform gray (ninth day) to gray-black (tenth day). Subsequently the color of its undersides fluctuated between gray, deep slate blue, and black, there being no apparent relation between color and incident light or fish activity. Frequently, however, the color of the undersides was lighter just after feeding. Sometimes the back and the base of the dorsal and anal fins were the same color as the under- sides. The three lateral stripes were unvaryingly the same. By the twenty-first day the fish’s dorsal and ventral surfaces were both jet black, and no further changes were observed. When the fish was preserved after 25 days, it retained its melanistic coloration in alcohol. During the lousefish’s 2 5 -day confinement, the color of the pool walls changed markedly as the diatom Melosira proliferated from es- sentially nothing to a growth 2 inches thick. The fish’s gradual color change paralleled this growth. The darkening may have been caused by a general increase in the amount of melanin. On the other hand, a faculty for rapid color change has been reported for Echeneis naucrates (Beebe and Tee-Van, 1933:222; Nichols in LaGorce, 1939:163; Sanborn, 1932:89; Town- send, 1927:171) and for Remora remora (Maul, 1956:50, 66), and thus it is possible that the observed changes were highly trans- itory. Another explanation for this color change derives from observations on echeneids which regularly attach. These fish have been stated either to lack countershading (Cott, 1940:43; Fincher, 1948:283) or to have reversed coun- tershading (Norman and Fraser, 1949:176). A lack of countershading supposedly results from a failure to maintain constant orientation to a light source; a reversal of countershading arises from the fact that the echeneid’s belly is usually more brightly illuminated than its back. The first explanation possibly fits the color change described above for the lousefish. When captured it had normal countershading, but began to lose it after a period in which swimming was often inverted or on the side. When free in the pelagic environment the lousefish presumably swims in a normal attitude. REFERENCES American Fisheries Society, i960. A List of Common and Scientific Names of Fishes from the United States and Canada. Am. Fisheries Soc., Ann Arbor, Spec. Publ. 2, 102 pp. Beebe, W., and J. Tee-Van. 1933. Field Book of the Shore Fishes of Bermuda. G. P. Putnam’s Sons, New York and London, xiv + 337 pp., 25 pis., 112 figs. Cott, H. B. 1940. Adaptive Coloration in Animals. Oxford Univ. Press, New York, xxxii -j- 508 pp., 49 pis., 84 figs. Jordan, D. S. 1907. Fishes. Henry Holt and Co., New York, xv -|- 789 pp., 18 pis., 673 figs. and B. W. Evermann. 1898. The Fishes of North and Middle America. Bull. U. S. Natl. Mus. (47), Part 3, pp. xxiv -f- 2183-3136. Lagorce, J. O. (Ed.) 1939. The Book of Fishes. Natl. Geogr. Soc., Washington. 367 pp., illus. Maul, G. E. 1956. Monografia dos peixes do Museu Municipal do Funchal. Ordem Disco- cephali. Bob Mus. Funchal (9)23:5-75, 5 figs., 21 tables. Menzies, A. 1791. Descriptions of three new Biology of the Lousefish — Strasburg animals found in the Pacific Ocean. Trans. Linn. Soc. London 1:187-188, 1 pi. Norman, J. R., and F. C. Fraser. 1949. Field Book of Giant Fishes. G. P. Putnam’s Sons, New York, xxii -|- 376 pp., 8 pis., 97 figs. Pincher, C. 1948. A Study of Fish. Duell, Sloan and Pearce, New York. 343 pp., 300 figs., 16 tables. Sanborn, E. R. 1932. Commensalism. Bull. N. Y. Zool. Soc. 35(3) :74, 86-92, 8 figs. Schultz, L. P. 1943. Fishes of the Phoenix and Samoan Islands. Bull. U. S. Natl. Mus. (180), x + 316 pp., 9 pis., 27 figs. Smith, J. L. B. 1950. The Sea Fishes of South- ern Africa. Central News Agency, Capetown, xvi -f- 550 pp., 103 pis., text-figs. 265 Strasburg, D. W. 1957. Notes on the respira- tion of small Remora remora. Copeia 1957 (1) : 5 8-60, 1 table. 1959. Notes on the diet and correlating structures of some central Pacific echeneid fishes. Copeia 1959 (3) : 244-248, 1 fig., 2 tables. Szidat, L., and A. Nani. 1951. Las remoras del Atlantico austral con un estudio de su nutricion natural y de sus parasitos. Rev. Inst. Invest. Mus. Argent. Cienc. Nat. 2(6): 385- 417, 14 figs. Townsend, C. H. 1927. Records of color changes among fishes. I. Three color changes of the shark-sucker ( Echeneis nau crates) . Bull. N. Y. Zool. Soc. 30(6) :170-171, 1 pi. The Planktonic Shrimp, Lucifer chacei sp. nov., (Sergestidae: Luciferinae), the Pacific Twin of the Atlantic Lucifer faxoni Thomas E. Bowman1 ABSTRACT: Lucifer chacei sp. nov., closely related to the Atlantic L. faxoni and identified as the latter species by previous authors, is described and compared with L. faxoni. It is widely distributed in the tropical Pacific, and like L. faxoni inhabits coastal waters. The planktonic shrimp genus Lucifer (fam- ily Sergestidae) contains six currently recog- nized species, of which all have been reported from the Indo-Pacific, but only two, L. typus Milne Edwards and L. faxoni Borradaile, are known to occur in the Atlantic. An unpublished study which I recently completed on the dis- tribution of the two Atlantic species off the southeastern coast of the United States shows that L. faxoni is essentially a coastal species. Since coastal plankters more often than not are restricted to one ocean, or even to a single coast of one ocean, it seemed advisable to re- examine the evidence for the occurrence in the Indo-Pacific of L. faxoni, the type locality of which is in the Atlantic, off Chesapeake Bay (restricted by Holthuis, 1959). For this study Atlantic specimens of L. faxoni from Bloody Bay, Tobago, West Indies, were compared with Pacific specimens of Lucifer from Eniwetok, Marshall Islands, and from several of the Society Islands. Both Atlantic and Pacific specimens key out to L. faxoni in Han- sen (1919), but they are specifically distinct, and a new species is established herein for the Pacific form. Surprisingly, the common short-eyestalked Atlantic species, L. faxoni , has never been de- scribed and illustrated adequately (Hansen’s drawings of L. faxoni in his 1919 monograph are actually of the new Pacific species) ; hence illustrations of taxonomically important features of L. faxoni are given herein. 1 Division of Crustacea, Smithsonian Institution, Washington, D.C. 20560. Manuscript received March 7, 1966. Lucifer chacei , new species Figs. 1-4 Lucifer faxoni Borradaile. — Hansen, 1919: 61-63, pi. 5, figs. 3a-3i. — Edmondson, 1925:5 .—Hiatt, 1 947 : 241-242 .— Chace, 1955:4. Leucifer reynaudi H. Milne Edwards. — Ed- mondson, 1923:35. MATERIAL EXAMINED: MARSHALL ISLANDS: Eniwetok Atoll; lagoon, 4.8 km west of Parry I., net tow at depth of approximately 3 m, 26- 27 July I960, 2435-0800 hours, by Woods Hole Oceanographic Institution, $ holotype (usnm 113327), $ allotype (usnm 113328) and 55 paratypes. Rongelap Atoll; 1 km off Yugui I., depth of water 24 m, dip net under night light, 30 July 1946, by Earl S. Herald, 3 $ , in poor condition, reported as L. faxoni by Chace (1955). tuamotu islands: Tikahau Atoll ; lagoon, south of deep water pass, net tow at depth of approximately 4 m, 12 April 1957, 2015-2030 hours, station 19 of Smithsonian Bredin 1957 Expedition, 43 specimens, society islands (Smithsonian-Bredin 1957 Expedition stations): Tahiti: Papeete Harbor, dip net under night light, 4 May, 2100-2130 hours, station 99, 40+ specimens. Moorea: Opunohu Bay, east side, net tows, 9 May, 1530 hours, station 114, 43 juveniles. Bora Bora: East of Farepiti Point, dip net under night light, 24 April, 2000-2030 hours, station 52, 50+ spec- imens; off Teraia Point, depth of water 30 m, net tows, 25 April, 0900-1100 hours, station 55, 24 specimens; west of north end of Toopua I., depth of water 13 m, dip net under light, 25 April, 2030-2100 hours, station 63, 30+ specimens. 266 Lucifer chacei n. sp. — Bowman 267 Fig. 1. Lucifer chacei: a, female, lateral; b, anterior end of male head, lateral; c, anterior end of female head, lateral; d, anterior end of male head, dorsal; e, male posterior head and anterior pereon, lateral; f, female uropod and telson, lateral; g, male telson, lateral; h, female telson, lateral; i, male telson, dorsal; j, apex of male telson, dorsal. 268 PACIFIC SCIENCE, VoL XXI, April 1967 Fig. 2. a-b, Lucifer chacei: a, male 6th abdominal somite and telson, lateral; b, endopod of left male 2nd pleopod, anterior, c-k, Lucifer faxoni : c, anterior end of male head, dorsal; d, same, lateral; e, anterior end of female head, lateral; f, male 6th abdominal somite and telson, lateral; g, endopod of left male 2nd pleo- pod, anterior; h, male telson, lateral; i, immature male telson, lateral; j, apex of male telson, dorsal; k, apex of exopod of male uropod, dorsal. Lucifer chacei n. sp. — Bowman 269 Fig. 3. a—c, Lucifer chacei : a, petasma in situ, medial; b, petasma sheath with its processus ventralis, and proximal lobe at base of sheath ; c, petasma sheath of another male, with processus ventralis displaced. d-e, Lucifer faxoni : d, petasma in situ, medial; e, petasma sheath with processus ventralis displaced. diagnosis: A Pacific species belonging to Hansen’s (1919) "Group B” (species with short eyestalks), closely resembling the western At- lantic L. faxoni in having the apex of the petasma acute, without transverse lines or pro- truding plates or scabrousness, and in having a slender processus ventralis with acute apex. The diagnostic features of L. chacei in com- parison with L. faxoni are shown in Table 1. The new species is named for my colleague, Fenner A. Chace, Jr., in recognition of his many contributions to decapod crustacean taxonomy. remarks: Of the remaining species of Luci- fer, only L. hanseni Nobili agrees with L. faxoni and L. chacei in having a slender, acutely ending processus ventralis. However, L. hanseni can be identified immediately by the uniquely shaped uropodal exopod, in which the outer tooth is located considerably proximad of the distal margin. Previous workers reporting L. faxoni from the Pacific naturally based their identifications on Hansen’s (1919) monograph. None had reason to suspect that Hansen’s description and illustrations dealt with an undescribed species. Although Hansen stated that the Copenhagen Museum possessed several samples of Atlantic L. faxoni, he could not have compared them carefully with his Siboga material. I cannot be- lieve that the distinguished Danish carcinologist would have overlooked the differences separat- ing the two species. L. chacei is so similar to L. faxoni that it is possible that their genetic divergence is at the subspecific rather than the specific level. They probably have been derived from a common ancestor. The actual level of divergence from this ancestor cannot be ascertained from avail- able collections, and, since the two forms are completely isolated from one another, there is no possibility of gene flow between them at present. If plans for a sea level canal connecting the Caribbean Sea and the Pacific Ocean are carried out, the picture may change drastically. 270 PACIFIC SCIENCE, Vol. XXI, April 1967 Fig. 4. Lucifer chacei, female, thelycum: a, ventral, b, lateral. A, P, Anterior and posterior margins of ventral process; P3, base of 3rd pereopod; SP, spermatophore ; ATR, atrium; SEM. REC., seminal receptacle. distribution of L. chacei : east indies: many localities (Hansen, 1919). Hawaiian is- lands: Oahu, Molokai (Hiatt, 1947). line is- lands: Fanning Island (Edmondson, 1923). tuamotu islands: Tikahau Atoll, society is- lands: Tahiti (Papeete Harbor) ; Moorea (Paopao Bay) ; Bora Bora (south of Farepiti Pt., olf Tereia Pt., west of Toopua I.), mar- shall islands: Rongelap Atoll (off Yugui I.); Eniwetok Atoll (west of Parry I.). It is probably widespread in the tropical Pacific, and, like its Atlantic counterpart, L. faxoni , is prob- ably a coastal species. Along the east coast of Australia it is replaced by L. penicillifer Han- sen (Gordon, 1956). reproduction: I have examined the female TABLE 1 Diagnostic Features of L. chacei Compared with L. faxoni Lucifer faxoni borradaile Lucifer chacei, new species 1. Last segment of peduncle of ant. 2 in $ reaches 1. beyond eye and nearly to distal margin of 1st segment of ant. 1 peduncle, in $ reaches beyond middle of cornea and to distal third of 1st seg- ment of ant. 1 peduncle. 2. Rostrum reaches proximal border of statocyst. 3. Ventral cushion of $ telson much broader than its posterior height. 4. Sheath of petasma curved. Last segment of peduncle of ant. 2 in S reaches middle of cornea and distal third of 1st segment of ant. 1 peduncle, in $ reaches proximal margin of cornea and to distal third of 1st segment of ant. 1 peduncle. 2. Rostrum reaches almost to distal end of statocyst. 1 3. Ventral cushion of $ telson about as broad as its posterior height. 4. Sheath of petasma straight. Lucifer chacei n. sp. — Bowman 271 external reproductive system in specimens cleared in lactic acid. The thelycum consists of a conical median ventral process between the bases of the third pereopods, behind which is a longitudinal slit leading into the atrium, a sclerotized pouch formed by a median depres- sion of the sternum. When present, the sper- matophore neck is inserted into the atrium through the slit. The large paired seminal re- ceptacles lie posterior and lateral to the atrium. In my cleared material it was evident that the seminal receptacles connect by ducts to the atrium, but the nature of the ducts was not clear. I could not detect the female genital openings, which Burkenroad (1934) and Gor- don (1956) found on the coxae of the third pereopods. The entire system is rather complex and, as Gordon pointed out, is in need of critical study using histological techniques. Several females from Bora Bora had clusters of eggs attached by short stalks to the ischia of the third pereopods. The eggs were probably about ready to hatch, for well-developed nauplii could be seen through the egg membranes. Be- cause of the ease with which eggs become de- tached from the third pereopods in preserved specimens, Gordon (1956) doubts that they stay attached until the nauplii emerge, a period of more than 36 hours according to Brooks (1882). The Bora Bora specimens prove that at least some of the eggs remain attached until eclosion. REFERENCES Brooks, W. K. 1882. Lucifer , a study in mor- phology. Phil. Trans. Roy. Soc. London 173: 57-137, pis. 1-11. Burkenroad, Martin D. 1934. The Penae- idea of Louisiana with a discussion of their world relationships. Bull. Am. Mus. Nat. Hist. 68 (art. 2) :6l-l43. Chace, Fenner A., Jr. 1955. Notes on shrimps from the Marshall Islands. Proc. U. S. Natl. Mus. 105(3349) :l-22. Edmondson, Charles H. 1923. Crustacea from Palmrya and Fanning Islands. B. P. Bishop Mus. Bull. 5:1-43, pis. 1-2. 1925. Crustacea, pp. 3-62. In: C. H. Edmondson, W. K. Fisher, H. L. Clark, A. L. Treadwell, and J. A. Cushman, Marine Zoology of Tropical Central Pacific. B. P. Bishop Mus. Bull. 27 :i — ii, 3-148, pis. 1—11. Gordon, Isabella. 1956. The Sergestidae of the Great Barrier Reef Expedition. Great Barrier Reef Exped. Sci. Rept. 6(5) :323- 333. Hansen, H. J. 1919. The Sergestidae of the Siboga Expedition. Siboga-Expeditie, Mono- graph 38, 65 pp., 5 pis. Hiatt, Robert W. 1947. Ghost prawns (sub- family Luciferinae) in Hawaii. Pacif. Sci. 1 (4) :24l-242. Holthuis, Lipke K. 1959. The Crustacea Decapoda of Suriname (Dutch Guiana). Zook Verh., Leiden, no. 44, 296 pp., 16 pis. Revision of the Genus Pandanus Stickman, Part 21 The Pandanus monticola Group in Queensland, Australia Harold St. John1 In Australia the section Acrostigma of the genus Pandanus occurs only in the rain forests of eastern Queensland. The first known species there was P. monticola F. Muell. It is here given a full description and illustrations, as is one new species segregated from it. Section Acrostigma Pandanus monticola F. Muell. (sect. Acro- stigma), Fragm. Phytog. Austral. 5:42, 1865; 7:63, 1870; 8:220, 1874; Mar- telli, Webbia 4(2) :pl. 30, fig. 21-27, 1914 Figs. 232-234 description (from St. John 26,237) : Shrubs 1-3 m tall, simple, erect or decumbent; stem 25-40 mm in diameter, yellowish, unarmed, bearing in a terminal plume the arching leaves 1.8-2.18 m long, 4 cm wide near the base, 4.2 cm wide at the middle, subcoriaceous, ligulate, gradually long tapering to a subulate tip which 10 cm down is 7 mm wide, the very base unarmed and amplexicaul but beginning 5 cm up the margins with spines 1.5-2 mm long, 1.5-5 mm apart, the lowest ascending but the others salient, subulate, stramineous; the nearby midrib with prickles 0. 5-0.8 mm long, 1.5-6 mm apart, the base thick like a boss, the apex reflexed and subulate-conic; at the mid- section the margins with prickles 0. 5-0.9 mm long, 2-5 mm apart, subulate, appressed as- cending; the nearby midrib unarmed; near the tip the margins and midrib below with the teeth 0.3-0. 5 mm long, 1-3 mm apart, subu- late-serrate, ascending, above the two pleats with teeth 0. 5-0.7 mm long, similar but broader-based and occasionally double; pistil- late inflorescence terminal, erect, 2-headed, the upper syncarp larger, but with a small, secon- 1 B. P. Bishop Museum, Honolulu, Hawaii. Manu- script received December 3, 1962. dary, subterminal one below; peduncle 50 cm long, 1 cm in diameter, trigonous, leafy bracted, the middle bracts 65 cm long, 3-4 cm wide; the larger syncarp 9-5 cm long, 7.5 cm in diameter, ovoid, bearing very numerous drupes, these 24-28 mm long, 2-3 mm wide, 1.5-2. 5 mm thick, yellow, narrowly lance- subulate, 4-6-angled, upper % free ; p ileus 13-14 mm long, subulate, sharply angled on drying, the surface smooth, the tip curved to- ward the apex of the syncarp; style 6-8 mm long, subulate, cartilaginous, yellowish, smooth, arcuate; stigma distal, 5-6 mm long, linear, brown ; endocarp in lower ninth, 4-6 mm long, cartilaginous, yellowish, the walls 0.1 mm thick; seeds 4 mm long, broad oblong-ellip- soid; apical mesocarp with a central core of fibers and with fleshy sides ; basal mesocarp very small, fibrous and fleshy. holotype: Australia, Rockingham’s Bay proximis (staminate only), (mel), specimen examined! This was published as "montibus sinu Rockingham’s Bay proximis.” Isotype (k) ! In the Paris herbarium there is an original collection, with part of a leaf and half of a syncarp, that is a far better specimen than the holotype preserved in Melbourne. It was dis- tributed by von Mueller in 1874, and the data is in his handwriting: "Rockingham’s Bay, on mountains, most probably this belongs to P. monticola , of which I had no flowers, when I described it, fruit small, bracts white.” (p) ! specimens examined: Queensland, without locality, L. J. Brass 2,128 (a) ; in valle Dalrympli, 22 Oct. 1864, /. Dallachy (mel) ; Russell River, F. von Mueller (mel) ; Mt. Bellenden-Ker, 1895, Mrs. Gribble (mel) ; Cairns, Fresh Water Creek, common, moist forest, 500 ft alt, Jan. 29, 1958, H. St. John 26,237 (bish). 272 Page 385: Revision of Pandanus , 21. Queensland — St. John 273 9CM. . Fig. 232. Pandanus monticola F. Muell., from holotype. a, Staminate spike and bract, X 1; b, stamens and axis, X 10; c, leaf, showing upper side of apex, X 1; d, leaf apex, lower side, X 1. 0 icm.. 1 , . ■ , _ t — i Fig. 233. Pandanus monticola F. Muell., from Queensland, Russell R., 1886, F. von Mueller (mel). a, Syncarp and peduncle, X 1 ; b, part of peduncle, X 4 ; c, drupe, lateral view, X 1 \ d, drupe, longitudinal median section, X 1 ; e, drupe, apical view, XI;/, drupe, lateral view, X 4 ; g, drupe, longitudinal median section, X 4; h, drupe, apical view, X 4; style and stigma, X 1 0; /, leaf base, lower side, X 1 ; k> lea^ middle, lower side, X 1 ; h venation of lower leaf surface, X 4. 274 Fig. 234. Pandanus monticola F. Muell., a-j from St. John 26,237 ; k, from "valle Dalrympli,” /. Dal- lachy (mel) . a, Drupe, lateral view, X 1; b, drupe, longitudinal median section, X 1 ; c, drupe, lateral view, X 1; ^ drupe, longitudinal median section, X 4 ; e, drupe, apical view, X 4 ; style and stigma, X 4; g, leaf base, lower side, X 1 \ h, leaf middle, lower side, X 1 ; h leaf apex, lower side, X 1; j, leaf apex, upper side, X 1; ^ drupe, lateral view, X 4. 275 Fig. 235. Pandanus pluvisilvaticus St. John, from holotype. a, Drupe, lateral view, XI \b, drupe, longi- tudinal median section, y l\ c, drupe, lateral view, X 4 ; d, drupe, longitudinal median section, X 4 ; e, drupe, apical view, X 4; f, style and stigma, X 10; g, leaf base, lower side, X 1; h, leaf middle, lower side, X 1; i, leaf apex, lower side, X 1 j j, leaf apex, upper side, X 1- 276 Fig. 236. Pandanus pluvisilvaticus St. John, from St. John 26,234. a, Staminate inflorescence, X 1/3; b, staminate spike, X 1; c, stamens and axis, X 10; d, leaf base, lower side, X 1; leaf middle, lower side, XI;/, leaf apex, lower side, X 1; g, leaf apex, upper side, X 1. 277 278 Page 391: Revision of Pandanus , 21. Queensland — St. John 279 discussion: P. monticola F. Muell. belongs in the section Aero stigma. Its author, Baron von Mueller, published upon it three times. In 1865 he published the binomial for it as a doubtful species, and gave a casual reference to fruit characters. The type locality was Rockingham’s Bay, but the authenticating speci- men preserved in the Melbourne herbarium, collected in 1864 by J. Dallachy, consists of one leaf and parts of a staminate inflorescence. This is taken to be the holotype. His second publication, in 1870, was based upon study of a collection from "valle Dal- rympli,” also by J. Dallachy. Here, von Mueller positively adopted the species and gave a good description, including characters of the scarlet fruit. In 1874 he published on it a third time, supplementing the herbage characters, and de- scribing the staminate flowers. Pandanus pluvisilvaticus sp. nov. (sect. Aero stigma) Figs. 235-237 diagnosis holotypi: Frutex 4 m altus, cauli- bus plerumque simplicibus 3 cm diametro erectis inermibus, radicibus fulturosis nullis, foliis 1.45- 1.66 m longis proxima basem 4.1 cm latis in medio 3. 8-3. 9 cm latis ligulatis subcoriaceis supra viridibus infra pallide viridibus sensim in apice subulato diminuentibus eo in puncto 10 cm ex apice 4.5 mm lato in sectione M-formatis, basi inermi amplexicauli sed ex 2.5-3 cm marginibus cum aculeis 1.5-2 mm longis 2-5 mm separatis subulatis adscendenti- bus stramineis, midnervo proximali inermi, in sectione mediali marginibus cum aculeis 0.5-1 mm longis 3-7 mm separatis subulatis adpresse adscendentibus, midnervo infra proximali cum aculeis 0.3-0. 5 mm longis 5-10 mm separatis subulatis adscendentibus, proxima apicem marginibus et midnervo infra subulatoserratis dentibus 0.4-0. 6 mm longis eis marginalium 1-2 mm separatis illis midnervi 2-5 mm separatis, supra plicis lateralibus cum serrulis 0. 5-0.8 mm longis 2-5 mm separatis, in- florescentia foeminea terminali, pedunculo 44 cm longo 7-8 mm diametro acute trigono cum bracteis foliosis paucis superiore 39 cm longa 2 cm lata, syncarpio ex 6-7 bracteis albis sub- petaloideis sustento, eis medialibus superisque 7-10 cm longis 3 cm latis lanceolatis minute aciculari-ciliatis, syncarpio 6 cm longo 5.5 cm diametro latiter obovoideo solitario terminali erecto cum drupis multis adpressis, drupis 20- 22 mm longis 4 mm latis 3 mm crassis rubris pilei 6.5-12 mm longo 5-6-anguloso libero corpore laevi ellipsoideo sed argute anguloso, parte % supera libera, stylo 3. 5-4. 5 mm longo subulato subcurvato plerumque distaliter curvato, stigmate 3-4 mm longo lineari distali brunneo papilloso, endocarpio in parte y4 infera 7 mm longo cartilagineo luteo lateribus 0.2 mm crassis in apice in lateribus cavernae extento, semine 5 mm longo 3-4 mm diametro latiter ellipsoideo sed in apice concavo, mesocarpio apicali in centro cum fibris fortibus et cum lateribus carnosis, mesocarpio basali sparso fibroso et carnoso. description of all specimens examined: Shrub 4 m tall; stems mostly simple, pale brown, 3 cm in diameter, erect, unarmed; prop roots none; leaves 1.45-2.33 m long, 3. 8-5. 4 cm wide near the base, 3.3-5. 1 cm wide at the middle, ligulate, subcoriaceous, green above, pale green below, gradually tapering to the subulate apex, this 10 cm down 4.5 mm wide, in section M -shaped, the base unarmed, am- plexicaul but beginning at 2.5-3 cm up the margins with prickles 1.5-2 mm long, 1-5 mm apart, subulate, ascending, stramineous, the nearby midrib unarmed, or with reflexed prickles 1-1.5 mm long, 2-5 mm apart; at midsection the margins with prickles 0.5-1 mm long, 3-7 mm apart, subulate, appressed as- cending ; the nearby midrib below with prickles 0.3-1 mm long, 3-10 mm apart, subulate, as- cending; near the tip the margins and midrib below subulate-serrulate, the teeth 0.4-0. 6 mm long, those of the margins 1-2 mm apart but those of the midrib 2-5 mm apart; above the lateral pleats with ascending serrulations 0.5— 0.8 mm long, 2-5 mm apart; pistillate inflores- Fig. 237. Pandanus pluvisilvaticus St. John, a-c, From holotype: a ( upper left), habit, and figure of Dr. S. T. Blake of Brisbane; b ( upper right), inflorescence in anthesis, showing the white bracts; c ( lower left), leafy branch with mature syncarp ; d (lower right), from St. John 26,264, leafy branch with mature syncarp. 280 PACIFIC SCIENCE, Vol. XXI, April 1967 cence terminal; peduncle 35-65 cm long, 7-8 mm in diameter, sharply trigonous, with a few leafy bracts, the uppermost one 39 cm long, 2 cm wide; syncarp subtended by 6-7 pure white, subpetaloid bracts, the middle and upper ones 7-10 cm long, 3 cm wide, lanceolate, finely acicular ciliate; syncarp 6-11 cm long, 5.5- 7 cm in diameter, broadly obovoid, single, terminal, erect, bearing numerous, crowded drupes, these 20-28 mm long, 4 mm in diam- eter, 3 mm thick, red to orange; pileus 5-6- angled, upper l/2-% of drupe free, pileus 6.5- 12 mm long, the surface smooth, the body ellipsoid but with sharp ridges and deep rounded valleys; style 3. 5-4. 5 mm long, subu- late, slightly curved, usually away from apex of syncarp; stigma 3-4 mm long, distal, linear, brown, papillose; endocarp in lower %, and 7-11 mm long, cartilaginous, yellowish, the walls 0.2 mm thick, and at its summit enclosing all but the top of a fat discoid cavity; seed 5-6 mm long, 3-5 mm in diameter, broad ellipsoid except for the concave apex; apical mesocarp with strong central fibers and soft pith forming the side tissues; basal mesocarp sparse and this fibrous and fleshy. Staminate plants 10 m tall, 20 mm in diam- eter; prop roots none; leaves 2.4 m long, 5 cm wide near the base but at the middle 5.2 cm wide, the base amplexicaul, entire, begin- ning 5-6 cm up the margins with spines 2-3 mm long, 1-6 mm apart, stout serrae with subulate tips, single or a few doubled or trebled, yellowish; the midrib below beginning at 12.5 cm up with stout retrorse serrae, 1-1.5 mm long, 2-8 mm apart; at the midsection the margins with teeth 1—1.5 mm long, 3-6 mm apart, subulate, appressed ascending; the mid- rib below sharp, raised, with similar teeth, but heavier based; near the apex the margins and the midrib below and the two secondary pleats above with serrae 0.8-1 mm long, 1-2 mm apart ; the blade gradually narrowed to the subulate trigonous apex, the very tip lost, but about 10 cm down the tip 10 mm wide; staminate inflorescence 65 cm long, sparsely leafy bracted, the middle bract of the peduncle 110 cm long, 6 cm wide, foliaceous, the midrib below and the margins spiny to serrate, the apex subulate, the peduncle 40 cm long, the main raceme with 8—10 lateral racemes, each subtended by a white, semipetaloid bract, the lower ones 40 cm long, 3 cm wide, the apical half green and foliaceous ; lateral racemes 6-10 cm long, including the 5-7 -mm stipe, 15 mm in diameter, densely flowered; stamens distinct; free filament 0.7-1. 6 mm long; anthers 6-7 mm long, linear, yellow, bearing an apical prolongation of the connective 1-2 mm long, subulate. holotypus: Australia, North Queensland, Kuranda, Black Mt. Road, rain forest with Acacia, Calamus, and Gahnia, 1,000 ft alt, Feb. 4, 1958, H. St. John 26,233 (bish). specimens examined: Australia, North Queensland, 11 miles N. of Mossman, rain forest with Calamus, Myristica, and Hibiscus tiliaceus, 20 ft alt, Feb. 6, 1958, St. John 26,264 (bish); 5 miles N. of Gadgarra Forest Station, 5 miles E. of Yungabarra, rain forest, 2,130 ft alt, Feb. 11, 1958, St. John 26,277 (drupes sterile) (bish) ; Bloomfield R., 1883, Barnard (mel) ; Russell R., 1886, F. von Mueller (mel) ; near Mulgrave R., [ F . von Mueller ] 144 (mel) ; Daintree R., 1882, Pentzke (mel) ; 11 miles N. of Mossman, rain forest with Calamus, Myristica, Hibiscus tilia- ceus, 20 ft alt, Feb. 5, 1958, St. John 26,263 (bish). discussion: P. pluvisilvaticus is a member of the section Acrostigma, as is its closest rela- tive P. monticola F. Muell., a species with the endocarp 4-6 mm long, and with the seed forming °f the drupe; pileus 1 3—14 mm long ; stigma 5-6 mm long ; anthers oblong-linear; free filament tips 0. 1-0.2 mm long; and the leaves 32-43 mm wide. P. plu- visilvaticus has the endocarp 7-11 mm long, and with the seed forming %-!/2 of the drupe ; pileus 6.5-12 mm long; stigma 3-4 mm long; anthers tapering upward ; free filament tips 0.7-1. 3 mm long; and the leaves 36-58 mm wide. P. pluvisilvaticus is actually a common species, but it has been confused with the rather poorly described P. monticola F. Muell., which occurs in eastern Queensland in the rain forests at from 18° to 21° South. It has been reported by von Mueller from the Russell River at 17° 30' South, but he also had a col- Page 393: Revision of Pandanus , 21. Queensland — St. John 281 lection of P. monticola with exactly duplicating data. There may have been a confusion in the data of these two collections. P. pluvisilvaticus occurs in the rain forests of eastern Queensland from 17° to 16° South, and is known from numerous collections. It shows some variability in the width of the leaf blades and considerable variation in the leaf spines, but it seems to represent a taxon of fairly wide distribution. Its fruit characters are dependable as a basis for separation from the older P. monticola F. Muell. The new epithet is coined from the Latin pluvia, rain, and silvaticus, woodsy, and is given with reference to the habitat of the species, the rain forest. Revision of the Genus Pandanus Stickman, Part 22 A New Species (Section Hombronia) from New Caledonia Harold St. John1 Numerous botanists in the tropics of the Pacific and of the eastern hemisphere have assisted the writer by making new collections of Pandanus. The new species here announced is named in honor of one of these cooperators, Prof. H. S. McKee of the University of Sydney. Pandanus Mc-Keei sp. nov. (sect. Hombronia') Figs. 238-239 diagnosis holotypi : Arbor 5 m alta ramosa, radicibus fulturosis nullis, foliis 1.5 m longis proxima basem 7.2 cm latis in medio 6.5 cm latis coriaceis supra obscure viridibus infra pallide viridibus basi rubra ligulatis gradatim ex media ad apicem diminuentibus, basi paene latiore exarmata sed ex 11-12 cm marginibus cum aciculato-serris 1-1.5 mm longis 1-5 mm distantibus eis in puncto nigro, midnervo toto exarmato, circa medim marginibus cum serris nigris 0.3-0. 5 mm longis 2-6 mm distantibus, circa apicem marginibus cum dentibus nigris crenulatis 0.1-0. 3 mm longis 1-3 mm distan- tibus ; pedunculo folioso, syncarpio solitario apicali 36 cm longo 17 cm diametro anguste oblongo-ellipsoideo cum circa 286 phalangibus ordinatis in helicis et sereis verticalibus, pha- langibus 4. 2-4.6 cm longis 2.4-3. 1 cm latis 1.5- 2 cm crassis rubro-brunneis sed intra pallide luteis elliptico-oblongis in medio subcontracts parte supera % libera apice concavo in sicco brunneis, lateribus minime curvatis et cum fissuris multis brunneis longitudinalibus, 5-6- angulosis, suturis lateralibus nullis, carpellis 5-7 in piano unico laterali ordinato, stigmatibus 1.5- 2. 6 mm latis brunneis cordatis vel obliquiter cordatis in latere ad apicem syncarpii obtutis, endocarpio in parte % infera et 12-14 mm longo osseoso brunneo lateribus 3—4 mm crassis, seminibus 6-8 mm longis 2-3 mm diametro ellipsoideo, mesocarpio caverno unico 25 mm longo pluri-fibroso et cum membranis albis 1 B. P. Bishop Museum, Honolulu, Hawaii. Manu- script received December 3, 1962. medullosis formanti, mesocarpio basali fibroso et cavernoso 3-6 mm longo. description of holotype: Tree 5 m tall, branched several times; prop roots none; leaves 1.5 m long, 7.2 cm wide near the base, 6.5 cm wide at the middle, coriaceous, dark-green above, light-green below, base red, ligulate, gradually tapering from the middle to the apex, the base scarcely widened, unarmed, beginning at 11-12 cm the margins acicular-serrate, the teeth 1-1.5 mm long, 1-5 mm apart, black- tipped; the midrib unarmed throughout; at the midsection the margins black serrulate, the teeth 0.3-0. 5 mm long, 2-6 mm apart; near the apex the margins black crenulate, the teeth 0. 1-0.3 mm long; peduncle leafy bracted; syncarp solitary, terminal, 36 cm long, 17 cm in diameter, narrowly oblong-ellipsoid, the pha- langes about 286, in spiral and in vertical rows; phalanges 4.2 -4.6 cm long, 2.4-3. 1 cm wide, 1.5—2 cm thick, reddish-brown without, light-yellow within, elliptic-oblong, slightly contracted at midsection, upper I/4 free; apex concave, dull, brownish when dry, the exposed sides gently curving and with numerous longi- tudinal, brown cracks, 5-6-angled ; lateral sutures none; carpels 5-7, in one lateral plane; stigmas 1.5-2. 6 mm wide, brown, cordate or obliquely so, facing sideways toward the apex of the syncarp; endocarp at lower and 12-14 mm long, bony, brown, the lateral walls 3-4 mm thick; seeds 6-8 mm long 2-3 mm in diameter, ellipsoid; apical mesocarp forming one cavern 25 mm long, with an open passage above each seed, with numerous, strong, longi- tudinal fibers and pale medullary membranes; basal mesocarp fibrous and cavernous, 3-6 mm. long. holotypus: New Caledonia, 7 km. S. of Riviere des Pirogues, on bank of creek, 50 m alt, 15 Oct., 1955, H. S. McKee 5,229 (bish). Isotype (L) ! 282 "WO OS Page 395: Revision of Pandanus , 22. New Caledonia — St. John 283 Fig. 238. Pandanus Mc-Keei St. John, from holotype. a, Syncarp, lateral view, X 1/5; b, drupe, proximal view, X 1 ; c, drupe, lateral view, X 1 ; d, drupe, longitudinal median section, X 1 ; e, drupe, apical view, X 1 ; f, g, stigmas, proximal view, X 4. 1 cm,. 284 PACIFIC SCIENCE, Vol. XXI, April 1967 a Fig. 239- Pandanus Mc-Keei St. John, from holotype. a, Leaf middle, lower side, X 1; A leaf base, lower side, X 1; c, leaf apex, upper side, X 1- Page 397: Revision of Pandanus , 22. New Caledonia — St. John 285 discussion: P. Mc-Keei is a member of the section Hombronia. There its closest relative is P. Balansae (Brongn.) Solms of New Cale- donia, which species has the syncarp 17—18 cm long, 13-14 cm in diameter; phalanges 5.5 cm long, 2.5 cm wide, the apex pyramidal to truncate apex which is half as wide as the phalange; carpels 3-6; and endocarp at the lower third of the phalange. P. Mc-Keei has the syncarp 28.8 cm long, 12.8 cm in di- ameter; phalanges 4.2-4. 6 cm long, 2.4-3. 1 cm wide, the convex apex nearly as wide as the body; carpels 5-7 ; and the endocarp at the lower I/4 of the phalange. The species is dedicated to Prof. H. S. McKee, the collector. Soil- Vegetation Relationships in Hawaiian Kipukas1 D. Mueller-Dombois and C. H. Lamoureux2 Kipuka, the Hawaiian word for "opening/’ has come into scientific usage as a term used to designate an older area on the slopes of vol- canic mountains that has been surrounded by more recent lava flows. Kipukas are common landscape features on the slopes of Mauna Loa and Kilauea volcanoes on the island of Hawaii, where they can be readily recognized as islands of denser vegetation in the vast, sparsely vege- tated areas. They range in size from a few square meters to hundreds of acres. Kipukas are of special interest for several reasons. As vegetation islands they provide seed- source centers for the invasion of vegetation on new volcanic material. As vegetation islands they represent somewhat simplified ecosystems, analogous to bogs or lakes, that are very suitable for studying internal ecological relationships. The isolation of small populations in kipukas provides unique opportunities for evolutionary studies. So far, very little ecological work has been done with Hawaiian kipukas. Need for such work has arisen in Hawaii Volcanoes National Park, where the Park Service is confronted with the task of interpreting certain kipuka features to the Park visitors. Kipuka Puaulu, popularly known as "Bird Park,” has been accessible to the public for some time and the nearby Kipuka Ki is soon to be opened. For this reason the present study was begun in these two kipukas. Rock described the flora of both kipukas in an undated manuscript (probably written around 1910) and reported a few general ecological observations. He remarked upon the unique and complex composition of arborescent species from which he judged both kipukas to be "of great age.” However, as an approximation he cited the estimate of Professor T. Jaggar (geologist at the Hawaii Volcano Observatory at that 1 The study was financed through U.S. Government Contract No. 14-10-0434-1504 to Dr. M. S. Doty, "Bioecological investigations of Hawaii Volcanoes National Park.” Manuscript received April 15, 1966. 2 Department of Botany, University of Hawaii. time), which placed the kipuka’s origin within the Christian era (i.e., less than 2,000 years). Rock recorded 40 arborescent native species forming a complex forest type in Kipuka Pu- aulu. Only half this number of tree species were found in Kipuka Ki. He also noted the presence of two vegetation types in Kipuka Puaulu, a complex forest type containing many tree species and a Metrosideros- dominated type. He believed that soil differences were respon- sible for the presence of these two types of forest. A general description of the kipuka soils is given in the Soil Survey report for the Ter- ritory of Hawaii (Cline et ah, 1955), where the soils were classified as Latosolic Brown Forest soils derived from two layers of volcanic ash. The primary objectives of this present study were to determine the floras of both kipukas, to describe the vegetation types present in each, and to determine what soil-vegetation relation- ships exist in these places. DESCRIPTION OF AREA Both kipukas occur at an elevation of from 1200 to 1300 m on the southeast slope of Mauna Loa approximately 3 km northwest of Kilauea crater (Fig. 1). The central elevation of Kipuka Ki is about 60 m higher than that of Kipuka Puaulu. Both are surrounded and separated by recent beds of rough aa lava. Their boundaries are about 800 m apart. Kipuka Puaulu is about 42 hectares and Kipuka Ki about 18 hectares in size. The climate is char- acterized by a rather uniform mean annual tern- perature of 16°C, which is 7°C cooler than that experienced at sea level. The mean variation be- tween the warmest month (August) and the coolest (February) is only 3.5 °C. Occasional freezing temperatures can be expected during February nights. Approximate annual rainfall is 1500 mm, varying monthly from about 25 mm in June to 200 mm in January. According to Krajina’s (1963) zonal classification, the ki- 286 Soil-Vegetation Relationships in Kipukas — Mueller-Dqmbois and Lamoureux 287 air photo taken October 1954.) pukas occur in the lower Metrosideros zone, whose climate is described as humid marine tropical (or subtropical) with common clouding. The kipukas are somewhat more sheltered from the windward rains than much of this zone, and Rock (1913) described them as being occupied by dry-mixed forest. Both kipukas are situated on moderate south slopes and have an irregularly undulating topog- raphy with a few short, steep slopes, several level areas, a few larger somewhat inclined areas, and scattered small pocket-like depres- sions. Two distinct vegetation formations were found in Kipuka Puaulu: a closed to semi-open forest type (Fig. 1, type 1), and a savannah type with a dense grass cover and scattered trees of Metrosideros and Acacia koa (Fig. 1, type 2). Kipuka Ki is dominated by a moderately stocked forest vegetation type, which in places is also semi-open (Fig. 1, type 1), but it lacks the very dense or closed forest stand segments found in Kipuka Puaulu: a closed to semi-open also in Kipuka Ki, representing there, however, mainly a transition zone in which occasional lava rocks protrude to the surface (Fig. 1, type 2). Characteristically, no rocks are found near the surface in either kipuka with the ex- ception of the transitory savannah in Kipuka Kb Within the forest formation of both kipukas several smaller plant communities can be recog- nized. One of the more obvious associations, common to both kipukas, is characterized by a ground cover of Microlepia setosa , a lush fern up to 1 m tall. The tree layer is dominated by Acacia koa and Sap Indus saponaria . This plant association occurs on level to moderately sloping ground. METHODS For the purpose of comparing the soils of the two kipukas, soil pits were dug in each kipuka in the Microlepia community near a tall Acacia koa tree in a level place. A level place near a tall koa tree was also chosen for a soil pit in the savannah for making a comparison between the soils of the forest and the savannah formation within Kipuka Puaulu (Fig. 1). The reason for choosing a level place was that the soils there were presumably not influenced by lateral seep- age. 288 Each pit was dug to a depth of 2 m. The soil horizons were described as to depth, material, and color, and samples were collected for lab- oratory analysis. The soil samples included three sets, one for microbiological analyses (now be- ing conducted), one for current soil moisture analysis, and one for other soil tests. In addi- tion, the three soil profiles were prepared as soil monoliths after the method of Smith and Moodie (1947) for further mega- and micro- scopic inspection and as permanent records. Subsequent soil tests carried out included determination of moisture equivalents (by the centrifuge method), permanent wilting per- centages (by the sunflower method), organic carbon (by the Walkley-Black wct-combustion method), and pH (by electric pH meter). Herbarium specimens were prepared. One set has been deposited in the herbarium of the University of Hawaii, and a second set in the herbarium of Hawaii Volcanoes National Park. RESULTS AND DISCUSSION A. Soils The soils give convincing evidence that they have been derived from volcanic ash and not from old, disintegrated lava as has been as- sumed by the authors who published the nature trail guide for Kipuka Puaulu (1961 edition). Ash strata were found to the depth of 2 m to which all soil pits were dug and there was no sign of parent material change at this depth. Rock (undated) indicated that the soil in K - puka Puaulu was nearly 6 m deep. The maxi- mum soil depth in Kipuka Ki is not known. Ash was deposited not at one time but in several stages, probably extending over many hundreds of years. Corresponding ash layers that appear to have originated from the same eruptions can be found in all soils we examined. Noteworthy are two thin red pumice layers that occur in each soil. One occurs in the lower pro- file at 100 cm depth in the soil of Kipuka Ki, at 140 cm in the forest soil of Kipuka Puaulu, and at 145 cm in the savannah soil (Fig. 2). A second red pumice layer is found in all soils nearer the surface, at 60 cm in Kipuka Ki, at 70 cm in the forest soil of Kipuka Puaulu, and at 85 cm in the savannah soil. PACIFIC SCIENCE, Vol. XXI, April 1967 Ash deposits were composed of at least five different materials: a fine, dusty gray ash with scattered pebbles up to 5 mm in diameter, a gravelly ash with basaltic and variously vesicular pebbles up to 1 cm in diameter, a black vitreous ash, a yellow-olive pumice, and the red pumice mentioned above. The fine, dusty gray ash occurs at a depth of 15-20 cm from the surface in all profiles. It is most pronounced in the savannah soil and least so in the soil of Kipuka Ki. This layer looks like the leached layer of a podzolic soil. However, there are three arguments against this interpretation. First, the layer is brightest under savannah, which has the least acid surface layer (Table 1). Second, it was horizontally continu- ous only in Kipuka Puaulu, whereas it occurred in local pockets in Kipuka Ki. Third, Went- worth (1938), in his study of ash formations around Kilauea Crater, described a "gray-laven- der, fine sand-size ash” near the surface in sev- eral places which seems to fit this layer. The gravelly ash was described by Went- worth as "basalt in glass” and is well shown in the savannah soil, where it recurs as a thin layer (usually ± 5 cm thick) at depths of 30, 50, and 70 cm. Black vitreous ash appears as a layer 20 cm deep in all three soils, from 50—70 cm depth in Kipuka Ki, from 60-80 cm in the forest soil of Kipuka Puaulu, and from 75-95 cm in the savannah soil. It recurs at three places above this layer (at 65 cm, 45 cm, and 25 cm) in the savannah soil. These black layers are black not only from ash but also, perhaps more domi- nantly so, from an extremely high incorporation of organic carbon (between 10.1 and 15.7%, Table 1). A yellow-olive pumice layer (called "reticu- lite” by Wentworth) is found in the savannah soil incorporated into the black layer at 25 cm depth. Some of this pumice occurs also in both forest soils beneath the fine gray ash layer (Cl), but here it is less abundant and less well strati- fied (Fig. 2). The lower ash deposits, from the thick black layer (Alb) down, in both soils of Kipuka Puaulu are not stratified horizontally, whereas the upper ones are more or less horizontally stratified (see Fig. 2, P± and P2 )• Angles of Soil-Vegetation Relationships in Kipukas — Mueller-Dombois and Lamoureux 289 departure were between 20 and 30°. This fact indicates that there have been some relief changes throughout the build-up of the soil to its present surface level. This suggests some- thing about the origin of Kipuka Puaulu, which may apply to Kipuka Ki as well. It appears probable that lateral translocation of ash has occurred after deposition as a result of wind or water erosion, especially during the early stages when the kipuka was only sparsely vegetated. A small kipuka of about 1 hectare in the Kau Desert south of Kilauea crater, which is just "in-the-making,” shows that it has originated as a small dune ecosystem. Gray-black sandy ash was deposited here in a thin layer on a large flat area of smooth pahoehoe lava. Wind has swept Ki P| P2 HORIZON NAME Fig. 2. Comparison of horizons of kipuka soils {Ki, forest soil of Kipuka Ki ; Pv forest soil of Kipuka Puaulu; P2, savannah soil of Kipuka Puaulu). Color symbols from Munsell charts refer to air-dry soil. No- menclature of horizons after the 1962 Supplement to the Agriculture Handbook No. 18, Soil Survey Manual. Some Parameters of the Kipuka Soils 290 PACIFIC SCIENCE, VoL XXI, April 1967 § § 2 o o xp o C a lp S O C _c C/3 >N o NO < PO PQ c 6 m 3 CN 3 Walkley-Black values. 4 Measured electrometrically. 5 Ki = Kipuka Ki, forest soil. 6 Pi = Kipuka Puaulu, forest soil. 7 Pa = Kipuka Puaulu, savannah soil Soil-Vegetation Relationships in Kipukas — Mueller-Dombois and Lamoureux 291 up much of this ash and redeposited it as a dune at a place where the smooth lava was inter- cepted by a rough aa flow. The ash-dune now represents an island supporting pioneer vegeta- tion. This process is accumulative since the vege- tation, once established, catches more eolian de- posits and in turn contributes organic matter, soon forming a moisture- and nutrient-improved habitat that also differs in elevation from its surroundings. It is quite conceivable that such elevated dune ecosystems can be surrounded by subsequent lava flows. Such an occurrence, on a much larger scale, could account for the origin of the kipukas discussed here, although addi- tional evidence to support this hypothesis must be obtained. Also, many, if not most, of the Hawaiian kipukas, such as Kipuka Nene, un- doubtedly have developed merely by the dis- integration in situ of older lavas, the kipuka area being subsequently surrounded by newer flows. The upper ash deposits in the kipuka soils are more or less horizontal with respect to the present soil surface, a form of deposition which Powers (1948) calls "blanket deposits." The "blanket deposits" in the savannah soil show that there have been at least 9 ash deposits in Kipuka Puaulu since establishment of the thick black horizon (Alb). Not all of these may have been derived from different explosions, but Powers has discovered ash from at least 26 eruptions in the area that occurred later than the big Kilauea ash explosion of 1790. The latest recorded near Kipuka Puaulu was from the 1924 eruption. This shows that the soil is not of one (old) age, but is of several ages from older to younger, and the surface soil may even be much younger than the surrounding rough aa flow, rather than older as indicated by Rock. The surrounding flow is prehistoric, thus at least pre-1778. Fragments of charcoal were found in both kipukas in the forest soils. They occurred at 70 cm depth in Kipuka Ki and at 80 cm in Kipuka Puaulu. This indicates two facts. First, there was fire in both kipukas at an earlier date in their development; and second, both had woody vegetation growing on them at that time. Al- though charcoal was not found in the savannah soil, fire may explain its origin. It is interesting that the savannah soil looks quite different from the forest soils which, in spite of being in two separate kipukas, show much similarity in ap- pearance. Both forest soils are deeply melanized, dark brown in color, and are rather uniformly enriched with organic carbon (Table 1). The savannah soil shows more clearly the parent material, because of less uniform melanization. Here organic carbon content fluctuates greatly between soil horizons. These two patterns, that is, the more uniform color and organic carbon distribution in the forest soils and the greater variation in color and organic carbon distribu- tion in the savannah soil, are undoubtedly asso- ciated with past rooting zones. One may assume that a mixed, well-stocked forest occupies the soil volume more uniformly than does a domi- nantly grass-covered savannah. The grass and ground vegetation roots may have been more restricted to the black horizon zones. Such a concentration in rooting depth was also found at present at the soil surface of the savannah soil. This pattern supports the assumption that the savannah originated after a fire. It is prob- able that the fire occurred when the 20 cm-thick, black layer, the buried A1 horizon (Alb), was at the surface supporting actively growing veg- etation, because the charcoal was found right at the lower boundary of this layer in both forest soils (Fig. 2). Therefore, the savannah may be quite old. The C-14 date of the 70 cm -deep charcoal in Kipuka Ki came to 2,170 dz 200 years, i.e.9 about 220 years b.c.3 Analyses of potentially available water, or- ganic carbon, and pH show no significant dif- ferences between savannah soil and forest soil in Kipuka Puaulu, so that neither soil water nor nutrient differences can be assumed to be responsible for the difference in vegetation. Moreover, there is no distinctive topographic pattern associated with either type, so that the savannah’s origin is not attributable to environ- mental differences related to topography. B. Flora Between November 1963 and March 1965 botanical surveys were made of both kipukas. Voucher specimens have been deposited in the herbarium of the Department of Botany, Uni- versity of Hawaii, and duplicate specimens in 3 Sample GX0394, Geochron Laboratories, Inc. 292 PACIFIC SCIENCE, Vol. XXI, April 1967 the herbarium of Hawaii Volcanoes National Park. At the end of this article will be found a check list of the plants found in the two ki- pukas. It includes also records from Rock (un- dated, and 1913), and Fagerlund and Mitchell (1944), as well as specimens in the herbaria of Hawaii Volcanoes National Park and the Ber- nice P. Bishop Museum, Honolulu, Hawaii. Table 2 summarizes the information provided in the check list. It shows that Kipuka Puaulu now contains, and has contained, significantly more species of vascular plants than has Kipuka Ki. Table 3 provides an analysis of the numbers of species common to both kipukas and of those found only in Kipuka Puaulu or Kipuka Ki. This indicates that, while each kipuka contains species which the other does not, Kipuka Puaulu has a significantly greater number of unique species than does Kipuka Ki. Thus, the observations recorded in Tables 2 and 3 and in the check list agree with Rock’s (1913) observation that there are more species TABLE 2 Numbers of Species, Varieties, and Forms of Vascular Plants Recorded from Kipuka Puaulu and Kipuka Ki PLANTS ALL DATA KIPUKA PUAULU KIPUKA KI Total number of spp. Native spp. Native trees Introduced spp. 158 ( 104) * 86 (52) 42 (21) 72 (52) 145 (92) 81 (48) 42 (21) 64 (44) 73 (63) 36 (30) 15 (11) 37 (33) * Figures outside parentheses include all spp. ever Figures within parentheses include all spp. growing in 1963-65. recorded. naturally TABLE 3 Distribution of Species, Varieties, and Forms of Vascular Plants Between Kipuka Puaulu and Kipuka Ki plants COMMON TO BOTH KIPUKAS KIPUKA PUAULU ONLY KIPUKA KI ONLY Total number of spp. Native spp. Native trees Introduced spp. 60 (51)* 31 (26) 15 (11) 29 (25) 85 (41) 50 (22) 27 (10) 35 (19) 13 (12) 5 (4) -(-) 8 (8) * Figures outside parentheses include all spp. ever recorded. Figures within parentheses include all spp. growing naturally in 1963-65. in Kipuka Puaulu. The number of native trees now growing in Kipuka Puaulu (21) is almost twice as large as that in Kipuka Ki (11). How- ever, Rock reported in 1913 that there were at least 40 native tree species growing in Kipuka Puaulu. Even allowing for differences of taxo- nomic opinion, the decrease in number of tree species during the last 50 years appears quite remarkable. In his book, "Indigenous Trees of the Hawaiian Islands,” Rock (1913) included 19 photographs of tree species in Kipuka Pu- aulu. Fifteen of these photographs show bits of landscapes and ground vegetation, which at that time appeared badly abused by cattle graz- ing. Many areas appear barren or show trampled ground vegetation, and several pictures show broken-down trees. From the photographic rec- ord one could assume that the present savannah formation is caused entirely by cattle grazing. However, two photographs show what appear to be sections of the present savannah forma- tion. One of these shows a dense cover of Pteridium, which today is also well established in the savannah. Inasmuch as fire has definitely occurred in both kipukas, it is believed that fire may have created openings in the forest that were aggravated and maintained by subsequent cattle grazing. It seems probable that cattle were guided in their grazing habits by this fire- conditioned vegetation pattern, since a denser ground vegetation would be found in the open, coupled with fewer obstacles for the movement of cattle. Increased cloud interception and fog drip in the forest (Ekern, 1964) may also have contributed to maintaining the pattern. This is indicated by the greater current moisture con- tent in the lower profile of the forest soil (Table 1) and the location of the kipukas in a zone of common cloud occurrence (Krajina, 1963). There are several possible explanations for the larger number of both native and introduced species in Kipuka Puaulu. 1. The larger number of native species in Kipuka Puaulu may be related to: (a) larger area. Both kipukas are so much larger than the "minimal” area-size of forest stand communities cited in the literature (Ellen- berg [1956] gives 500 m2; Cain and Castro [1959], < 20,000 m2 for tropical rain forest), that one may think that size is not a factor. Soil-Vegetation Relationships in Kipukas — Mueller-Dombois and Lamoureux 293 However, such minimal area calculations are based on the more common species. From the records it is quite clear that the now extinct species were extremely rare. The smaller size of Kipuka Ki can be used, therefore, as one ex- planation for its smaller number of indigenous species. (b) greater age. Rock (1913) believed that Kipuka Ki was more recent in origin than Puaulu, because of the common assumption that an older area would have more species. His idea cannot be disproved from current evidence, but one observation points in the opposite direc- tion. The amount of organic carbon did not decrease in the lower profile of Kipuka Ki, whereas it did so in both soils of Kipuka Puaulu (Table 1). This may indicate vegetative activ- ity at an earlier date in Kipuka Ki as compared with Puaulu. (c) GREATER DIVERSIFICATION IN HABITATS. This factor in Kipuka Puaulu is not expected from observations made so far. Both kipukas have similar topographic variations and deep, rich soils. Also, the distribution of tree species is not as likely to be affected by small-scale environmental variations as is that of herba- ceous plants. (d) DIFFERENT HISTORY OF DISTURBANCE. Little definite information is available on differences in disturbance-history. We know only that three important disturbance factors operated in both kipukas: fire, cattle grazing, and pig damage. Current pig damage appears to be less in Kipuka Ki. Past cattle grazing also was probably less devastating here. It is pos- sible, however, that fire eliminated a few trees, either directly, or indirectly by competition of more aggressive plants that followed the fire in both kipukas. In this connection, the chance of the smaller, isolated kipuka to be restocked with rare species would be less than that of the larger one, which also may have provided a greater chance of survival of rare tree species simply because of its larger size. (e) DIFFERENCES IN RAINFALL AND PRO- DUCTIVITY. It was interesting to find that the current moisture distribution downward in the soils differed between the kipukas. The current soil water content increased considerably in the bottom part of the profile in Kipuka Ki and was higher than in the soils of Kipuka Puaulu, whereas the upper part of the profile was drier than that of the soils in Kipuka Puaulu. This indicates a different rain shower pattern be- tween the kipukas. This may be a random pat- tern, however, which then would have no bear- ing on the total amount received. Except for the lower profile parts (B3 and C2), there was little difference in the amount of organic carbon in the two forest soils, indicating a similar pro- ductivity in both kipukas. Thus, the differences in number of species cannot be related to dif- ferences in productivity. 2. The larger number of introduced weed species in Kipuka Puaulu may be caused by (a) its greater exposure to man and cattle, and (b) its larger sun-exposed area, which favors the establishment of shade-intolerant weeds. It is interesting that the fewer weed species in Ki- puka Ki occupy more ground. Some of them have formed dominant communities. C. V egetation Several obvious plant communities occur un- der the forest cover. They are represented by native and introduced plants as follows: NATIVE Microlepia association Nephrolepis association Peperomia patches Pipturus shrub strata Coprosma thickets INTRODUCED Commel'ma association Rubus penetrans association Solanum association Dactylis patches Com m,eUna-N ep h rolepis mixed community Nephrolepis communities and Dactylis patches are common also in open areas. Cop- rosma thickets are characteristic only for Kipuka Puaulu. The Rubus penetrans and Solanum as- sociations are characteristic for Kipuka Ki. Only one small Solanum patch was observed in Ki- puka Puaulu. All other associations occur in both kipukas. Peperomia patches seem to be estab- lished on ground that has been rather recently scarified by pigs, and form there a pioneer com- munity in shaded habitats. Similarly, Coprosma thickets are associated with pig scarification, 294 PACIFIC SCIENCE, VoL XXI, April 1967 which is particularly pronounced under larger Sapindus trees where the pigs seem to search for their fruits. SUMMARY AND CONCLUSIONS 1. The soil of both kipukas is derived from several ash deposits. The lower, sloping ones in Kipuka Puaulu differ from the upper ones, which are stratified horizontally. 2. Charcoal was found in both kipukas un- der forest in association with a buried, black surface horizon, at 70 cm depth in Kipuka Ki and at 80 cm in Puaulu. 3. The C-14 analysis of the 70 cm-deep charcoal in Kipuka Ki indicates that a fire occurred at about 220 B.c. 4. The forest soils of both kipukas are uni- formly melanized, showing considerable mega- scopic similarity, and differ markedly from the savannah soil, which showed melanization re- stricted to narrow layers and which exposed a clear parent material stratification. 5. The soil parameters tested indicated no significant differences between the forest soil and the savannah soil of Kipuka Puaulu in terms of soil water, organic carbon, and pH. 6. The forest soils of the kipukas differ only in current soil moisture distribution and organic carbon content of the lower horizons (B3 and C2). The work so far is only an introduction to the plant ecology of Hawaiian kipukas and points to the need for the following further research : 1. Analysis of photographs. It would be profitable to examine all photographs Rock made of Kipuka Puaulu and, if possible, to identify some spots for rephotographing. This could reveal certain interesting successional changes over the last 50 years. 2. Current observation indicates reoccupa- tion of the savannah by forest. This appears to be accomplished by sucker growth of Acacia koa. Invasion of trees by seed seems practically impossible. It would be interesting to study the rate of reinvasion, now, when there is no more interference by cattle. 3. Studies of cloud interception. Differences in soil water supply as a result of fog drip should be investigated, to determine the role this environmental factor plays in influencing the rate of reinvasion of forest into the savan- nah. 4. Measuring pig damage. Current observa- tion indicates that pigs affect the forest vegeta- tion in two ways, (a) By scarifying the surface, they eliminate ground vegetation and provide ideal seed beds for tree seed germination of that which is left. Many formerly pig-scarified areas seem to come back in thickets of tree seedlings of Sapindus and Coprosma. (b) During periods of food scarcity or over-population pigs seem to gnaw away the bark of trees, particularly of Coprosma, thus damaging them severely, e.g., by providing entrance avenues for pathogens. The food habits of pigs should be studied in connection with population counts to explain their influence on vegetation patterns. 5. Quadrat studies of vegetation patterns. These should be done in particular with Pepe- romia , as a probable native pioneer on pig- scarified ground; with Commelina and Nephro- lepis mixed associations, to determine whether Commelina takes over the habitats occupied by the native fern, Nephrolepis ; and with the two weed communities formed by Rubus penetrans and Solanum pseudocapsicum , to determine their effect on the native Microlepia association. 6. An ecological survey of all kipukas and their surroundings should be made in an at- tempt to assess their development in succession and their influence on the vegetation of the sur- rounding more recent volcanic material. CHECK LIST OF PLANTS IN THE KIPUKAS This check list includes all of the vascular plant species of Kipuka Puaulu and Kipuka Ki as of May 1965. The symbols used are: * = native Hawaiian species ; # — native tree ; -f- = growing, apparently naturally, in kipuka in 1963-65; a = growing in kipuka in 1963- 65 only as individuals recently planted by Na- tional Park Service ; b = specimens, collected between 1930 and I960, in herbaria at Hawaii Volcanoes National Park or B. P. Bishop Museum, but species not found growing in kipuka in 1963-65 ; c = reported by Fagerlund and Mitchell (1944), but no specimens avail- able; d = reported by Rock (undated, and 1913), but no more recent specimens available. Soil-Vegetation Relationships in Kipukas— Mueller-Dombois and Lamoureux 295 SPECIES KIPUKA PUAULU KIPUKA KI PTERIDOPHYTA ASPIDIACEAE * Athyrium sandwichianum Presl. + Cyclosorus dentatus ( Forsk. ) Ching C. parasiticus (L.) Farwell 4- * Cyrtomium caryotideum (Wall.) Presl. + * Dryopteris glabra (Brack.) Kuntze b * D. hawaiiensis (Hillebr.) Christ + + * D. latifrons (Brack.) Kuntze + * D. pale ace a (Swartz) Christensen + 4- * Elaphoglossum conforme (Swartz) Schott b ASPLENIACEAE * Asplenium adiantum- nigrum L. + * A. cf. caudatum Forst. f. 4- + * A. macraei Hook, et Grev. c b BLECHNACEAE * Sadleria cyatheoides Kaulf. + + DAVALLIACEAE * Nephrolepis exaltata (L.) Schott + + POLYPODIACEAE * Pleopeltis thunbergiana Kaulf. 4- + PSILOTACEAE * Psiloium nudum (L.) Griseb. 4- PTERIDACEAE * Cibotium chamissoi Kaulf. d * C. glaucum (Smith) Hook, et Am. 4- 4- * Conio gramme pilosa (Brack.) Hieron. 4- * Microlepia setosa (Smith) Alston 4- + * Pellaea ternifolia (Cav.) Link b SPECIES KIPUKA PUAULU KIPUKA KI * Pteridium aquilinum (L.) Kuhn 4- * Pteris cretica L. + + * P. excels a Gaud. 4- MONOCOTYLEDONAE COMMELINACEAE Commelina diffusa Burm. 4- + CYPERACEAE * Car ex macloviana D’Urv. var. subfusca (Boott) Kiikenth. 4- 4- * C. wahuensis C. A. Meyer var. rubiginosa R. W. Krauss 4- Cyperus brevifolius (Rottb.) Hassk. 4- 4- * C. hillebrandii Boeck. b * C. polystachyus Rottb. 4- GRAMINEAE Agrostis retrofracia Willd. 4- Anthoxanthum odoratum L. 4- + Briza minor L. b Bromus commutatus Schrad. c B. rigidus Roth c + B. se cal in us L. b B. unioloides (Willd.) H.B.K. 4- + Cynodon dactyl on (L.) Pers. 4- + Dactylis glomerata L. + 4- Digitaria pruriens (Trin.) Buese d Festuca dertonensis (All.) Asch. et Graebn. b Holcus lanatus L. f * Panicum tenuifolium Hook, et Arn. b Paspalum con'jugatum Berg. 4- P. dilitatum Poir. + 4- P. urvillei Steud. 4- Poa annua L. 4- P. pratensis L. 4- Set aria geniculata (Lam. ) Beauv. 4- 296 PACIFIC SCIENCE, VoL XXI, April 1967 SPECIES KIPUKA PUAULU KIPUKA KI Sporobolus africanus (Poir.) Robyns et Tournay + Stenotaphrum secundatum (Walt.) Kuntze b Unidentified grass + + IRIDACEAE X Tritonia crocosmaeflora Lemoine + LILIACEAE Cordyline terminalis (I.) Kunth + * S mil ax sandwicensis Kunth d ZINGIBERACEAE Hedychium coronarium Koenig + DICOTYLEDONAE AMARANTHACEAE * # Charpentiera obovata Gaud. + a APOCYNACEAE * Alyxia olivaeformis Gaud. + * # Ochrosia sandwicensis A. Gray a ARALIACEAE Brassaia actinophylla F. Muell. b * # Cheirodendron trigynum (Gaud.) Heller + celastraceae * # Perroitetia sandwicensis A. Gray + COMPOSITAE Achillea millefolium L. b Bidens pilosa L. + + Cirsium lanceolatum (L.) Hill. + Erigeron albidus (Willd.) A. Gray b E. canadensis L. b Hypochaeris radicata L. + + Senecio sylvaticus L. c Sonchus asper L. + + S. oleraceus L. c SPECIES KIPUKA PUAULU KIPUKA KI CONVOLVULACEAE * Ipomoea indica (Burm.) Merr. + + CRUCIFERAE Lepidium virginicum L. b Sisymbrium officinale (L.) Scop. b EP ACRID ACE AE * Styphelia tameiameiae (Cham.) F. Muell. + + EUPHORBIACEAE Aleurites moluccana (L.) Willd. a FLACOURTIACEAE * # Xylosma hawaiiensis Seem. var. hillebrandii (Wawra) Sleumer b GENTIAN ACE AE Centaurium umbellatum Gilib. b GERANIACEAE Geranium carolinianum L. var. australe (Benth.) Fosb. + HYPERICACEAE Hypericum mutilum L. + LABIATAE Mentha sp. + LAURACEAE Per sea americana Mill. + LEGUMINOSAE * # Acacia koa A. Gray + + Desmodium uncinatum (Jacq.) DC. + + * # Sophora chrysophylla (Salisb.) Seem. + + LOBELIACEAE * # Clermontia hawaiiensis (Hillebr.) Rock d * # Clermontia sp. a LORANTHACEAE * Korthalsella complanata (Van Tiegh.) Engl. + LYTHRACEAE Cuphea carthaginensis (Jacq.) Macbride b Ly thrum maritimum H. B. K. + + Soil-Vegetation Relationships in Kipukas — Mueller-Dombois and Lamoureux 297 SPECIES KIPUKA PUAULU KIPUKA KI MALVACEAE * # Hibiscadelphus giffardianus Rock a a * # H. hualalaiensis Rock a * # Kokia rockii Lewt. a Modiola caroliniana (L.) G. Don + + MENISPERMACEAE * Cocculus ferrandianus Gaud. + MORACEAE Ficus carica L. b MYOPORACEAE * # Myoporum sandwicense A. Gray var. fauriei (Lev.) Kraenzl. + + MYRSINACEAE * Embelia pacifica Hillebr. + * # My r sine lessertiana A. DC. + + MYRTACEAE * # Metrosideros polymorpha Gaud. + + Psidium cattleianum Sabine + P. guajava L. + NYCTAGINACEAE * # Heimerliodendron brunonianum (Endl.) Skottsb. + a OLEACEAE * # Osmanthus sandwicensis (A. Gray) Knobl. + + ONAGRACEAE Oenothera stricta Ledeb. + OXALIDACEAE Oxalis corniculata L. + + PAPAVERACEAE * Argemone glauca L. ex Pope b PASSIFLORACEAE Passiflora ligularis Juss. + PHYTOLACCACEAE * Phytolacca sandwicensis Endl. + SPECIES KIPUKA PUAULU KIPUKA KI PIPERACEAE * Peperomia cookiana C. DC. + + * P. hypoleuca Miq. + * P. leptostachya Hook, et Arn. b '*■ P. reflexa Dietr. var. reflexa C * P. reflexa Dietr. var. parvifolia C. DC. + + PITTOSPORACEAE * # Pittosporum hosmeri Rock var. longifolium Rock a * # P. hosmeri Rock var. saint-johnii Sherff a PLANTAGINACEAE Plantago lanceolata L. + + POLYGONACEAE Rumex acetosella L. + + PRIMULACEAE Anagallis arvensis L. C + RANUNCULACEAE Ranunculus muricatus L. c b RHAMNACEAE * # Alphitonia ponderosa Hillebr. a ROSACEAE Fragaria vesca L. forma alba (Ehrh.) Rydb. + + Prunus persica (L.) Batsch + C * Rubus hawaiiensis A. Gray _L * R. macraei A. Gray d R. penetrans L. H. Bailey + + R. rosaefolius Smith + + RUBIACEAE * # Coprosma cymosa Hillebr. C * # C. rhynchocarpa A. Gray + + * # Gouldia ter min alls (Hook, et Arn.) Hillebr. var. antiqua Fosb. forma antiqua c 298 PACIFIC SCIENCE, Vol. XXI, April 1967 SPECIES * #G. terminal! s (Hook. et Arn.) Hillebr. var. anti qua Fosb. forma acuta Fosb. * # G. terminals (Hook. et Arn.) Hillebr. var. konaensis Fosb. forma konaensis * # Psychotria hawaiiensis (A. Gray) Fosb. var. hillebrandii (Rock) Fosb. RUTACEAE * # Pa gar a dipetala (Mann) Engl. var. geminicarpa (Rock) St. John * # F. mauiense (Mann) Engl. var. an ceps (Rock) St. John * # F. mauiense ( Mann ) Engl. var. anceps (Rock) St. John forma petiolulatum (Rock) St. John * # F agar a sp. * # Pelea hawaiiensis Wawra var. gaudichaudii (St. John) Stone * # P. puauluensis St. John * # P. zahlbruckneri Rock * # Pelea sp. * # Pelea sp. * # Pelea sp. SAPINDACEAE * # Dodonaea viscosa (L.) Jacq. var. spathulata (Sm.) Benth. * # Sapindus saponaria L. scrophulariaceae Linaria canadensis (L.) Dumont Veronica plebeia R. Br. V . serpyllifolia L. SOLANACEAE * # Nothocestrum breviflorum A. Gray * # N. longifolium A. Gray KIPUKA KIPUKA KIPUKA KIPUKA PUAULU KI SPECIES PUAULU KI Physalis peruviana L. + + Solanum pseudocapsicum L. + + THYMELAEACEAE * Wilkstroemia phillyreaejolia A. Gray b TROPAEOLACEAE b Tropaeolum majus L. + UMBELLIFERAE Hydrocotyle sibthorpioides + Lam. var. oedipoda + Deg. et Greenwell + URTICACEAE * # Pipturus hawaiiensis Lev. + + * # Urera sandwicensis Wedd. + + VERBENACEAE Verbena litoralis H. B. K. + + b b d + + + d d d + + b + + b d REFERENCES Buck, P. H. 1953. Explorers of the Pacific. Bernice P. Bishop Museum, Special Publ. 43. Cain, S. A., and G. M. de Oliveira Castro. 1959. Manual of Vegetation Analysis. Harper and Brothers, New York. 325 pp. Cline, M. G., et al. 1955. Soil Survey of the Territory of Hawaii. Soil Survey Series 1939, No. 25. 644 pp. Ekern, P. C. 1964. Direct interception of cloud water on Lanaihale, Hawaii. Soil Sci. of Am. Proc. 28(3) :4l9-421. Ellenberg, H. 1956. Grundlagen der Vegeta- tionsgliederung. I. Aufgaben und Methoden der Vegetationskunde. Eugen Ulmer Verlag, Stuttgart. 136 pp. Fagerlund, G. O., and A. L. Mitchell. 1944. A check list of the plants, Hawaii National Park, Kilauea-Mauna Loa Section. Natural History Bull. No. 9. Hawaii Na- I tional Park (mimeographed). Hawaii Natural History Association. 1961. Kipuka Puaulu, Self-guiding Nature Trail. Hawaii Volcanoes National Park, Na- tional Park Service. Unnumbered pamphlet. 10 pp. Krajina, V. J. 1963. Biogeoclimatic zones on the Hawaiian Islands. Newsletter of the Hawaiian Botanical Society 2(7) :93— 98. Soil-Vegetation Relationships in Kipukas — Mueller-Dombois and Lamoureux 299 Powers, H. A. 1948. A chronology of the explosive eruptions of Kilauea. Pacif. Sci. 2(4) :278— 292. Rock, J. F. 1913- The Indigenous Trees of the Hawaiian Islands. Published under patron- age, Hawaii. 518 pp. Rock, J. F. undated. Kipuka Puaulu near the Volcano of Kilauea, Hawaii. Extract of un- published manuscript (typescript). Botany Department, University of Hawaii. 5 pp. Smith, W. H., and C. D. Moodie. 1947. Col- lection and preservation of soil profiles. Soil Sci. 64:61-69. Wentworth, C. K. 1938. Ash Formations on the Island of Hawaii. 3rd Spec. Rept. Hawaii Volcano Observatory. Honolulu. 183 pp. ' Manuscript Form. Manuscripts should be typed on one side of standard-size, white bond paper and double-spaced throughout. Pages should be consecu- tively numbered in upper right-hand corner. Sheets should not be fastened together in any way, and should be mailed flat. Inserts should be either typed on separate sheets or pasted on proper page, and point of insertion should be clearly indicated. Original copy and one carbon copy of manuscript should be submitted. The author should retain a car- bon copy. Although due care will be taken, the editors cannot be responsible for loss of manuscripts. Introduction and Summary. It is desirable to state the purpose and scope of the paper in an introductory paragraph and to give a summary of results at the end of the paper. Dictionary Style. It is recommended that authors fol- low capitalization, spelling, compounding, abbrevia- tions, etc., given in Webster’s New International Dic- tionary (unabridged), second edition; or, if desired, the Oxford Dictionary. Abbreviations of titles of pub- lications should, if possible, follow those given in World List of Scientific Periodicals. Footnotes. Footnotes should be used sparingly and never for citing references (see later). When used, footnotes should be consecutively numbered by supe- rior figures throughout the body of the paper. Foot- notes should be typed in the body of the manuscript on a line immediately below the citation, and sepa- rated from the text by lines running across the page. Citations of Printed Sources. All references cited should be listed alphabetically by author at the end of the paper, typed double-spaced. References to books and to papers in periodicals should conform to the following models: Batzo, Roderick L., and J. K. Ripkin. 1849. A Treatise on Pacific Gastropods. Rice and Shipley, Boston, vii 326 pp., 8 figs., 1 map. Crawford, David L. 1920*. New or interesting Psyllidae of the Pacific Coast (Homop.). Proc. Hawaii. Ent. Soc. 4(l):12-l4. 1920^. The sandalwoods of Hawaii. Proc. Hawaii. Ent. Soc. 4(2) : 374-375, 13 pis. In the text, sources should be referred to by author, date, and page, as follows: "It was noted (Rock, 1916:18) that . . .” or "Rock (1916:21-24) says ...” Quotations. Quoted matter of fewer than five printed lines (about 200 characters) should be given in the text in the usual form, using double quote marks. Longer quotations should be set flush with left mar- gin. The author is responsible for the accuracy of quoted material. Numbers. Decimals, measurements, money, percent- ages, time; enumerations in which any figure is 10 or over; and isolated enumerations of 10 and over should be given in Arabic figures, rather than spelled out, except when the number begins a sentence. ILLUSTRATIVE MATTER Only the minimum number of illustrations required to supplement the text will be accepted by the editors. Reproduction costs of illustrations in excess of the number allowed by the editors will be paid by the author. Artwork for illustrations should be 8!/2 x 11 inches or smaller, and it should accompany manuscript, on separate sheets. Often more valuable than a photo- graph is a good line drawing. Figures and Graphs. Copy for figures and graphs should always be drawn large enough to allow for at least one-third reduction by the engraver. Copy should consist of carefully prepared line drawings in one color only, drawn in India ink on plain white draw- ing paper or tracing cloth. Co-ordinate paper with lines in light blue (a color which will not show in a photograph) may be used; but co-ordinates which should appear in the finished graph must be drawn in India ink. If original figures may not be conven- iently submitted with manuscript, duplicate rough sketches or photographic prints may be furnished to aid the editors in their decisions. It is strongly urged that an indication of scale be incorporated as a part of all drawings in which mag- nification and size are critical considerations. Photographs. Photographs should be chosen for clarity in portraying essential information. They should be printed for contrast, on glossy paper, and should be sent unmounted. They should be identified with serial number written in soft pencil on the back to corre- spond with list of captions. Illustrations will be returned to the author. Tables. Tabular matter should be kept to a minimum. Each table, prepared to conform with Pacific Science style, should be typed on a separate page, and its posi- tion indicated on the manuscript. Mathematical Formulas. Complicated formulas cannot be set by the printers. Authors should submit them as illustrations. Captions. Readily identifiable captions for figures, graphs, photographs, and other illustrative matter should be supplied on a separate page. PROOF Proof should be corrected immediately and returned at once to John S. Howe, Jr., assistant to the editors. Authors are reminded that the editors will allow only a minimum number of corrections on galley proof. Additions to the printed text and changes in style and content are not permitted. All queries on proof should be answered. If cor- rected proof is not received within four weeks after being sent to the author, author’s changes cannot be accepted. REPRINTS Reprints or separates should be ordered on the form provided and returned with author’s proof. All correspondence concerning separates must be directed to the printer, Heffernan Press Inc., 35 New Street, Worcester, Massachusetts 01605. The publisher does not keep a reprint file. >5 • H n VOL. XXI JULY 1967 NO. 3 PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION GEORGE PARARAS-CARAYANNIS Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964. Part I. Water Waves AUGUSTINE S. FURUMOTO Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964. Part II. Rayleigh Wave CAROL WRIGHT STEELE Marine Fungi of Hawaiian, Line, and Phoenix Islands L. G. SWARTZ Distribution of Birds in the Bering and Chukchi Seas GARETH J. NELSON Branchial Muscles of Five Eel Families MICHAEL SALMON Acoustical Behavior of Myripristis berndti in Hawaii GARY J. BRUSCA Ecology of Pelagic Amphipoda, I CARL M. BOYD Benthic and Pelagic Habitats of the Red Crab BRYANT T. SATHER and TERENCE A. ROGERS Inorganic Constituents of Muscles and Blood of Katsuwonus pelamis RICHARD J. KREJSA Life History and Migration of Cottus asper DON HUNSAKER II and PAUL BREESE Herpetofauna of the Hawaiian Islands CLAUDE ROGER Euphausia eximia and E. gibboides in the Equatorial Pacific RICHARD D. CAMPBELL Monobrachium parasitum, a One-Tentacled Hydroid / UNIVERSITY OF HAWAII PRESS BOARD OF EDITORS O. A. Bushnell, Editor-in-Chief Department of Microbiology, University of Hawaii John S. Howe, Jr., Assistant to the Editors University of Hawaii Press Gladys E. Baker Department of Botany University of Hawaii Sidney C. Hsiao Department of Zoology University of Hawaii George W. Chu Departments of Microbiology and Zoology University of Hawaii Charles H. Lamoureux Department of Botany University of Hawaii Maxwell S. Doty Department of Botany University of Hawaii John J. Naughton Department of Chemistry University of Hawaii Augustine S. Furumoto Hawaii Institute of Geophysics University of Hawaii Walter R. Steiger Department of Physics University of Hawaii Philip Helfrich Hawaii Institute of Marine Biology University of Hawaii Goro Uehara Department of Agronomy and Soil Science University of Hawaii Robert W. Sparks, Director University of Hawaii Press INFORMATION FOR AUTHORS Contributions to Pacific biological and physical science will be welcomed from authors in all parts of the world. (The fields of anthropology, agriculture, engineering, and medicine are not included.) Manu- scripts may be addressed to the Editor-in-Chief, PACIFIC SCIENCE, University of Hawaii, Honolulu, Hawaii 96822, or to individual members of the Board of Editors. Use of air mail is recommended for all communications. Manuscripts will be acknowledged when received and will be read promptly by members of the Board of Editors or other competent critics. Authors will be notified as soon as possible of the decision reached. Manuscripts of any length may be submitted, but it is suggested that authors inquire concerning possi- bilities of publication of papers of over 30 printed pages before sending their manuscripts. Authors should not overlook the need for good brief papers, presenting results of studies, notes and queries, com- munications to the editor, or other commentary. preparation of manuscript It is requested that authors following the style of Pacific Science described herein and exemplified in the journal. Authors should attempt to conform with the Style Manual for Biological Journals , Am. Inst. Biol. Sci. Washington. ( Continued on inside back cover ) 7 PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION VOL. XXI JULY 1967 NO. 3 Previous issue published April 10, 1967 CONTENTS PAGE A Study of the Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964. Part I. Water Waves. George Pararas-Carayannis .... 301 A Study of the Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964. Part IL Analysis of Rayleigh Wave. Augustine S. Furumoto 311 Fungus Populations in Marine Waters and Coastal Sands of the Hawaiian, Line, and Phoenix Islands. Carol Wright Steele 317 Distribution and Movements of Birds in the Bering and Chukchi Seas. L. G. Swartz 332 Branchial Muscles in Representatives of Five Eel Families. Gareth f. Nelson 348 Acoustical Behavior of the Menpachi, Myripristis berndti, in Hawaii. Michael Salmon 364 The Ecology of Pelagic Amphipoda, I. Species Accounts, Vertical Zonation and Migration of Amphipoda from the Waters off Southern California. Gary ]. Brusca 382 The Benthic and Pelagic Habitats of the Red Crab, Pleuroncodes planipes. Carl M. Boyd 394 Pacific Science is published quarterly by the University of Hawaii Press, in January, April, July, and October. Subscription prices: institutional, $10.00 a year, single copy, $3.00; individual, $5.00 a year, single copy, $1.25. Check or money order payable to University of Hawaii should be sent to University of Hawaii Press, 535 Ward Avenue, Honolulu, Hawaii 96814, U. S. A. Printed by Heffernan Press Inc., 35 New Street, Worcester, Massachusetts 01605. SMITHSONIAN WfeiTiON M 1 4 196? ? CONTENTS ( continued ) PAGE Inorganic Constituents of the Muscles and Blood of the Oceanic Skip- jack, Katsuwonus pelamis. Bryant T. Sat her and Terence A. Rogers ... 404 The Systematics of the Prickly Sculpin, Cottus asper Richardson, a Polytypic Species. Part II. Studies on the Life History, with Especial Reference to Migration. Richard f. Krejsa 4l4 Herpetofauna of the Hawaiian Islands. Don Hunsaker II and Paul Breese 423 I NOTES Note on the Distribution of Euphausia eximia and E. gibboides in the Equatorial Pacific. Claude Roger 429 Monobrachium parasitum, a ver Island. Richard D. One-Tentacled Hy droid, Collected at Vancou- Campbell 431 A Study of the Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964 Part I. Water Waves1 George Pararas-Carayannis ABSTRACT: The geologic history and the general geomorphology of the area affected by the March 27, 1964 Alaska earthquake are given. The tsunami-generat- ing area is determined and the extent of crustal displacement and the limits of the areas of subsidence and uplift, as revealed by geologic evidence, are discussed. The dimensions of this tsunami-generating area, its volume of crustal displacement, and the energy associated with the tsunami are calculated. Wave activity within and outside the generating area and the possible generating mechanisms for the tsunami are discussed. A wave refraction diagram of the Alaska tsunami for the north Pacific Ocean area is presented in The Aleutian Island Arc and the Aleutian Trench extend for 2800 km from Kamchatka to south-central Alaska along remarkably smooth curves which are convex toward the south (Fig. 1). The Arc forms the Alaska Peninsula and, according to Wilson (1954), intersects, north of Cook Inlet, a second tectonic arc that ex- tends northward from the vicinity of the Wrangell Mountains. However, Plafker (1965) regards this second segment as a continuation of the Aleutian Arc. Where the trench im- pinges on Alaska it loses its identity, although an offshore range of seamounts suggests it may once have extended around to the south to parallel the continental slope, as postulated by Menard and Dietz (1951). Concavity in the former shape of the trench on its eastern seg- ment is also suggested by the sedimentary arc defined by Wilson (1954), which embraces Kodiak Island and the Kenai Peninsula. As shown by Wilson, such concavity is to be ex- pected where two arcs meet at an acute angle, as is well exemplified where the Aleutian and Kuril-Kamchatka arcs intersect. It is also quite possible that large horizontal movements of crustal blocks have helped to change the shape of the Trench and Arc on their eastern seg- ments. However, no such evidence was found in a field study following the Good Friday earthquake (Berg et ak, in preparation). 1 Hawaii Institute of Geophysics Contribution No. 184. Manuscript received June 22, 1966. Figure 6. The nature of the termination of the eastern segment of the Aleutian Trench is obscured by thick sediments washed in from the conti- nental shelf against which it abuts offshore from Cape Suckling. The sediments are of geo- synclinal-dimensions in the sedimentary arc on Kodiak Island (Menard and Dietz, 1951) and as shown by drilling on the Kenai Penin- sula. Woollard et al. (I960) show there is geophysical evidence for at least 7 km of sedi- ments in Cook Inlet, a graben separating the primary arc from the offshore sedimentary arc. Sediment is about 2 km thick off Kodiak Island along the Aleutian Trench, thinning out to about 0.7 km south of Unimak Island in the deep water area, according to seismic measure- ments by Shor (1962). THE GENERATING AREA OF THE ALASKA TSUNAMI According to Van Dorn (1964), the tectonic dislocations associated with the Alaska earth- quake of March 27, 1964 ranged over a dis- tance of 800 km, from the upper portion of Prince William Sound to southwest of the Trinity Islands. The dislocations follow a di- pole pattern of positive and negative displace- ments on either side of a zero-line which, intersecting the east coast of Kodiak Island, continues northeast to the western side of Prince William Sound. There, changing direction, it 301 302 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 1. Generating area of the Alaska tsunami. Crossbatched area indicates ( — ) area of subsidence and ( + ) area of uplift. Heavy dashed lines indicate the backward-refracted wave fronts. Solid line marked by a zero is the axis of rotation (no elevation change). Other solid lines indicate tectonic axes. runs east along the upper part of the sound. The line roughly parallels the Aleutian Trench axis and separates the Kodiak geosyncline from the shelf geanticline. The areas north and west of this line have undergone negative elevation changes, whereas the east and south underwent positive changes. An extensive pattern of positive surface dis- locations under the sea is suspected to lie east of the island of Kodiak and along the conti- nental shelf bordering the Gulf of Alaska. The extent of these dislocations still needs to be confirmed by detailed bathymetric surveys of the area, although large positive displace- ments have been observed as far south as Middleton Island and southwest to Sitkinak Island. Wave refraction studies, described here, also strongly indicated that the tsunami- generating area was mainly in the belt of uplift and included a large segment of the conti- nental shelf and slope. The zone between the known areas of tec- tonic uplift closely corresponds to a major crustal fault defined by crustal seismic mea- surements conducted by the Department of Terrestrial Magnetism of the Carnegie Insti- tution of Washington (Woollard et al., I960). In view of the shallowness of the earthquake (20 km), it was concluded that the crustal dislocations occurred alongside a zone of tilt- ing or a surface rupture (Grantz et al., 1964), but a survey of the area failed to identify such a feature. The focal depth corresponds, how- ever, to the base of the granitic layer defined by Woollard’s analysis of the crustal measure- ments made by the Carnegie Institution. The total area of tectonic displacements asso- ciated with the Alaska earthquake of March 27, 1964 is estimated to be approximately 215,000 km2. This is the largest area known to be associated with a single earthquake within historic time. The magnitude of the Alaska earthquake was estimated to be from 8.4 to 8.75, which Alaska Earthquake and Tsunami, I — -Pararas-Carayannis 303 is greater than the 1906 San Francisco earth- quake (8.3), and equal to or greater than the I960 Chile earthquake (8.4). The epicenter of the earthquake was at 61.05° N, 147.7°W (USCGS, 1964), near the east shore of Una- kwik Inlet in northern Prince William Sound. Geological investigations have defined the land areas affected by the earthquake. To the east, the zone of deformation appears to die out between the Bering Glacier and Cape Yakataga. The northwestern limit of tectonic changes extends at least to the west side of Shelikof Strait and Cook Inlet (Plafker, 1965). The north inland limit is known only along the highway connecting Valdez and Fairbanks; it appears to extend in a northeasterly direction to the vicinity of the Wrangell Mountains, and quite possibly into the Alaska Range. The area of uplift covers about 105,000 km2 and extends from southern Kodiak Island north- east to Prince William Sound. It includes the southern and eastern parts of Prince William Sound, the coastal area as far east as the Bering Glacier, and the continental shelf and part of the slope to a depth contour of approximately 200 m. The maximum uplift on land was 10 m at the southwest end of Montague Island, but is suspected to have been considerably more off- shore. Uplift also occurred along the extreme southeastern coasts of Kodiak Island and Sitka- lidak Island, and part or all of Sitkinak Island. The maximum measured uplift of Sitkalidak Island was 0.4 m. The estimated uplift of Sit- kinak Island was from 0.35 to 0.65 m and pos- sibly as much as 1.5 m (Plafker, 1965). The area that subsided included the northern and western parts of Prince William Sound, the western segment of the Chugach Mountains, portions of the lowlands north of them, most of the Kenai Peninsula, and almost all of the Kodiak Island group. This area of subsidence covers approximately 110,000 km2, and is 800 km long and 150 km wide. Plafker (1965) estimates that the volume of crust that has been depressed below its pre-earthquake level is about 115 km3. The seaward limits of the earthquake and the tsunami-generating area were determined by means of a series of refraction diagrams based on Snell’s Law of Refraction using the velocity equation for shallow water waves, C = \/gd. Such a method of preparing refraction diagrams has shown good results, especially if carried out on large-scale charts with detailed bathymetry (Johnson, O’Brien, and Isaacs, 1948). In constructing the refraction diagrams for the Alaska tsunami, the marigrams of different tide gauge stations around the Pacific were con- sulted and the total travel time of the first wave at each station was determined. Then refraction diagrams were constructed toward the earth- quake area from each tide gauge station in lengths of time equal to the calculated travel time for that station. It was assumed that the last wave front in each refraction diagram would correspond to a point on the boundary of the generating area, and if enough refracted wave fronts from different stations were plotted, an envelope defining the tsunami-generating area could be drawn. Wave fronts were refracted from Yakatat, Cape Yakataga, Seward, Uzinki, Kodiak, Old Harbor, Unalaska, Adak, Attu, and Honolulu. The last front of each of the refracted waves is shown by a heavy dashed line in Figure 1. The seaward boundary of the generating area is near the 200-m depth contour which defines the edge of the continental shelf. Maximum displacement of the ocean floor occurred along the continental shelf, from an area southeast of Kodiak Island, to an area close to Cape St. Elias south of the island of Kayak (Fig. 1). Geologic evidence, however, has shown positive land displacements as far north as Cape Suckling and as far east as the Bering Glacier. It is quite probable, there- fore, that the tsunami-generating area extended farther to the northeast, although waves gener- ated in such shallow water would reach tide gauges much later and their origin would not be identifiable. Unfortunately, this same wave refraction tech- nique could not be used to define the northern and western boundaries of the main tsunami- generating area, because conditions in Prince William Sound and elsewhere along the coast of Alaska were further complicated by local tsunamis, oscillations, and surge. In addition, no tide gauge stations were operating in the area, and personal accounts were conflicting as to arrival times of the different waves. The northward limit is assumed to be re- stricted by the land boundaries, and the western 304 PACIFIC SCIENCE, Vol. XXI, July 1967 limit to extend to the west side of Shelikof Strait and Cook Inlet. In estimating the travel time of the tsunami, corrections were made for the delay at the is- land of Kodiak in the arrival of the ground shocks from Prince William Sound. These cor- rections ranged from 1 minute to 6 minutes and were based on the fact that the Navy Weather Central on the island of Kodiak listed the time of the principal shock in Prince William Sound as 6 minutes later than the time listed by the U. S. Coast and Geodetic Survey. This would imply that the wave front generated on the northeast side of the disturbance area had a 6- minute head start on the wave front generated southeast of Kodiak. The tsunami-generating area covers an area 700 km long by 150 km wide, a total of about 105,000 km2. The volume of the uplifted crust along the continental shelf is about 96 km3. The energy associated with the tsunami has been estimated by Van Dorn (1964) to be of the order of 2.3 X 1021 ergs. This estimate is based on the source dimensions of an area 240 nau- tical miles by 100 nautical miles and an uplift of 1.8 m (6 ft) at the northeastern end of this area and zero at the southwestern end. This estimate, however, is considered low because the generating area had dimensions that were larger than those estimated by Van Dorn. Using our source dimensions, and assuming that the total energy was equal to the potential energy of the uplifted volume of water, the total energy for the tsunami in the Gulf of Alaska was calculated as follows: Et = \gh2A 6 1 = — (1-03) (.980) (103) (104) (1.832) (1.5 6 X 107) (7 X 107) = 5.88 X 1021 ergs where Et = Ep = total energy p — 1.03 g/cm3 = density g = 980 cm/sec2 h = height of displacement =1.83 m A = area 1 erg = g cm2 sec-2 The waves generated in the Gulf of Alaska were of an unusually long period, on the order of an hour or more. Their energy radiation was preferentially directed toward the southeast and this is why more damage was done to the North American coast than anywhere else east or south of the generating area. This preferential direc- tivity of energy radiation can be attributed to the orientation of the tectonic displacements along the continental shelf of the Gulf of Alaska, and the long period of the waves can be related to the long seiche period of the shallow shelf. According to Japanese seismologists (Iida, 1958), the generating area of a tsunami roughly corresponds to the distribution of the major aftershocks. This appears to be indeed the case in the Gulf of Alaska. There were 52 aftershocks of the Alaska earthquake. The largest had a magnitude of 6.7. The aftershocks occurred in an area from about 15 km north of Valdez to about 55 km south of Trinity Islands, and were heavily concentrated on the northeast and the southwest of the up- lifted region (USCGS, 1964), which also was the main tsunami-generating area. The vast area of tectonic movements indicates that wave crests were generated along one or more line sources from the region of maximum uplift. Thus, the shores of the Kenai Peninsula were struck within 20 minutes after the start of the earthquake, and those of Kodiak Island, within 34 minutes. Unfortunately, the violence of the earthquake j left south-central Alaska without a tide gauge | in operation. The only reliable record from the | generating area is the one that was obtained by personnel of the U. S. Navy Fleet Weather Station at Kodiak; it is shown in Figure 2. This | record has been corrected for the 1.7-m (5.6-ft) submergence of the area. Outside the immediate generating area, the record of Cape Yakataga, as constructed from the personal account of C. R. Bilderback, a resi- dent of the area, is the next most reliable record, j This record is the only one obtained outside the generating area that shows an initial drop in the water level (Berg et al., in preparation). Withdrawal of the water immediately following the earthquake has been reported from Kayak, Middleton, and Hinchinbrook islands, as well as from Rocky Bay and Nuka Bay, at the end Alaska Earthquake and Tsunami, I — Pararas-Carayannis 305 Fig. 2. Diagram of wave activity at Women’s Bay, Kodiak Island. (From visual observations made at Marginal Pier, Nyman Peninsula.) of the Kenai Peninsula, but these islands are inside the generating area. Yakatat, a coastal town 170 km southeast of Cape Yakataga, had a tide gauge in operation, and the marigram shows that a positive wave arrived first (Fig. 3). It is quite possible, therefore, that the first waves to arrive at Cape Yakataga had a differ- ent origin from that of the first waves to arrive at Yakatat. It could very well be that the Cape Yakataga waves traveled over the shallow por- tion of the shelf, whereas the Yakatat waves came from the open ocean. An interesting aspect of these two records is that of the difference in amplitude and period of the first waves to arrive at these two sites — which also supports the hypothesis of difference in origin (see Figs. 3 and 4). TSUNAMI GENERATED IN PRINCE WILLIAM SOUND The shallow continental shelf and the islands bordering the southern side of Prince William Sound, as well as the pattern of crustal displace- ments, confined the waves generated in this area to the Sound itself; very little energy escaped this closed region. Most of the energy was ex- pended in the narrow, deep fjords of the Sound, creating catastrophic waves and setting up reso- nating oscillations and surges that lasted for hours. In certain places maximum inundation occurred 5 or 6 hours later, at high tide. At Valdez, for example, the third wave came in at 2300, March 27, and the fourth one at 0145, March 28 (Brown, 1964). This last wave took the form of a tidal bore and inundated the 306 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 3- Marigram of wave activity at the town of Yakutat. downtown section of Valdez, ruining almost all the merchandise in the stores. These waves could not have come from the generating area outside Prince William Sound because if this were so, it would have taken them only 34 minutes to reach Valdez. It is more likely, then, that the waves at Valdez arrived in resonance at high tide, from the immediate area of Port Valdez. Maximum positive crustal displacement in Prince William Sound occurred along the north- west coast of Montague Island and in the area offshore. These earth movements caused a gra- dient in hydrostatic level and the resulting short-period wave raced through Knight Island Passage within 10 minutes and on toward Che- nega Island, inundating the village of Chenega to an elevation of 15.5 m and completely de- stroying it. This same wave continued north through Knight Island Passage and inundated Perry and Naked islands, but to lesser heights (Berg et ah, in preparation). Bathymetric surveys by the USCGS (1964) in the area off Montague Island and at the north end of Latouche Island revealed a num- ber of large submarine slides. It is possible, therefore, that the combination of submarine slides and the tilting of the ocean floor due to uplift created the solitary wave reported at Chenega village and at Perry and Naked islands. A second wave about 40 m high (125 ft) was reported coming out of the Valdez Narrows Alaska Earthquake and Tsunami, I — Pararas-Carayannis 307 and spreading across the Sound (Plafker and Mayo, 1965). This wave was caused by slump- ing of the glacial deltas in Port Valdez which had been shaken loose by the force of the earthquake. TSUNAMI MECHANISM Most tsunamis result from earthquakes hav- ing focal depths of less than 60 km. Iida (1958) has derived an empirical relation giv- ing the maximum focal depth H (in km) for an earthquake of magnitude M which has re- sulted in a detectable tsunami: M i§ 6.42 + 0.01 H (1) where M is the Richter magnitude given by log E(ergs) = 11.8 -f- 1.5 M (2) The focal depth of the Alaska earthquake was about 20 km. This was shallow enough to create tsunami waves even though the epicenter of the main shock was as much as 100 km inland from the coast. A number of shallower aftershocks over a large area ranging from Hinchinbrook Island to southeast Kodiak Island indicate that crustal movements over a wide area were in- volved. Undoubtedly these shallow aftershocks created smaller waves that could not be sepa- rated, in the tide gauge records, from reflections of the initial tsunami. If the tsunami waves that hit the island of Kodiak were the result of crustal movements only, then the first wave could be expected to 308 PACIFIC SCIENCE, Vol. XXI, July 1967 be the highest, at least within the generating area. At Uzinki, Kodiak City, Women’s Bay, and elsewhere on the island of Kodiak, how- ever, the third and fourth waves were the high- est. A theory of generation from a single pattern of crustal deformation is therefore not satisfactory here. Such factors as reflection from coastal boundaries, wave interaction, and reso- nance should be taken into consideration. Slumps or avalanches, similar to the ones that occurred in Prince William Sound, are usually localized; they can produce no large tsunamis that would travel across wide portions of the ocean. According to Wiegel (1954), not more than 2% of the potential energy of a falling or sliding body is converted into wave energy. In Prince William Sound, however, slumping and sliding when added to tectonic movements cre- ated tsunami waves of very large energy, but their effect was catastrophic only locally; very little of the energy escaped the Sound. SUMMARY AND CONCLUSIONS The Alaska earthquake of March 27, 1964 affected an area of approximately 215,000 km2, extending from the Wrangell Mountains at the northeast to the Trinity Islands in the south- west, and from the west side of Shelikof Strait and Cook Inlet east to the vicinity of the Bering Glacier. Geologic evidence has revealed a dipole pat- tern of positive and negative tectonic move- j ments resulting from this earthquake. The area of subsidence covers approximately 110,000 km2 and the volume of crust that has been depressed below its pre-earthquake level is about 115 km3. The area of uplift covers about 105,000 km2 and includes the southern and eastern parts of Prince William Sound, the coastal area as far east as the Bering Glacier, and a great part of the continental shelf and slope bordering the Gulf of Alaska. The seaward limits of the area affected by the Alaska earthquake and the tsunami-generat- ing area were determined by means of a series of wave refraction diagrams as shown in Figure 5, based on Snell’s Law of Refraction. The tsunami-generating area covers 140,000 km2 and includes the whole of the region of uplift and part of the region of subsidence. It extends from the Trinity Islands to the Bering Glacier and includes Shelikof Strait, Cook Inlet, and the continental shelf bordering the Gulf of Alaska to a depth of approximately 200 m. The total volume of displaced material in the tsu- nami-generating area was estimated to be 120 km3, and the energy associated with the tsunami was calculated to be in the order of 6 X 1021 ergs. As a result of this work the following con- clusions are drawn: Fig. 5. Diagram of wave fronts refracted toward the earthquake area from Attu Island ( dashed line), Adak Island ( solid line), and Unalaska Island ( dotted line). Alaska Earthquake and Tsunami, I — Pararas-Carayannis 309 1. Two main tsunami-generating areas can be distinguished : one along the continental shelf bordering the Gulf of Alaska; the other in Prince William Sound. 2. The main generating area in the Gulf of Alaska roughly corresponds to the geographic distribution of the major aftershocks. 3. The energy of the tsunamis generated in Prince William Sound was expended inside the Sound; not much energy escaped this closed region. 4. The long period of the waves generated in the Gulf of Alaska is related to the long seiche period of the shallow shelf. 5. The preferential radiation of energy to- ward the southeast is attributed to the orienta- tion of the tectonic displacements along the continental shelf of the Gulf of Alaska. 6. The waves arriving at Cape Yakataga had their origin in the shallow coastal area near the Bering Glacier, whereas the waves arriving at Yakatat traveled through the deeper waters. 7. In Prince William Sound two major tsu- namis were distinguished: one had its origin near the west coast of Montague Island, the other originated in the Port of Valdez. 8. Two types of tsunami-generating mecha- nisms were associated with the Alaska earth- quake: (a) waves generated directly by tectonic movements of the sea floor, and (b) waves generated indirectly from landslides, mudflows, and slumping of alluvial deposits. 9. In Prince William Sound both generation mechanisms were evident, while in the generat- ing area along the Gulf of Alaska, the generated tsunami was the direct result of tectonic move- ments. ACKNOWLEDGMENTS The work on which this paper is based was supported in part by the National Science Foun- dation under the United States-Japan program for cooperative research in the Pacific, through grant No. GF-153, and in part by the Office 310 of Naval Research through contract Nonr- 3748(03). I am particularly indebted to D. C. Cox, W. M. Adams, and G. P. Woollard for their advice and constructive criticism. I would also like to acknowledge with appre- ciation the advice, comments, and suggestions given to me by A. S. Furumoto, K. Kajiura, G. W. Groves, H. G. Loomis, G. R. Miller, and Robert Harvey. I also thank the members of the United States-Japan Cooperative Field Survey (E. Berg, D. C. Cox, A. S. Furumoto, K. Kajiura, H. Ka- wasumi, and E. Shima) for permission to use data from their report prior to publication. REFERENCES Berg, E., D. C. Cox, A. S. Furumoto, K. Kajiura, H. Kawasumi, and E. Shima. (In preparation.) Field Survey of the Tsunami of 27 March 1964 in Alaska. Hawaii Inst. Geo- phys. Rept. Series. Brown, D. L. 1964. Tsunamic Activity Accom- panying the Alaskan Earthquake of 27 March 1964. U. S. Army Engr. Dist., Anchorage, Alaska. 20 pp. Grantz, A., G. Plafker, and R. Kacha- doorian. 1964. Alaska’s Good Friday Earth- quake March 27, 1964: A Preliminary Geo- logic Evaluation. U. S. Geol. Surv. Circ. 491. 35 pp. Iida, K. 1958. Magnitude and energy of earth- quakes accompanied by tsunami, and tsunami energy. J. Earth Sci., Nagoya Univ. 6:101- 112. Johnson, J. W., P. O. O’Brien, and J. D. Isaacs. 1948. Graphical construction of wave refraction diagrams. H. O. Publ. No. 605. PACIFIC SCIENCE, Vol. XXI, July 1967 Menard, H. W. 1964. Marine Geology of the Pacific. McGraw-Hill Book Co., New York. Pp. 97-116. and R. S. Dietz. 1951. Submarine ge- ology of the Gulf of Alaska. Bull. Geol. Soc. Am. 62:239-253. Plafker, G. 1965. Tectonic deformation asso- ciated with the 1964 Alaska earthquake. Sci- ence 148:1675-1687. and L. R. Mayo. 1965. Tectonic De- formation, Subaqueous Slides and Destructive Waves Associated with the Alaskan March 27, 1964 Earthquake: An Interim Geologic Evaluation. U. S. Geol. Surv., Open File Rept. 19 pp. Shor, G. G., Jr. 1962. Seismic refraction stud- ies off the coast of Alaska: 1956-57. Bull. Geol. Soc. Am. 52:37-57. U. S. Coast and Geodetic Survey. 1964. Preliminary Report, Prince William Sound, Alaskan Earthquakes; March- April 1964. 83 pp. Van Dorn, G. W. 1964. Source mechanism of the tsunami of March 28, 1964 in Alaska. Chap. 10. In: Proc. Ninth Conference on Coastal Engineering, Am. Soc. Civil Engr., pp. 166— 190. Wiegel, R. L. 1954. Laboratory studies of gravity waves generated by the movement of s| a submerged body. Univ. Calif. Inst. Engr. Res., Ser. 3, Issue 362. Wilson, J. T. 1954. The development and structure of the crust. Chap. 4. In: Gerard P. Kuiper, ed., The Solar System, Vol. 2. The Earth as a Planet. Univ. of Chicago Press. Woollard, G. P., N. A. Ostenso, E. Thiel, and W. E. Bonini. I960. Gravity anomalies, crustal structure, and geology in Alaska. J. Geophys. Res. 65:1021-1037. A Study of the Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964 Part II. Analysis of Rayleigh Wave1 Augustine S. Furumoto ABSTRACT: The source mechanism of the Alaska earthquake of March 27, 1964 has been investigated by analyzing the Rayleigh wave recorded on the strain seis- mograph at Kipapa Station, Hawaii. The parameters that give the best fit to the observed data are: rupture length of 800 km, rupture velocity of 3 km/sec, and direction of rupture line of S30°W. The results of this analysis compare favorably with field data of elevation changes, with distribution of epicenters of aftershocks, and with the area of generation of the tsunami as obtained from sea-wave refraction diagrams. The United States-Japan Cooperative Field Survey of the Alaska Earthquake of March 27, 1964 (Berg et al., in preparation) resulted in an estimate of the length and size of the rupture zone of the earthquake. Corroboration for these results was sought from seismic data. Toksoz et al. (1965) have published a source mecha- nism analysis using surface wave data. Their results are as follows: rupture velocity, 3.0 km/ sec; rupture length, 600 km; azimuth of rup- ture, S50°W from the epicenter. These results, however, are at variance with the field survey data. Shortly after the field survey, an attempt at source mechanism analysis by surface wave methods was made by using the record of the strain seismograph at Kipapa Station, Hawaii. The results of this analysis are presented here because they are in somewhat better accord with field survey data. This study was supported by funds from the National Science Foundation under Grants GP- 2257 and GP-5111. METHOD OF ANALYSIS The analysis of source mechanism based on earthquake surface waves was developed by Ben- Menahem (1961). According to this method, if the Rayleigh wave is used the ratio of the amplitude spectrum of R3 to the amplitude spectrum of R2 can be related to directivity function D(f), D(f) = (v + cos0 (C \v~cose (1) where C is the phase velocity of the curve at frequency f, V is the velocity of rupture propa- gation, B is the length of the rupture, and 6 is the angle which the rupture line makes with the great circle path through the epicenter and observing station. A method using the Love 1 Hawaii Institute of Geophysics Contribution No. 185. Manuscript received June 22, 1966. wave has also been developed, but the present study utilizes the Rayleigh wave only. Ben-Menahem and Toksoz have applied the method of surface wave analysis to the study of the source mechanism for the Mongolian earth- quake of 1958 (Ben-Menahem and Toksoz, 1962) , the Alaska earthquake of 1958 (Ben- Menahem and Toksoz, 1963^), and the Kam- chatka earthquake of 1952 (Ben-Menahem and 311 312 PACIFIC SCIENCE, Vol. XXI, July 1967 Toksoz, 1963^). Wada and Ono (1963) have applied the method for the Chile earthquake of I960. For the Alaska earthquake of 1964, copies of records from the strain seismograph at Kipapa Station, Hawaii, were used. This strain seismo- graph consists of a quartz rod 80 ft long. It was installed by the California Institute of Technology in the spring of 1963. Figure 1 shows the traces of R2, R3, and R4. RESULTS OF ANALYSIS The Fourier spectra of R2, R3, and R4 are given in Figure 2. To form the ratios of am- plitudes R3/R2 and R3/R4, the decay of ampli- tudes with travel distance must be considered because the decay coefficient is frequency-depen- dent. The decay coefficient determined by Ben- Menahem and Toksoz (1963^) from empirical data was used for the corrections. The amplitude ratios of R3/R2 and R3/R4 are given in Figure 3. There is coherence be- tween the two ratio spectra at certain frequen- cies. Troughs of the spectra coincide at 0.0027 cps, 0.0056 cps, 0.0080 cps, and 0.0010 cps. Peaks agree at 0.0088 cps and 0.0111 cps. There is a peak at 0.0038 cps for R3/R4 and a peak Fig. 1. Upper : Phases R2 and G3. Window indi- cates the section of R2 that was used as data. Middle'. Trace of R3. Lower'. Trace of R4 and G5. Fig. 2. Upper: Fourier spectrum of R3. Lower: Fourier spectra of R2 and R4. The amplitude coordi- nate is in arbitrary units. Fig. 3. Directivity function, theoretical and ob- served. The amplitude coordinate is in arbitrary units. For the theoretical curve, V = 3 km/sec, 6 — 15°, and B — 800 km. 152* 150* !48* 146* S44* 142* 313 Alaska Earthquake and Tsunami, II — Furumoto Fig. 4. The rupture line and distribution of epicenters of aftershocks. 314 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 5. The rupture line and elevation changes. Alaska Earthquake and Tsunami, II — Furumoto 315 at 0.0044 cps for R3/R2. These two peaks prob- ably coincide, and the apparent lag between the two is due to inadequate resolution of the Fou- rier analysis at these frequencies. There are opposing patterns at 0.0068 cps. The best-fitting curve of the directivity func- tion with the R3/R2 spectrum is plotted on the upper section of Figure 3. In this curve the pa- rameters are: B = 800 km, V = 3-0 km/sec, and 0 = 15°. R3/R4 fits the curve also, except for the mismatch in the neighborhood of 0.0067 cps. The epicenter determined by the U. S. Coast and Geodetic Survey (1964) was 61.05 °N, 147. 5°W. The coordinates of the Kipapa Sta- tion are 21°25'24"N and 158°00'54"W. The direction of the station from the epicenter is Sl5.3°W. This defines the direction of the rup- ture line from the epicenter as S30°W. In Figure 4, the rupture line, as obtained from the present study, is superimposed on a map prepared by the U. S. Coast and Geodetic Survey (1964) which shows the epicenters of the main shock and the aftershocks of the Alaska earthquake. In general, the aftershock area defines the area of rupture. In the present case, the rupture line obtained from Rayleigh wave analysis extends 100 km beyond the after- shock area. Surveys of elevation changes after the Alaska earthquake show positive changes in the Prince William Sound area, and negative changes in the Kodiak Island area. In Figure 5, the calcu- lated line of rupture is superimposed on the map of elevation changes as prepared by Pa- raras-Carayannis (see his Fig. 1, on p. 302 of this issue). The rupture line runs diagonally across the section of positive changes. In this calculation the direction of the rupture line may vary about 5°. (This value is determined by the resolving power of the Fourier analysis.) If the direction of the rupture line is turned 5° clockwise, with the epicenter as the pivotal point, the rupture line will agree with the line of zero displacement from field observations. An inspection of the directivity function D(f) in equation (1) shows that the period- icity in terms of frequency of the peaks and troughs of the function is controlled by the length B of the rupture line. The peaks and troughs of R3/R2 and R3/R4 in Figure 3 are such that a length of B = 800 km fits the data best. The superimposition of the rupture line on the elevation-change map shows that the rupture line extends to the south 100 km be- yond the zone of elevation changes. On the other hand, if the total area of the observed ele- vation changes is considered, the zone has a length of 700-800 km (Plafker, 1965). The present analysis shows a discrepancy be- tween the direction of the calculated rupture line and the direction expected from field sur- vey, but the discrepancy is within the limits of error of the calculation. The length of the cal- culated rupture line agrees with that from field data. DISCUSSION The results of the field survey by the United States-Japan Cooperative Team (Berg et al., in preparation) have heavily influenced the anal- ysis presented here since the author was a mem- ber of the survey team. Perhaps because of this bias, the analysis should not be considered as an independent study but, rather, as additional evidence to strengthen the results proposed by the field survey. The rupture zone of the Alaska earthquake of 1964 has now been outlined con- sistently by four different methods: (a) field survey of elevation changes (Berg et al., in preparation; Plafker, 1965); (b) plot of epi- centers of aftershock (U. S. Coast and Geodetic Survey, 1964); (c) tsunami refraction diagrams (Pararas-Carayannis, p. 301-310, in this issue) ; and (d) seismic surface wave method (this paper) . REFERENCES Ben-Menahem, A. 1961. Radiation of seismic surface waves from finite moving sources. Bull. Seism. Soc. Am. 51:401-435. and M. N. Toksoz. 1962. Source mech- anism from spectra of long period seismic surface waves, 1. The Mongolian earthquake of December 4, 1957. J. Geophys. Res. 67:1943-1955. 1963^. Source mechanism from spectra of long period surface waves, 2. The Kam- chatka earthquake of November 4, 1952. Ibid. 68:5207-5222. 1963^. Source mechanism from spectra 31 6 PACIFIC SCIENCE, Vol. XXI, July 1967 of long period seismic surface waves, 3. The Alaska earthquake of July 10, 1958. Bull. Seism. Soc. Am. 53:905-919. Berg, E., D. C. Cox, A. S. Furumoto, K. Kajiura, H. Kawasumi, and E. Shima. (In preparation.) Field Survey of the Tsunami of 28 March 1964 in Alaska. Hawaii Inst. Geo- phys., Rept. Series. Plafker, G. 1965. Tectonic deformation asso- ciated with the 1964 Alaska earthquake. Sci- ence 148:1675-1687. Toksoz, M. N., A. Ben-Menahem, and D. Harkrider. 1965. Source mechanism of Alaska earthquake from long period seismic surface waves. (Abstr.) Trans. Am. Geophys. Union 46:154. U. S. Coast and Geodetic Survey. 1964. Preliminary Report, Prince William Sound, Alaskan Earthquakes, March-April 1964. 83 pp. Wada, T., and H. Ono. 1963. Source mecha- nism of the Chilean earthquake from spectra of long period surface waves. Zisin 16 (ser. II) : 181— 187. Fungus Populations in Marine Waters and Coastal Sands of the Hawaiian, Line, and Phoenix Islands1 Carol Wright Steele2 ABSTRACT: Saprophytic and facultative parasitic fungi present in the coastal waters and adjacent pelagic areas of the Hawaiian Islands, and in coastal sands of the Hawaiian, Line, and Phoenix islands, were isolated by plating methods. Isolates from all areas sampled indicate that abundant and varied fungus popu- lations do exist in these environments. The number of fungi obtained from the inshore neritic zone was seven times that obtained from the oceanic zone. The fungus Aureobasidium pullulans (De Bary) Arnaud was isolated repeatedly from oceanic waters. A comparison is made between the genera and the average number of isolates per liter of water known from the Atlantic Ocean with those found in this study of the Pacific Ocean. The number of fungi isolated from sand samples of the different islands ranged from 2 to 1,600 per gram. Species diversity was evident throughout the samples. The leeward Hawaiian islands had a higher aver- age number of isolates per gram than any other island group. In conclusion the problems of defining a marine fungus are discussed. Oceanic areas in different parts of the world have been shown to be habitats for marine fungi (Johnson and Sparrow, 1961). Investi- gators, however, have usually concentrated on particular groups of fungi by use of selective isolation methods (Barghoorn and Linder, 1944; Moore and Meyers, 1959; Jones, 1962). Only one extensive analysis of marine waters for a general fungus population is known, and it was made in the northwestern subtropical Atlantic Ocean (Roth et al., 1964). References to the occurrence of fungi in the Pacific Ocean are found (1) as incidental to studies of bac- teria in marine water (ZoBell, 1946); (2) in studies of specialized fungi such as lignicolous fungi (Cribb and Cribb, 1955, 1956, I960; Kohlmeyer, I960; Meyers and Reynolds, I960) and those on algae (Cribb and Cribb, 1955, 1956, I960); and (3) in studies of particular kinds of fungi, e.g., Phy corny cetes in Japanese waters (Kobayashi, 1953) and pathogenic spe- cies (Van IJden and Castelo Branco, 1961). Reports of fungi from terrestrial environments of islands in the central and southern Pacific 1 Prepared from a thesis submitted in partial ful- fillment of requirements for the Master of Science degree at the University of Hawaii. 2 Department of Botany, San Diego State College, San Diego, California, 92115. Manuscript received June 20, 1966. also are very limited. These include a few rec- ords of higher fungi collected in the Marshall Islands (Rogers, 1947), the Society Islands (Olive, 1957, 1958), and Raroia in the Tua- motu Archipelago (Cooke, 1961); Phyco- mycetes recovered by plating soils of the atolls of Bikini, Eniwetok, Rongerik, and Ronggelap (Sparrow, 1948); and Ascomy cetes and Fungi Imperfecti from dung and soil samples col- lected by Olive in the Society Islands (Peter- sen, I960). Consequently, as Cooke points out, the geographic distribution of fungi occurring on the islands of the Pacific is poorly known. Although studies have been initiated on the soils of the Hawaiian Islands (Baker, 1964) and the phyllosphere (Marsh, 1965), no study has been made of fungi occurring in marine waters and intertidal environments of these islands or elsewhere in the central Pacific. This investigation was undertaken to determine the occurrence and distribution of the saprophytic and facultative parasitic fungi which constitute the fungal populations of these habitats. MATERIALS AND METHODS Collection Isolations for fungi were made from 59 water samples and 50 sand samples collected 317 318 from inshore areas of the Hawaiian Islands and adjacent pelagic environments. Seventeen sand samples were taken from the Phoenix and Line islands. The water samples were taken from four zones: the surf or spray zone; the inshore neritic zone, 2-200 m from shore; the offshore neritic zone, 300 m to 2 km from shore; and the oceanic zone, 2 km or more off shore. Of these water samples 56 were taken at the surface and 3 were taken at depths down to 600 m. The sand samples were taken from a depth of 2 inches in the supra- tidal, intertidal, and subtidal zones as delimited by Hedgpeth (1957). The sample sites were selected to include a variety of shore environ- ments: leeward, windward, or areas of special interest (e.g., South Point, Hawaii, the south- ernmost point in the Hawaiian Islands; Wai- kiki Beach, probably the most frequently used beach on Oahu; and Midway Island, the north- ernmost inhabited island of the Hawaiian island chain.) Other samples were obtained as oppor- tunity offered: those from Kure Island, the Phoenix Islands, the Line Islands, and the open ocean. Figure 1 gives the geographical location of the water and sand collection sites. The surface water samples were collected in sterile 600-ml glass bottles with plastic screw caps. The closed bottles were immersed in the water, then opened and allowed to fill with water, closed underwater and brought to the surface. The depth samples were taken by send- ing a plastic water sampler of the van Dorn (1956) type to the designated depths. After the sampler was brought to the deck of the ship, a sterile 600-ml bottle was filled with water from the sampler. All sand samples were collected with sterile implements and placed in sterile 100-ml jars with plastic screw caps or in sterile poly- ethylene bags (nasco Whirl-Pak, Hydro Prod- ucts Co., San Diego, California). Salinity measurements were obtained by use of Quantabs SO 51 (Linayer Corp., Detroit, Michigan). Temperature was determined by standard centigrade thermometer, and depth by standard oceanographic methods. Isolation Two principle means of isolation were em- ployed: the pour-plate method (Salle, 1954) PACIFIC SCIENCE, Vol. XXI, July 1967 and the millipore filter method (Roth et al., 1964). Standard materials and procedures were used in the pour-plate method. Amounts plated for the water samples ranged from 0.5 ml to 5.0 ml. Dilutions of 1:10 to 1:100,000 were used for sand samples. Some samples were plated upon return to the laboratory; because of the distance between most sampling sites and the laboratory, however, most of the plating was done 24 to 48 hours after collection. All sam- ples were kept under refrigeration until plated. A control plate, uninoculated, was set up for each medium for every plating. Plates were also exposed to the air during plating opera- tions to determine the level of air contamina- tion in the laboratory. Materials for the millipore filter method included sterile cellulose-ester membranes with 0.45 [x porosity (Millipore Filter Corp., Bed- ford, Massachusetts). The samples were run through the millipore filter apparatus using a vacuum pump. The membranes with retained fungal elements were cultured on selective me- dia in presterilized pastic petri dishes. The amount put through the filter ranged from 100 ml to 600 ml per sample. Controls were included for testing agar sterility and air con- tamination. Several kinds of media were used: sodium caseinate agar (BBL 01-549, Fred and Waks- man, 1928, modified by Potter, 1957), Roth’s isolation medium (Roth et ah, 1964), Fell’s yeast agar (Fell et al., I960) and mycobiotic agar (difco 0689-02). All media except the mycobiotic agar were made with sea water col- lected at the sample sites. Bacterial growth was controlled by the incorporation of 0.05% chlor- amphenicol (Chloromycetin, Steri-Vial No. 65, Parke, Davis and Co.). Dilutions were made with Mcllvaine’s buffer solution, pH 7.0 (Machlis and Torrey, 1956). The mycobiotic agar was made with 250 ml sea water and 750 ml of distilled water in order to retain the selectivity of this medium in which Chloro- mycetin is already incorporated. The plates were placed in paper bags, sealed with adhesive tape, and incubated at 20°- 24°C. Pour-plates were incubated for three weeks and the millipore filter plates for 10 days at room temperature (20°-24°C). Marine Fungi from Central Pacific — Steele 319 Fig. 1. General geographical location of marine water and coastal sand collection sites, x, Marine water collection sites; o, coastal sand collection sites. 320 PACIFIC SCIENCE, Vol. XXI, July 1967 Other methods of isolation, such as baiting (Johnson and Sparrow, 1961), spread-plate technique (Buck and Cleverdon, I960) and cellulose plates made by placing a piece of sterile filter paper over an agar plate before inoculating (Baker, 1964) were attempted on some samples. The pour-plate and millipore filter methods, however, proved superior for this study. Identification Some fungi were identified directly from the isolation plates by a technique employing pressure sensitive tape (clear acetate tape, Scotch No. 800, Minnesota Mining and Manu- facturing Co.). This method is described by Roth, Orpurt, and Ahearn (1964). Other fungi were transferred to various selective me- dia to encourage sporulation. Those used most successfully were Czapek-Dox agar (difco 0339-01) and "V-8” juice agar (Wickerham et ah, 1946). RESULTS Populations in Water Variations in salinity and temperature for all water sample sites was slight, ranging be- tween 30 %o and 35 %o for salinity; between 20° and 25 °C for temperature. Inasmuch as sampling was done over the period from June 1964 to May 1965 this temperature range also reflects the slight variation characteristic of a tropical climate. All sample sites yielded fungus colonies (Table 1 ) ; 44 of the samples had counts of 50 colonies or under, whereas only 15 gave counts over 50. The standard plate count was TABLE l Zonal Distribution of Fungi in Marine Waters and Coastal Sands AVERAGE AVERAGE NUMBER NUMBER NUMBER NUMBER OF ISOLATES OF ISOLATES WATER SAMPLES SAMPLES PER ML SAND SAMPLES SAMPLES PER GM Main Hawaiian islands Main Hawaiian islands Kauai Kauai Surf zone 2 .16 Intertidal zone 5 218 Oahu Oahu Surf zone 9 .10 Supratidal zone 3 185 Inshore neritic zone 9 .83 Intertidal zone 3 6 Inshore bay zone 3 .28 Tidal pool zone 2 72 Offshore neritic zone 3 .61 Subtidal zone 3 14 Polluted zone 3 3.18 Maui Maui Intertidal zone 1 36 Surf zone 1 .17 Hawaii Hawaii Intertidal zone 6 77 Surf zone 3 .09 Leeward Hawaiian islands Inshore neritic zone 5 .43 Kure Leeward Hawaiian islands Intertidal zone 8 41 Midway Midway Surf zone 1 .20 Supratidal zone 1 1000 Inshore neritic zone 4 .031 Intertidal zone 6 790 Oceanic, Johnston Island 9 .066 Other leeward Oceanic, Oahu 7 .029 Hawaiian islands Intertidal zone 12 1220 Surf zone total 16 .12 Line Islands Inshore neritic zone total 18 .34 Intertidal zone 4 150 Inshore bay zone total 3 .28 Phoenix Islands Offshore neritic zone total 3 .61 Intertidal zone 13 80 Oceanic zone total 16 .045 Polluted zone total 3 3.18 59 .14 total 67 grand total Marine Fungi from Central Pacific — Steele not significant for the dilutions used. There- fore, the number of isolates from the total number of milliliters used in plating is re- corded as an average isolation return per ml of water samples. The average number of iso- lates from any one sample ranged from 0.06 to 3.94 isolates per ml, with the average for samples at 0.14. The number of species ranged from 1 to 17 per sample. More than 50% of the sites, however, returned only 2 to 7 dif- ferent species (Steele, 1965). The predominant genera and species by percentage of occur- rence are listed in Table 2. Table 3 lists the 126 species of fungi repre- senting 59 genera which were isolated from the water samples plated. The percentage of occurrence represents the number of water sam- ples in which a particular fungus occurred in reference to the total number of samples ana- lyzed. A tabulation of species isolated from the six zones sampled shows that they can be ranked in descending order for number of species per zone as follows: inshore neritic, 70; surf, 56; polluted zone, 28; oceanic, off John- ston Island, 23; oceanic, off Oahu, 20; and offshore neritic, 13. The inshore neritic zone was the richest area, having a higher average 321 number of isolates than either the surf or oceanic zones. A very low average number of isolates was obtained from both oceanic regions. The offshore oceanic area near Johnston Island returned more isolates than the comparable zone near Oahu, but both areas had about the same number of species. Aureobasidium pullulans and Rhodotorula spp. were common to both oceanic sites. These fungi were among those predominant in all isolations from water (Table 2). As might be expected, samples from the polluted areas off Oahu had the highest aver- age number of fungi: 3.18 per ml. This was an area of diverse speciation. Members of the Sphaeropsidales were common, as were species of Aspergillus, Penicillium, and Cephalo- sp or turn. Populations in Sand From 67 sand samples plated from four dif- ferent zones, 134 species of fungi representing 71 genera were recovered (Table 3). The fre- quency of predominant isolates by species is given as percentage of occurrence (Table 2). The average number of isolates per gm, ob- tained by the standard dilution plate counting TABLE 2 Predominant Genera and Species in Water and Samples WATER PERCENTAGE * OF OCCURRENCE SAND PERCENTAGE OF OCCURRENCE Yeasts 45.8 Aspergillus wentii 50.7 Rhodotorula spp. 27.1 Fusarium spp. 44.7 Fusarium spp. 22.0 Phialophora spp. 25.3 Cephalosporium curtipes 22.0 Penicillium spp. 22.3 Cladosporium cladosporioides 16.9 Aspergillus niger 20.8 C. epiphyllum 16.9 Y easts 20.8 Helminihosporium anomalum 16.9 Me gas ter sp. 17.6 Trichoderma lignorum 15.2 Masoniella grisea 16.4 Aspergillus niger 13.5 Aspergillus spp. 14.9 A. went ii 13.5 A. terreus 13.4 Aureobasidium pullulans 13.5 A. ustus 11.9 Phoma spp. 11.8 Trichoderma lignorum 11.9 Aspergillus spp. 11.8 Cladosporium cladosporioides 11.9 Black yeasts 11.8 C. epiphyllum 11.9 Pestalotia spp. 10.1 Cephalosporium roseo-griseum 11.9 Cladosporium herbarum 10.1 C. spp. 11.9 Aspergillus versicolor 10.1 C. acremonium 10.4 Penicillium spp. 10.1 C. curtipes 10.4 Penicillium lilacinum 10.4 * The percentage of occurrence represents the number of water or sand samples in which a particular fungus occurred in reference to the total of 59 water samples or 67 sand samples analyzed (Orpurt, 1964). 322 PACIFIC SCIENCE, Vol. XXI, July 1967 TABLE 3 Genera and Species Isolated From Marine Waters and Coastal Sands PERCENTAGE PERCENTAGE OCCURRENCE OCCURRENCE IN WATER IN SAND NAME SAMPLES SAMPLES PHYCOMYCETES Mucorales Cunningloamella echinulata Thaxter 1.4 C. sp. 1.4 Mucor globosus Fischer 1.4 Rhizopus nigricans Ehrenberg 1.6 2.9 Syncephalastrum racemosum (Cohn) Schroeter 1.6 1.4 ASCOMYCETES Chaeiomium olivaceum Cooke and Ellis 1.4 C. sp. 1.6 Melanomma sp. 1.4 Melanospora lagenaria (Pers.) Fuckel 1.4 M. sp. 1.4 Microascus intermedins Emmons and Dodge 1.4 M. trigonosporus Emmons and Dodge 1.4 Neurospora sp. 5.0 2.9 Sporormia sp. 2.9 BASIDIOMYCETES Sp. indet. 1.6 DEUTEROMYCETES Sphaeropsidales Amerosporium sp. 1.4 Aposphaeria sp. 5.0* Coniothyrium juckelii Sacc. 1.6 Cytosporina sp. 3.3 Diplodia sp. 1.4 Diplodina sp. 1.6 Macrophoma sp. 1.6 Peyronellaea sp. 3.3 Phoma hibernica Grimes 3.3 2.9 Pboma spp. 11.8 7.1+ Phomopsis sp. 1.6 Phyllosticta sp. 1.6 Piggotia sp. 1.6 Pyrenochaeta sp. 1.6 1.4+ Sphaeronaema spinella Kalchb. 1.6 Sporonema sp. 1.4 Melanconiales Pestalotia sp. 10.1 5.9+ Phylactaena sp. 1.6 Moniliales Sporobolomycetaceae Sporobolomyces sp. 2.9+ Moniliaceae Acremonium sp. 1.4+ Acrostalagmus cinnabarinus Corda 1.6 1.4 Aleurisma carnis (Brooks and Hansford) Bisby 1.4+ Allescheriella crocea (Mont.) Hughes 1.6 Aspergillus amstelodami (Mangin) Thom 1.4+ Marine Fungi from Central Pacific — Steele 323 TABLE 3 ( continued ) PERCENTAGE PERCENTAGE OCCURRENCE OCCURRENCE IN WATER IN SAND NAME SAMPLES SAMPLES A. mespiiosus Raper and Thom 1.4 A. Candidas Link 4 At A. came us (van Tiegh) Bloch. 4 At A. clavatus Desm. 2.9X A. effusus Tiraboschi 1 At A. flavipes (Bainier and Sartory) Thom and Church 3.3 2.9 A. flavus Link 1.6* 1.4 A. fumi galas Fres. 1.6 4At A. granulosus Raper and Thom 7.1 A. itaconicus Kinoshita 1.6* A. janus Raper and Thom 1.6 4.4 A. luchuensis Inui 1.6 2.9 A. micro-virido-citrinus Cost, and Lucet 1.6 A. nidulans (Eidam) Wint. 1.4 A. niger van Tiegh. 13.5* 20.8+ A. niveus Bloch. 1.4 A. ochraceus Wilhelm 3.3 A. oryzae (Ahlburg) Cohn 1.6 2.9 A. panamensis Raper and Thom 2.9 A. proliferans G. Smith 1.6* A. re strict us G. Smith 1.4 A. ruber (Spieckermann and Bremer) Thom and Church 1.4 A. sulphureus (Fres.) Thom and Church 3.3 1.4 A. sydowi (Bain, and Sart.) Thom and Church 3.3* 7.1$ A. turn aril Kita 1.6* 2.9$ A. terreus Thom 3.3 13.4$ A. unguis (Emile-Weil and Gaudin) Thom and Raper 1.4$ A. ustus (Bainier) Thom and Church 1.6 11.9$ A. versicolor (Vuill.) Tiraboschi 10.1* 5.9$ A. weniii Wehmer 13.5* 50.7$$ A. spp. 11.8* 14.9 Botryophialophora marina Linder 2.9 Cephalosporium acremonium Corda 3.3* 10.4+$ C. asperum Marchal 1.4 C. coremioides Raillo 1.4 C. curtipes Sacc. 22.0* 10.4 C. humicola Oudemans 3.3* 1.4 C. roseo-griseum Saksena 6.7* 11.9$ C. sp. 1.6 11.9 Fusidium viride Grove 1.6 1.4 Gliocladium pmbriatum Gilman and Abbott 1.4 Malbranchea sp. 1.4 Moeszia sp. 1.6 Monilia brunnea Gilman and Abbott 1.4 Monocillium sp. 1.6 5.9$ Paecilomyces fusisporus Saksena 1.6* P. varioti Bainier 1.6 2.9 Penicillium albidum Sopp 1.4 P. brevi-camp actum Dierckx 3.3 P. canescens Sopp 3.3 P. caseicolum Bainier 1.6 1.4 P. charlesii Smith 1.6 1.4$ 324 PACIFIC SCIENCE, Vol. XXI, July 1967 TABLE 3 ( continued ) PERCENTAGE PERCENTAGE OCCURRENCE OCCURRENCE IN WATER IN SAND NAME SAMPLES SAMPLES P. chermesinum Biourge 1.4 P. citrinum Thom 1.6 8.9J P. commune Thom 1.6 P. corylophilum Dierckx 1.4t P. cyaneo-fulvum Biourge 1.6 P. cyaneum (Bainier and Sartory) Biourge 1.6 P. janthinellum Biourge 3.3 P. kapuscinskii Zaleski 1.4t P. lanosum Westling 6.7 1.4 P. lanoso-coeruleum Thom 5.0 P. lilacinum Thom 6.7* 10.4 P. miczynskii Zaleski 1.4 P. nigricans (Bainier) Thom 8.4 8.9+ P. notatum Westling 5.0 P. oxalicum Currie and Thom 1.6 1.4 P. piscarium Westling 1.4 P. purpurrescens Sopp 1.6 P. raciborskii Zaleski 1.6 P. rotundum Raper and Fennell 1.4t P. simplicissimum (Oud.) Thom 1.6* P. steckii Zaleski 1.6 4.4J P , velutinum van Beyma 2.9 P. spp. 10.1* 22.3W Rhinotrichum sp. 1.4++ Scopulariopsis brevicaulis Bainier 4A% S. brumptii Salvanet-Duval 1-4+ 5. carbonaria Morton and G. Smith 1.6 5. croci van Beyma 1.4t S. fimicola (Cost, and Mat.) Vuill. 1.6 2.9 S. sp. 1.6 7.1 Sepedonium sp. 1.6 2.9 Spicaria simplicissima Oudemans 1.6* Sporotrichum epigaeum Brunard 2.9 Trichoderma album Preuss 3.3 T. glaucum Abbott 1.6 2.9 T. koningi Oudemans 5.0* 1.4 T. lignorum (Tode) Harz 15.2* 11.9 Trinacrium sp. 1.4 T ritirachium album Limber 1.4 T. purpureum (Saito) Beyma 3.3* Varicosporium sp. 1.4 V ertici Ilium candelabrum Bonorden 1.6 V. terrestre (Link) Lindau 3.3 1.4 Dematiaceae Acrostaphylus sp. 1.4 Acrotheca sp. 1.4 Alternaria fasciculata Cooke and Ellis 1.6 1.4 A. geophila Daszewska 1.6 A. humicola Oudemans 3-3 A. tenuis Nees 3.3 Aureobasidium mansonii (Cast.) Cooke 1.4t A. pullulans (De Bary) Arnaud 13.5* 8.9tt A. sp. 1.6 4.4t Marine Fungi from Central Pacific — Steele 325 TABLE 3 ( continued ) PERCENTAGE PERCENTAGE OCCURRENCE OCCURRENCE IN WATER IN SAND NAME SAMPLES SAMPLES Bispora sp. 1.4 Catenularia sp. 1.6 Chalaropsis sp. 1.6 Chloridium sp. 1.6 Cladosporium cladosporioides (Fres.) de Vries 16.9* 11.9+t C. epiphyllum Persoon 16.9* 11.9 C . herb arum (Persoon) Link 10.1 8.9tt C. lignicolum Corda 3.3 C. spp. 10.4 Cordana sp. 2.9+ Curvularia geniculata (Tracy and Earle) Boedijn 6.7 1.4+ C. interseminata (Berkeley and Ravenel) Gilman 3.3 C. pallescens Boedijn 3.3 4.4 C. subulata (Nees) Boedijn 3.3* C. tetramera (McKinney) Boedijn 3.3 Dendryphion sp. 1.6 Gliobotrys alboviridis von Hohnel 1.4 Gliomastix convoluta (Harz) Mason 8.9 Gonytrichum macrocladum (Sacc.) Hughes 1.6 G. sp. 1.4 Hansford/a togoensis Hughes 1.6 H. sp. 1.6 Helminthosporium anomalum Gilman and Abbott 16.9 1.4 H. sativum Pammel, King and Bakke 1.6 Heterosporium sp. 1.4 Hormodendrum cladosporioides (Fresenius) Sacc. 1.6 Humicola grisea Tragen 1.6 H. lanuginosa (Griff and Maubl.) Bunce 1.4 H. nigrescens Omvik 1.4 H. sp. 2.9 Macrosporium cladosporioides Desm. 1.6 M. sacrinaeforme Cavara 1.6 Masoniella grisea (Smith) Smith 5.0 16.4+ Megaster sp. 17.6+$ Menispora apicalis Berk, and Curt. 1.6 Nigrospora sphaerica (Sacc.) Mason 3.3 8.9+ Oidiodendron citrinum 1.4 0. griseum Robak. 1.4 Passalora sp. 1.4 Periconia byssoides Persoon 1.4 P. hispidnla (Pers. ex Pers.) Mason and M. B. Ellis 1.4 Phialophora sp. 8.4 25.3+$ Scolecotrichum sp. 1.4+ Stachybotrys atra Corda 1.4 S. lobulata Berkeley 4.4 Stemphylium botryosom Wallrath 1.6 S. macro sporoideum (Berkeley and Broome) Sacc. 3.3 Torula allii (Harz) Sacc. 1.6 T. lucifuga Oudemans 4.4$ T. sp. 1.4$ Trichocladium sp. 1.4 Zygosporium masonii Hughes 4.4 Stilbaceae 326 PACIFIC SCIENCE, VoL XXI, July 1967 TABLE 3 ( continued ) PERCENTAGE PERCENTAGE OCCURRENCE OCCURRENCE IN WATER IN SAND NAME SAMPLES SAMPLES Didymostilbe sp. 1.6 Graphium sp. 1.4 Harpographium sp. 1.6 Synnematium jonesii Speare 3.3 1.4 Tuberculariaceae Cylindrocarpon didymum (Hartung) Wo lien weber 2.9+ C. radicicola Wollenweber 1.4 Epicoccum purpurascens Ehrenb. 1.6 E. sp. 2.9 Fusarium merismoides Corda 1.6 F. spp. 22.0* 44.7++ Hymenella sp. 4.4+ Myrothecium roridum Tode 3.3 M. venue aria (Alb. and Schw.) Ditmar ex Fr. 1.4 M. sp. 1.6 1.4 Yeasts Rhodotorula spp. 27.1* 1.4 Orange yeasts 1.6 8.9++ Pink yeasts 2.9+ Black and orange yeasts 1.4 Black yeasts 11.8* 8.9++ Yeasts 45.8* 20.8++ My celia sterilia (Dematiaceae) 28.8* 37.3++ My celia sterilia (Moniliaceae) 16.9* 40.1++ * Isolated from oceanic zone, f Isolated from the Line Islands. X Isolated from the Phoenix Islands. technique, is given in Table 1. A total of 37 samples had fewer than 100 isolates per sam- ple; the remaining 30 samples had more than 100 isolates per sample. The sand of the lee- ward Hawaiian islands had the highest number of isolates, over 500 per sample, but the num- ber of different species was lower than in com- parable samples from the Phoenix Islands. The sands returned from 1 to 35 species per sample. The majority yielded from 3 to 9 species each. The intertidal sands of the main Hawaiian islands returned the highest number of species, a total of 68. Black Sand Beach, Hawaii, and Kaena Point, Oahu, each yielded totals of 16 species. The supratidal zone of Kuhio Beach, Oahu had the highest number of species for sites in that zone: 35 species among 16 genera. The zone total was 53 spe- cies. The subtidal zone returned the lowest number of species, only 22. In Table 2, the 18 fungi occurring most frequently in water and sand are listed. Among these, 9 are common to both areas although of different rank for percentage of occurrence; 7 occur in water and not sand, and 8 occur in sand, not in water. Of those common to both water and sand, yeasts, aspergilli, and peni- cillia were common to all sand samples. Neither Rhodotorula spp. which is penultimate in rank for water, nor Aureobasidium pullulans, also frequent in water samples, was predominant for sand samples. The control plates poured when both water and sand samples were plated showed no growth. Only one colony was observed on a plate exposed to determine the level of air contamination in the laboratory. DISCUSSION Isolates obtained from water samples indi- cate that abundant and varied fungus popula- Marine Fungi from Central Pacific — Steele tions do exist in this environment, but that frequencies vary with zones. The oceanic zone had fewer isolates than did the other zones studied in the pelagic region. This result was expected, inasmuch as oceanic regions are known to have lower populations of marine organisms than do regions closer to land. Of the 59 water samples, 50 were taken from areas that are strongly influenced by the presence of oceanic islands, whereas the other 9 were from areas well away from any shore. Areas near islands are known to support abundant and varied marine life. Ships may also be a source of water pollution, and therefore modified populations might be expected in shipping lanes. Differences can also be observed between the two oceanic locations studied, as reflected in differences in kind of fungi present but not in numbers. The oceanic area off Johnston Island contained more yeasts than did surface samples obtained near the island of Oahu. The latter had more fungi which would be classified as terrestrial, such as aspergilli and penicillia. The high yeast population observed from the oceanic areas is in agreement with the findings of both ZoBell (1946) and of Fell and Van Uden (1963). Members of the genus Rhodotorula were isolated consistently from all water samples, including the depth samples. Roth et ah (1962) have noted the common occurrence of these yeasts in oceanic localities, an observation now confirmed by these studies for the Pacific Ocean. When the three zones — surf, inshore neritic, and offshore neritic — are compared, a correla- tion is found between numbers of isolates per ml and location of zone. This is not unex- pected. This correlation is particularly clear in the data for these zones in Oahu. The surf zone, which is an unstable area with constant wave action, returned 0.10 isolates per ml com- pared with 0.33 for offshore neritic and 0.83 for inshore neritic zones. Species of Curvularia, Alternaria, and Helminth osporium were iso- lated repeatedly from the inshore neritic zone. This zone, in the area of Oahu, is richer in number of species than is the same zone on Hawaii. This might be due to pollution, as Oahu has a greater population and has a major port for shipping. Members of the order 327 Sphaeropsidales, of the Fungi Imperfecti, were frequently found in the Oahu samples. These fungi are parasitic on plants. The polluted area contained much floating debris which could serve as a source of these fungi. Some water samples were taken from tidal pools in the intertidal zone. In every case there was a high number of isolates. Intertidal pool populations may be affected by higher temper- atures, higher organic content, salinity levels, or wave action. One area, however, that of a large bay on the windward side of Oahu, did not return a high yield of isolates. This area has a high fresh-water run-off which reduces the salinity at the surface as much as 5 %o dur- ing the rainy season. Even though the area had a lower population than expected, it did have great diversity, explained perhaps by the run-off factor. The high number of aspergilli and penicillia isolated was to be expected because of their proclivity for sporalation, cosmopolitan habitat, and their great adapt- ability. A good percentage of these might be run-off and/or air contaminants introduced into the water. The fact that more fungi were isolated from the sand than from the water supports the well-known observation that microbial popu- lations are higher in relation to fixed surfaces. Examination of Table 3 also shows that the general population of sand is quite different than that of water. The high number of isolates from the sand samples taken from the leeward Hawaiian islands and the few species among them is in direct contrast with the low number of isolates and high number of species in the samples from the Phoenix Islands and the Line Islands. There could be many reasons for this. The elapsed time before plating the leeward sam- ples was greater than for those of the Phoenix and Line islands. Variations in bird population, temperature, and humidity, and shore stability among the leeward Hawaiian islands, may be critical factors controlling fungus populations. If the bird population serves as a control, a survey for keratinophilic and coprophilous spe- cies might be rewarding. Such a survey should also extend to other islands with large bird populations. Another factor influencing fungus populations, as reflected in the high number 328 of species returned from the Phoenix and Line islands, is the fact that Gardner Island, Wash- ington Island, and Palmyra Island are, or have been, inhabited. The supratidal zone gave the highest average number of isolates per gm. This zone is a very stable area, a mixture of soil and sand. When the average number of isolates is compared for the supratidal, intertidal, and subtidal zones on Oahu, the intertidal displays the lowest average number of isolates, which may be explained by the influence of constant washing by the waves. The tidal pool area of the sand, like the tidal pool area of the water, is characterized by the presence of more fungi than are found in the surrounding areas of each zone. Two of the sand areas sampled are unique among sands for their color and composition, namely a green sand beach and a black sand beach. The black sand beach had 259 isolates distributed through 12 genera and 200 species, while the green sand beach had 28 isolates, in 10 genera and 12 species. Although 4 genera were common to both, only 1 species, Asper- gillus terreus, occurred at both sites. Both beaches are on the island of Hawaii. One rea- son for this difference may be that the black sand beach is continually exposed to contamina- tion from people and their litter, while the green sand beach is in a very remote location. Another reason could be the difference in the chemical composition of the sands. Black sand is formed from basalt lava rock and cinders, and green sand is formed by the release of oli- vine crystals which are in the basalt lava rock. PACIFIC SCIENCE, VoL XXI, July 1967 Most of the Phycomycetes and the Ascomy- cetes were isolated from sand. These fungi are known to adhere to substrates. Only one species predominates over the others, Aspergillus wentii, as shown in Table 2. When plating four samples on mycobiotic agar from beaches that are local tourist attractions on Oahu, one po- tential pathogen was found: Microascus inter - medius, which has been isolated from a number of soils by mouse passage. Several species of this genus are known to be etiologic agents of dermatophytoses and onychomycosis in man (Barron et ah, 1961). It is interesting to compare the results ob- tained by Roth et al. (1964) in the Atlantic with the results obtained from this study in the Pacific. Roth took 227 water samples and iden- tified 41 genera among his isolates. This study encompassed 59 water samples and resulted in the identification of 59 genera. Of these, 29 genera were common to both lists. The Atlantic list included 11 genera not reported in this study; this study includes 29 genera not reported by Roth et al. (1964). Table 4 shows the difference in average num- ber of isolates per liter between the samples taken in the Atlantic Ocean and those taken in the Pacific Ocean. Furthermore, it is a striking fact that no Pa- cific sample, either water or sand, was without fungi, whereas Roth et al. (1964) recovered fungi from only 80% of their 227 samples. Because they did not include sand samples in their study, comparisons can be made only be- tween water samples. The maximum number of TABLE 4 Comparison of Numbers of Isolates from the Atlantic Ocean and the Pacific Ocean ATLANTIC OCEAN (ROTH ET AL. 1964) PACIFIC OCEAN (1965) DEPTH NUMBER OF SAMPLES AVERAGE NUMBER ISOLATES PER LITER DEPTH NUMBER OF SAMPLES AVERAGE NUMBER ISOLATES PER LITER 500 to 5 11.1 600 m 1 13.0 1000 m 17 12.1 300 m 1 6.0 9 4.9 300 m 1 10.0 2 1.0 Surface 12 15.5 Surface 1 18.0 to 79 17.5 1 55.0 500 m 43 15.1 1 30.0 16 3.0 1 73.0 Marine Fungi from Central Pacific — Steele species per offshore sample in the Atlantic was 6. In the Pacific the number ranged from 1 to 11. Although the maximum number of species per Pacific sample exceeds that reported for the Atlantic, the total number of species found in each area is approximately the same: 133 species in the Atlantic compared with 127 in the Pa- cific. The diversity of genera is notably higher in the Pacific samples. Samples from both areas yielded fungi that were not identified. There are both differences and similarities between the kinds of fungi obtained from the two regions. A total of 30 species were common to Atlantic and Pacific waters. Neither Asco- mycetes nor Basidiomycetes were reported for the Atlantic. From the Pacific water 2 Ascomy- cetes and 1 Basidiomycete were recovered. Roth et al noted the low incidence of Phycomycetes, reporting only 6. Five species were identified in this Pacific study, but only 2 were from water samples. Species of Sphaeropsidales occurred in similar numbers in both oceans: 7 in the At- lantic and 5 in the Pacific, but only 2 species were common to both. In comparing the species of fungi which are predominant in each area, some species are found on both lists. Roth et al. distinguished 8 fungi each from the eulittoral and oceanic water samples as their dominant species. Of these, Aureohasidium pullulans was most common in the oceanic zone and it did not occur in their eulittoral list. In the Pacific, this fungus was common in water but ranked sev- enth among 18 species. Cladosporium species (sic) were the most common fungi in the At- lantic eulittoral samples, and occupied second place in the oceanic list. Two species, Clado- sporium cladosporioides and C. epiphyllum, ranked high among those frequent in both wa- ter and sand samples of the Pacific. Tricho- derma lignorum, the last species on the Atlantic list of eulittoral dominants, is absent from the corresponding oceanic list. In the Pacific it oc- curred in both water and sand samples, although much more frequently in water than in sand. In summary, the differences suggest that the Pacific fungus population is different from that of the Atlantic. Having established the fact that fungi can be isolated from Pacific as well as Atlantic water and shores, there still remains the prob- lem of how one determines whether or not a 329 certain isolate is a marine fungus. There is no single diagnostic test. The criteria which have been suggested are summarized by Roth et al. (1964) and may be stated as: (1) the isolate must grow and reproduce exclusively or pre- dominantly in the sea or on intertidal substrata, or (2) the isolate must grow and reproduce at an optimal level in the normal salinity of the oceans. None of these criteria can be supported without qualification. There is still no means of concrete demonstration of growth and repro- duction of a fungus in vivo in marine environ- ments except for those fungi growing on nat- ural (and introduced) substrata. If growth and reproduction at normal salinity levels is ac- cepted as a criterion, this allows for the inclu- sion of fungi found in salt lakes (Anastasiou, 1963). Moreover, Gray (1963) has shown that many fungi of terrestrial origin are capable of better growth in sea water media than in fresh water media. If growth is limited to natural substrata, then the possibility of free-living ma- rine fungi is excluded. As more investigations of marine habitats are undertaken, the number of genera and spe- cies isolated from them will undoubtedly in- crease. Many of the genera found in the Pacific are known from other marine locations, e.g., species of Aureohasidium, Macrophoma, Phoma , Diplodina, Diplodia, Epicoccum, Fusidium, Cladosporium, Alternaria, and Macro sporium (Johnson and Sparrow, 1961). Repeated isola- tion, however, is not confirmatory evidence, but is only suggestive. Roth et al. (1964) contend that, until further distributional and physiolog- ical data are obtained, fungi so isolated should be regarded as of incidental occurrence in the sea. Nor is the rare isolation of a species from only marine habitats reliable evidence, as this may reflect only the randomness of the sam- pling and the samplers. Curiously, Dendry- phyiella arenaria Nicot, which is dominant in Atlantic eulittoral samples and known only from marine sources, was not among the Pa- cific isolates. Conversely, Botryophialophora marina, also known only from marine sources, was found in the Pacific but not in the Atlantic. In conclusion, a working definition for a marine fungus is proposed: A marine fungus is one which is capable of producing successive generations by sexual and/or asexual means in 330 PACIFIC SCIENCE, VoL XXI, July 1967 natural oceanic waters or on intertidal substrata. Until experimental means of proof are devised, the data presented here will serve as contribu- tory evidence for the distribution of fungi iso- lated from marine habitats. ACKNOWLEDGMENTS The author wishes to express her sincere ap- preciation to Dr. Gladys E. Baker, Department of Botany, University of Hawaii, for guidance and suggestions throughout this study. She would also like to express her thanks to Dr. C. H. Lamoureux and Dr. E. S. Reese, mem- bers of her thesis committee. REFERENCES Anastasiou, C. J. 1963. Fungi from salt lakes. II. Ascomycetes and Fungi Imperfecti from the Salton Sea. Nova Hedwigia 6:243-276. Baker, G. E. 1964. Fungi in Hawaii. Hawaiian Botanical Society Newsletter 3:23-29. Barghoorn, E. S., and D. H. Linder. 1944. Marine fungi: their taxonomy and biology. Farlowia 1:395-467. Barron, G. L., R. F. Cain, and J. C. Gilman. 1961. The genus Microascus. Can. J. Bot. 39:1609-1631. Buck, J. D., and H. C. Cleverdon. I960. The spread plate as a method for the enumera- tion of marine bacteria. Limnol. Oceanog. 5:78-80. Cooke, W. B. 1961. Fungi from Raroia in the Tuamotu Archipelago. Pacific Sci. 15(2) : 186-188. Cribb, A. B., and J. W. Cribb. 1954. Three species of fungi parasitic on marine algae in Tasmania. Pap. Univ. Qd. Dept. Bot. 3:9-13. • 1955. Marine fungi from Queensland. I. Ibid. 3:77-81. 1956. Marine fungi from Queensland. II. Ibid. 3:97-105. I960. Marine fungi from Queensland. III. Ibid. 4:41-48. Dorn, W. G. Van. 1956. Large volume water samplers. Trans. Am. Geophys. Union 37:628-684. Fell, J. W., D. G. Ahearn, S. P. Meyers, and F. J. Roth, Jr. I960. Isolation of yeasts from Biscayne Bay, Florida and adjacent ben- thic areas. Limnol. Oceanog. 5:366-371. and Van Uden, N. 1963. Yeasts in marine environments, pp. 329-340. In: C. H. Oppenheimer, ed., Symposium on Marine Mi- crobiology. Charles C Thomas, Springfield, Illinois. Fred, E. B., and S. A. Waksman. 1928. A Laboratory Manual of General Microbiology. McGraw-Hill Book Co., New York. Gray, W. D. 1963. Growth of fungi in sea water medium. Appl. Microbiol. 11:501- 505. Hedgpeth, J. W. 1957. Treatise on Marine Ecology and Paleoecology. Geol. Soc. Am., Mem. 67, Vol. I. Johnson, T. W., Jr., and F. K. Sparrow, Jr. 1961. Fungi in Oceans and Estuaries. Hafner Publishing Co., New York. Jones, E. B. G. 1962. Marine fungi. Trans. Brit. Mycol. Soc. 45:93-114. Kobayashi, Y. 1953. Studies on the marine Phycomycetes. Bull. Nat. Sci. Museum (To- kyo) 33:53-65. Kohlmeyer, J. I960. Wood-inhabiting marine fungi from the Pacific Northwest and Cali- fornia. Nova Hedwigia 2:383-398. Machlis, L., and J. G. Torrey. 1956. Plants in Action. A Laboratory Manual of Plant Physiology. W. H. Freeman and Co., San Francisco. Marsh, D. H. 1965. Microorganisms of the Phyllosphere, with Particular Reference to Fungi Occurring on the Dominant Plants of Biogeoclimatic Zones of the Hawaiian Is- lands. Unpublished M.S. thesis, University of Hawaii, Honolulu, Hawaii. Meyers, S. P., and E. S. Reynolds. I960. Oc- currence of lignicolous fungi in northern At- lantic and Pacific marine localities. Can. J. Bot. 38:217-266. Moore, R. T., and S. P. Meyers. 1959. Tha- lassiomycetes. I. Principles of delimitation of the marine mycota with the description of a new equatically adapted deuteromycete genus. Mycologia 51:871-876. Olive, L. S. 1957. Tulasnellaceae of Tahiti. A revision of the family. Mycologia 49:663- 679. 1958. The lower Basidiomycetes of Ta- Marine Fungi from Central Pacific — Steele hiti, I. Bull. Torrey Bot. Club 85(1) : 5—27; 85(2) :89-110. Orpurt, P. A. 1964. The microfungal flora of bat cave soils from Eleuthera Island, The Bahamas. Can. J. Bot. 42:1629-1633. Petersen, R. H. I960. Some soil and copro- philous fungi from the South Pacific area. Mycologia 52:552-556. Rogers, D. P. 1947. Fungi of the Marshall Islands, Central Pacific Ocean. Pacific Sci. 1:92-107. Roth, F. J., Jr., D. G. Ahearn, J. W. Fell, S. P. Meyers, and S. A. Meyer. 1962. Ecol- ogy and taxonomy of yeasts isolated from various marine substrates. Limnol. Oceanog. 7:178-185. P. A. Orpurt, and D. G. Ahearn. 1964. Occurrence and distribution of fungi in a subtropical marine environment. Can. J. Bot. 42:375-383. Salle, A. J. 1954. Fundamental Principles of 331 Bacteriology. 4th ed. McGraw-Hill Book Co., New York. Sparrow, F. K. 1948. Soil Phycomycetes from Bikini, Eniwetok, Rongerik, and Ronggelap atolls. Mycologia 40:445-453. Steele, C. S. W. 1965. Fungus Populations in Marine Waters and Coastal Sands of the Hawaiian, Line, and Phoenix Islands. Un- published M.S. thesis, University of Hawaii, Honolulu, Hawaii. Wickerham, L. J., M. H. Flickinger, and K. A. Furton. 1946. A modification of Henrici’s vegetable- juice sporulation medium for yeasts. J. Bact. 52:611. Van Uden, N., and R. Castelo Branco. 1961. M etschnikowiella zobellii sp. nov. and Metschnikowiella krissii sp. nov., two yeasts from the Pacific Ocean pathogenic for Daph- nia magna. J. Gen. Microbiol. 26: 141-148. Zobell, C. E. 1946. Marine Microbiology. Chronica Botanica Co., Waltham, Mass. Distribution and Movements of Birds in the Bering and Chukchi Seas L. G. Swartz1 This paper reports on pelagic observations of about 29 species of birds in the northern Bering Sea and the Chukchi Sea during a cruise in the late summer of I960. This work represents part of a larger study of the sea bird colonies at Cape Thompson, Alaska (Swartz, 1966). The problems presented by the offshore dis- tribution and movements of sea birds have proved refractory to many workers. Primarily, efforts to delineate and solve these problems have been incidental to other objectives of sea voyages and have centered in the North At- lantic and Barents Sea. Wynne-Edwards (1935) has brought together much of this scattered work from the North Atlantic, and Belopolski (1957) summarized data from the Barents Sea. Recently, Kuroda (I960) and Shuntov (1961) have published observations extending into the Bering Sea. Jacques (1930) is the only worker to publish substantial pelagic observations north of Bering Strait. THE ENVIRONMENT An intensive investigation of the Chukchi Sea (Fig. 1) began in 1959 when the De- partment of Oceanography of the University of Washington, under the direction of Dr. R. H. Fleming, and the Bureau of Commercial Fisheries sent their respective research vessels, the "Brown Bear” and the "John Cobb,” to the area. These organizations together under- took extensive physical, chemical, and biologi- cal investigations. In I960, the Department of Oceanography sent the "Brown Bear” to these waters again in order to extend and verify the results of the 1959 cruise. Two publications (Wolfe, I960, 1962) in- clude brief summaries of the scope of the ma- rine programs but present little actual data. 1 Biological Sciences Department, University of Alaska, College, Alaska. This work was done under Contract No. AT- (04-3) -310 with the United States Atomic Energy Commission. Manuscript received June 29, 1966. The results of these projects have been pre- sented in preliminary form in several reports prepared for the individual financing agencies. Formal publication has been made of some of the University of Washington work (Creager and McManus, 1961; Fleming et ah, 1961; Fleming and Heggarty, 1962), and a large volume has recently been published including the work of many individuals which provides comprehensive coverage of almost all aspects of the marine environment included within the scope of this paper (see Swartz, 1966). The following brief summary of the char- acteristics of this environment is summarized from personal experience, the preliminary re- ports mentioned, the published works, and from personal communication and conversation with the individuals involved in the marine programs. The Chukchi Sea, in which most of the Fig. 1. The Bering and Chukchi seas showing the area included within this study. Major sea bird colonies near the cruise track are indicated with the circular symbol. 332 Birds in Bering and Chukchi Seas — Swartz pelagic bird observations were made, is shal- low with a relatively featureless bottom (Fig. 2). In the area sampled, few depths exceed 35 fathoms. Current and temperature patterns are complex. In general, current flow is north- ward through Bering Strait (Fig. 4). Patterns of current flow in the northern Chukchi Sea seem to be affected strongly by winds. Tem- perature patterns at 5-m depths are shown in Figure 5. The invertebrate fauna is rich and abundant, but fishes are not conspicuously abundant in either species or individuals, though of course their numbers are adequate to support sea bird colonies in apparent pros- perity (Swartz, 1966). ACKNOWLEDGMENTS Dr. R. H. Fleming, of the University of Washington, offered to place a man aboard the University of Washington research vessel, "Brown Bear," to observe sea birds during a portion of its I960 oceanographic cruise (Brown Bear Cruise 268). Mr. E. J. Wil- loughby was selected to be the observer and deserves great credit for the zeal and accuracy of his work. The cooperation of Dr. Fleming, the crew, and the research staff of the "Brown Bear” is gratefully acknowledged. Willoughby’s activities were financed to a major degree by a program directed by F. S. L. Williamson of the Arctic Health Research Center. I am greatly indebted to Mr. William- son for permission to include in this paper the observations on species not breeding at Cape Thompson. These species were initially in his province and this paper could not have had its present form without his generous cooperation. PROCEDURE The portion of the cruise track of the "Brown Bear" which is included in this paper is shown in Figure 3. Willoughby boarded the "Brown Bear” on August 6 near Cape Thomp- son and disembarked at Nome on August 28. Most observations were made from the fly- ( ing bridge where the view in all directions was relatively unobstructed. In order to achieve an objective index of abundance and movements, 10-minute-long counts of all birds seen and 333 records of their activities were made at inter- vals throughout the 24-hour period. Detailed observations were continued between the 10- minute counts to the extent permitted by weather, visibility, and the endurance of the observer. Over 600 entries pertaining to sea birds were made in addition to 10-minute counts. Latitude and longitude are known for each entry. Observations including time, posi- tion, surface temperature of the sea, wind speed and direction, wet and dry bulb air tempera- ture, barometer readings, precipitation, size and direction of swell, and approximate visibility were recorded several times a day from routine readings made by Willoughby and other ship personnel. All of these data were examined and those which proved meaningful in inter- preting the avian observations are discussed at the appropriate place. Since conditions of visi- bility varied widely from day to day, no effort was made to convert the data to absolute abun- dance per square unit of sea surface as was done by Kuroda (I960). During the same interval that Willoughby was making observations at sea, a shore party was conducting investigations of the large col- onies at Cape Thompson. It was hoped that comparisons of behavior at the breeding cliffs with offshore observations would yield signifi- cant information not otherwise obtainable. With the exception of the expected observa- tion that departure of flocks from the cliffs produced a rise in numbers observed at sea, this hope was not realized. RESULTS Below, listed phylogenetically, are discus- sions of distribution, abundance, and move- ments of birds seen from the "Brown Bear.” Unless specifically noted, all species were pre- viously observed in the Bering or Chukchi seas by Jacques (1930) or Shuntov (1961), the only authors who have published substantial offshore observations which overlap those re- ported here. Loons (Gavia sp.) Two sightings of unidentified loons were made, both close to shore (Fig. 6). Four of these birds were seen at 69°46'N, 163°17'W 334 PACIFIC SCIENCE, Vol. XXI, July 1967 1961.) Birds in Bering and Chukchi Seas — Swartz 335 Fig. 3. The Bering and Chukchi seas, showing included within this study. (Modified from Fleming between Point Lay and Icy Cape; another was seen near Chamisso Island in Kotzebue Sound. Fulmar (Fulmarus glacialis) Fulmars were seen on 11 occasions (Fig. 6), ordinarily as single birds. Two birds were seen together near Bering Strait and three or four near Little Diomede Island amid a large num- ber of alcids. Feeding concentrations as seen by Kuroda (1960:59) were not observed but, on the other hand, Fulmars were not abundant during any part of this voyage. Shuntov (1961: 1063) described Fulmars as the most abundant bird in the Bering Sea west of the Pribilofs. Jacques (1930:360-361) remarked on their abundance near the Pribilofs and off East Cape but, in common with observations from the ’'Brown Bear,” he saw them only occasionally in the Chukchi Sea. All the Fulmars observed (14 or 15) were the light phase, which agrees with the observations of Jacques (1930:361) that in the Arctic the light phase greatly pre- dominates. i Shearwaters Probably all shearwaters seen (Fig. 7) were Slender-billed Shearwaters ( Puffinus tenuiros- the portion of the cruise track of the "Brown Bear’’ et al., I960.) iris'), but they could not always be identified with certainty. It is possible that some were Sooty Shearwaters, though no records are known from north of the Aleutian Islands (Gabrielson and Lincoln, 1959:80-81). The former species, which breeds in the southern hemisphere, spends its nonbreeding season in northern waters and has been collected and observed as far north as Point Barrow (Gabri- elson and Lincoln, 1959:78). Several of the sightings reported here were feeding flocks, on one occasion near Cape Thompson compris- ing between 500 and 1,000 individuals. In no case, however, did abundance approach the concentrations that have been observed by other authors south of Bering Strait (see Gabrielson and Lincoln, 1959:79; Shuntov, 1961:1061- 1062). Cormorants (Phalacrocorax) Cormorants were identified on four occasions (Fig. 7), chiefly within sight of nesting cliffs. One doubtful sighting was made near Chamisso Island in Kotzebue Sound. Cormorants were presumably all Pelagic Cormorants (P. pelagi- cus), but doubt exists in some cases. An obser- vation about 20 miles from Little Diomede 336 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 4. Surface currents (5.0 m). Vector length indicates speed of current. (From Fleming and Heggarty, 1960.) Fig. 5. Surface (5.0 m) isotherms. (From Fleming et al., I960.) Fig. 6. Observations of loons, O ; the Fulmar, $ ; Pectoral Sandpiper, □ ; unidentified sandpipers, H; and the Long-billed Dowitcher, A. Fig. 7. Observations of the Slender-billed Shearwater, O ’> unidentified shearwaters, ® ; cormorants, □ ; the Old Squaw, Common Eider, A; Spectacled Eider, ▲; and unidentified eiders, Q. Birds in Bering and Chukchi Seas — Swartz Island represents the maximum distance these birds were seen from shore or nesting colonies. Jacques (1930:362) apparently did not ob- serve cormorants north of Bering Strait, but his voyage did not bring him close to breeding colonies so this is not surprising. Old Squaw (Clangula hyemalis) Numerous Old Squaws were seen very close to shore between Cape Thompson and Point Hope by the shore party, but only a single pelagic observation (near Bering Strait) was made (Fig. 7). This pattern of distribution is probably typical of this species during the breeding season. Common Eider (Somateria mollissima) One sighting was made of 19 Common Ei- ders close to the shore of Cape Lisburne (Fig. 7). Spectacled Eider (Lampronetta fischeri) Two sightings were made, one single indi- vidual and one flock of four, both near the Cape Lisburne cliffs (Fig. 7). Unidentified Eiders Four sightings of small flocks of unidentified eiders were made, only one more than a few miles offshore (Fig. 7). Pectoral Sandpiper (Erolia melanotos) This species was identified only at one loca- tion, off Point Lay (Fig. 6), when a single bird landed on the deck of the "Brown Bear” and walked about for 5 minutes. A Pectoral Sandpiper was seen flying 15 minutes later and may have been the same bird. Jacques (1930:353-366) did not observe this species. Shuntov (1961:1066) observed sandpipers in the Bering Sea, but did not identify the species. Unidentified Sandpipers Sandpipers which could not definitely be identified were seen on four occasions (Fig. 6), two of which were more than 100 miles from shore. Long-billed Dowitcher (Limnodromus scolopaceus) (?) A single individual, probably of this species, 337 was observed about 30 miles off the coast near Kivalina (Fig. 6). Red Phalarope (Phalaropus fulicarius) All identified phalaropes were of this spe- cies, but it is possible that some Northern Phalaropes ( [Lobipes lobatus ) were present in the area. Jacques (1930:364) commented to similar effect that probably all the phalaropes he saw in the Arctic Ocean were Red Phalaropes. Red Phalaropes were seen at 28 locations, mostly in groups of 3-6, although 13 solitary individuals were seen. Observations were widely scattered over the course of the cruise, but none were made south of Bering Strait (Fig. 8). U nidentified Ph alar op es In many cases, it was not possible definitely to identify phalaropes. No doubt most, if not all, of the unidentified birds were Red Phalaropes (Fig. 8). Pomarine Jaeger (Stercorarius pomarinus) Seven scattered sightings of this species were made, all north of 67°N (Fig. 9). Four of these were single birds, two sightings were of two birds, and one of "several.” Jacques (1930:357) found it to be common or abun- dant north of Bering Strait during about the same time of year. Shuntov (1961:1065) ob- served a northerly movement of Pomarine, Parasitic, and Long-tailed Jaegers in the south- ern Bering Sea in the end of May and begin- ning of June, which probably represented mi- gration to breeding grounds. He saw Pomarine Jaegers commonly, but only infrequently ob- served the other species. Parasitic Jaeger (S. parasiticus) This species was seen on 12 occasions (Fig. 9). As was the case with the previous species, all sightings were made north of 67 °N. Eight sightings were of single individuals, one of three, two of two, and one of "several.” Long-tailed Jaeger (S. longicaudus) This species, though far more abundant as a breeding bird at least in the Cape Thompson area than the two preceding species, was al- most entirely absent in the pelagic observations. Only two birds were seen (Fig. 9). 338 PACIFIC SCIENCE, VoL XXI, July 1967 Fig. 8. Observations of the Red Phalarope, 05 and unidentified phalaropes, q. Fig. 9- Observations of the Pomarine Jaeger, 05 Parasitic Jaeger, Long-tailed Jaeger, A; and un- identified jaegers, A- Fig. 10. Observations of the Sabine’s Gull, 05 Arctic Tern, q; Yellow Wagtail, A 5 and the Water Pipit, A- Fig. 11. Abundance, distribution, and movements of the Pigeon Guillemot, 0 5 Kittlitz’s Murrelet, #; Para- keet Auklet, □; Crested Auklet, Least Anklet, A; unidentified auklet, A; Horned Puffin, Q; and Tufted Puffin, 3 • Numbers and direction of flight in both species of puffins is indicated by size of symbol and direction of vector: o, 1-5; Q, "some”; Q, hundreds. C indicates the birds were circling. Birds in Bering and Chukchi Seas — Swartz 339 Fig. 12. Abundance, distribution, and movements of Glaucous Gulls {open and half open circles ) and Herring Gulls {black circles). Numbers and direction of flight of Glaucous Gulls indicated by size of sym- bols and direction of vectors; smallest circle, 1-5; °, 6-10; 0> 11-20; 3 > 21-40; Q, more than 150. C indicates circling, F indicates following the ship, and X indicates birds on the water. Fig. 13. Abundance, distribution, and movements of Black-legged Kittiwakes. Numbers and direction of flight are indicated by size of symbols and length and direction of vectors: smaller circle, 1-5; °, 11-20; shortest arrow, 1-5; -» 6-10; » 11-20; C indicates circling and X indicates birds on the water. Unidentified Jaeger Jaegers were seen but could not be identified on 15 occasions. These observations were widely scattered but all were north of 67 °N (Fig. 9). Glaucous Gull (Larus hyperboreus) Adult and immature Glaucous Gulls were most abundant at the cliffs and beaches where they fed extensively on the eggs and chicks of other species, especially murres. They were seen frequently to about 25 miles offshore and only occasionally at greater distances (Figs. 12 and 14). This species and, to a lesser extent, kittiwakes often followed the ship for up to several hours at a time. Both species fed on garbage thrown overboard. Although some tendency to fly into the wind was observed from shore and from the ship, no large move- ments of Glaucous Gulls in response to wind were evident. Such movements by larids have been observed by other authors, however (e.g., Harrison, 1955:109-110). Herring Gull (Larus argentatus) Two doubtful sightings of immature birds (possibly the same bird sighted at different hours) were made 16-18 miles west northwest of Cape Thompson on August 24. Groups of two, four, and three individuals were sighted near shore in the vicinity of Port Clarence (Fig. 12). All Herring Gulls sighted were immature. Black-legged Kittiwake (Rissa tridactyla) Both adult and immature kittiwakes were common on the open ocean (Fig. 13). Adults were often in winter plumage. The adults in breeding plumage and immatures were most abundant near the nesting cliffs, but the radius of their daily movements from the cliffs is not 340 clearly demonstrated (Fig. 15). It seems likely that breeding kittiwakes do not regularly fly out to sea as far as murres, but rather feed closer to shore. Regular patterns of movements which may exist are not evident from the data. Kittiwakes do not exhibit as distinct a pattern of daily activity fluctuations as do murres (Swartz, 1966), and do not characteristically fly as straight a course, so that possible flight trends might tend to be obscured. No move- ment, abundance, or distributional phenomena associated with daily rhythms are evident. One kittiwake in winter plumage was col- lected at about 70°50'N, 165°30'W near the edge of the polar ice pack. The reproductive tract was undeveloped and brood patches were not present. It is likely that many of the adult kittiwakes seen on the open ocean far from shore are nonbreeders. Sabine's Gull (Xema sabini) These birds were seen on six occasions and all but one were immatures (Fig. 10). The only adults (six birds) were seen at 70°46'N, 165°42'W, near the northernmost point on the cruise track. No particular distributional pattern is evident. Arctic Tern (Sterna paradisaea) Arctic Terns were seen on 11 occasions in groups of from 1 to 17 birds, of which 7 were immature and the rest (43) were adults. Most of the sightings were within about 50 miles of a shoreline and all but 1 were within 40 miles of land (Fig. 10) . Murres (Uria lomvia and U. aalge) Thick-billed and Common Murres are very similar in appearance and could not be differ- entiated consistently under the conditions pre- vailing at sea. They are therefore considered together and such differences as exist between them are discussed at the appropriate places. The abundance and distribution of murres is plotted in Figures 16, 17, and 18. Due to the large total number of observations, only ten-minute count data are presented. Murres were the most abundant birds on the Chukchi Sea and were almost always visible from the ship even far from shore. Murres were seen in all but 24 (16%) of the 146 ten-minute PACIFIC SCIENCE, Vol. XXI, July 1967 counts. Of these 24 negative counts 19 were made in an area near and within Kotzebue Sound in which very few birds of any kind were seen. Clear correlations between murre distribution and water temperature were found in this study. A striking decrease in the number of murres was noted on August 14 as the ship passed from colder waters to the warmer waters near Kotzebue Sound, crossing the 9°, 10°, 11°, and 12°C isotherms as plotted by Fleming et al. (I960: Fig. 5). (These isotherms repre- sent the temperature 5 meters below the sur- face.) R. H. Fleming (personal communica- tion) has observed this correlation on previous cruises in the area. Storer (1952:185) showed that the main breeding range of the Thick- billed Murre lies in areas where the August surface water temperatures are below 10°C and that the temperature tolerance of the Com- mon Murre tends to be somewhat higher. Storer (1952:187) cited Salomonsen’s (1944) claim that low temperatures retard spring molt and breeding in murres. This is clearly a local or individual response and would reinforce the contention that water temperatures could in part account for local distribution patterns. As the ship passed deeper into Kotzebue Sound on August 14 a subsequent sharp drop in water temperature occurred, but no increase in murres was evident except near Chamisso and Puffin islands, where small numbers of Common Murres were seen. Grinnell (1900:7) reported "immense numbers" of Thick-billed Murres breeding on these islands but did not report Common Murres. Neither species seems any longer to be an abundant breeder. The situation with regard to distribution and temperature is complex and at present re- mains unclear. Possibly not only water temper- ature but salinity, food supply, depth, and distance from breeding concentrations are interrelated factors. The speculation that a pro- gressive northern range extension of the Com- mon Murre is occurring which is correlated with long-term warming trends is provocative. Far too few data are available at present, how- ever, to consider the hypothesis in detail. In colder waters, murres were seen on all but 3% of the 10-minute counts. Even in these cases, murres were observed between the count- 10 5000- 300( 20 30 40 49 60 • 5,000 + STATUTE MILES FROM • 3,5 00+ NEAREST KNOWN COLONY 70 91 96 100 8 128 8 98 110 137 16 2.50 — O 80- • % rj* 20 "\h. | ••• * § »i« l|A*l 80 -HV4- 90 92 137 30 40 50 60 70 STATUTE MILLS FROM NEAREST KNOWN COLONY Fig. 14. Abundance of Glaucous Gulls vs. distance from the nearest shore. Since Glaucous Gulls may nest at many points along the shore, comparison is made between abundance and distance from the nearest land. All entries are included. Fig. 15. Abundance of Black-legged Kittiwakes vs. distance from the nearest known colony. Because of incomplete knowledge of nesting colonies, observations from 10-minute counts near and in Kotzebue Sound are not included. All other observations are plotted. Fig. 16. Abundance of murres vs. distance from the nearest known colony. Due to the great amount of data available, only 10-minute counts are plotted. Because of incomplete knowledge of nesting colonies, data from 10-minute counts in and near Kotzebue Sound are omitted. 341 342 PACIFIC SCIENCE, Voi XXI, July 1967 Fig. 17. Abundance, distribution, and movements of murres ( 10-minute counts only). Abundance is in- dicated by length of vector and, when not in flight, by numbers at the point of observation. The number is printed on the vector when more than 100. C indicates circling; X indicates that the birds were on the water. Cape Lisburne {top) and Cape Thompson colonies are indicated by cross hatching. 1-5 6-10 » 21-30 - 11-20 » 31-50 - ing intervals. Although distinguishing the two species was difficult unless the birds came very close to the ship, it appears that away from the colonies at Cape Thompson and Cape Lis- burne (in general, 5 miles offshore and be- yond), Thick-billed Murres greatly outnumber the Common Murres, probably making up more than 90% of the murre population on the open ocean. On the nesting cliffs, the population is believed to include 60% Thick-billed Murres (Swartz, 1966), implying that Common Murres prefer shallower water than do thick-bills. This is consistent with data on food habits (Swartz, 1966), which imply that Common Murres feed in shallower water. The area on the open ocean in which the fewest murres were seen was near the ice pack at about 70°50/N, 166°0Q'W. » 51_80 » 81-100 — » The greatest number of murres was found within about 40 statute miles from the nearest colony (Fig. 16). It is apparent from Figures 16, 17, and 18 and from direct observation of feeding activities made from the ship that the usual feeding activity of breeding birds takes place within about 40 miles of the nesting cliffs and mostly within about 30 miles. Since murres are strong flyers and are capable of fly- ing at least 50 mph (Vaughan, 1937:123; Baxter and Rintoul, 1953; Portaz, 1928; Frow- hawk, 1928, in: Tuck, 1960:23)s a feeding dis- tance of 30-40 miles seems reasonable. Feeding areas for the Cape Thompson colonies seem to be primarily south of Point Hope, and those for the Cape Lisburne colonies north and west of the Cape Lisburne cliffs, although some over- Birds in Bering and Chukchi Seas — Swartz 343 Fig. 18. Inset from Figure 17. Symbols as in Figure 17. Cross hatching indicates the location of the Cape Thompson breeding cliffs. 344 lap of feeding areas may sometimes occur near Point Hope. No particular portion of the ocean off the cliffs seems to be favored for feeding. Comparison of distribution with bottom types (Sparks and Pereyra, 1960:7^) does not reveal clear correlations. Movements of murres in the open sea within about 40 miles of the nearest colony are strongly oriented by the location of the nesting colony and are little influenced by winds. Re- gardless of wind direction, murre flight is over- whelmingly oriented either toward or away from nesting areas (Figs. 17 and 18). Beyond the limit of daily feeding flights (about 40 miles), no significant flight trends are evident, in response either to weather or to colony loca- tion. Local or short-term orientations to winds may be striking. Takeoffs from both water and cliffs are made into the wind whenever pos- sible. In the immediate area of the nesting cliffs, flight patterns of murres approaching the cliffs are perceptibly influenced by winds. Sev- eral authors whose observations were made mostly from shore have noted flights of murres influenced by wind. Alexander (1935:299) observed feeding flights of Thick-billed Murres near Dungeness Point in England which usu- ally proceeded against the wind. The same author reiterates that movements are related to winds but are more related to tides and currents. Fay and Cade (1959:123) suggest that movements of murres at St. Lawrence Island are correlated with tidal currents. Alex- ander (1935:299) stated that birds are carried by water currents away from the feeding wa- ters and fly back to -regain their initial position. In the Cape Thompson area, however, neither tides nor currents are strong and probably have little influence on murre movements. It is pos- sible in the Cape Thompson area that winds may play the same displacing role that water movements may play elsewhere. Strong winds are common in the area, and Harrison (1955: 110) and several others have noted unusual flights of alcids following strong winds. At Cape Thompson, flocks of murres com- ing in often approach the coastline 1 or 2 miles downwind from their nesting location and fly against the wind relatively close to shore. Under foggy conditions, which are especially PACIFIC SCIENCE, Vol. XXI, July 1967 common when the sea ice is still present, the murres appear to use the shoreline as a guide and fly only a few feet above the beach. While flying against the wind, murres, like many other species fly lower where friction with the sub- strate slows the air movements. Willoughby at sea and members of the shore party all repeatedly observed this tendency. Often, birds approaching the cliffs begin to gain altitude when about 5 miles from shore. Birds leaving the cliffs at this distance from shore generally fly lower than those approach- ing the cliffs, often within a few feet of the water. Flocks of murres, both approaching and leav- ing the cliffs, are largest close to the cliffs, although flocking of the departing birds seems to take place farther out to sea than does the breaking up of arriving flocks. Flocks flying away from the cliffs break up as the distance from the colony increases, as though the birds spread out to fill in the areas away from the breeding center. Viewed from the shore, ap- proaching flocks are seen to retain their integ- rity until single birds or groups of birds break off to occupy their own nesting cliffs. This is most conspicuous at the ends of the colonies, where V-shaped flocks flying along the coast- line can be seen to gradually lose their identity while flying along the nesting cliffs. Four Thick-billed Murres were shot at sea. A male and a female were shot on August 22 at 67°53/N, 166°09/W; both showed evidence of having bred. A male and a female were shot on August 20 at 67°38/N, 165°45^W. The male was molting into the winter plumage but, from the presence of a regressing brood patch and testes still somewhat enlarged, presumably had bred. The female was molting extensively. It possessed no brood patch, and the ovary and follicles were minute; this was probably a non- breeding bird. Tuck (1961:82-119) presented data which seem to indicate that most murres seen far at sea are young which have not yet reached breeding age. Both Shuntov (1961 : 1059-1061) and Jacques (1930:357) observed Thick-billed and Common Murres. Beyond the implication that murres were abundant, only Shuntov (1961: 1059-1061) offered observations of real value in working out the broad outlines of murre Birds in Bering and Chukchi Seas — Swartz distribution and movement in this area. He stated that both species have similar patterns and that the main wintering waters are located between the edge of the ice and the Alaska Peninsula, primarily in Bristol Bay extending out to Unimak Island, but also extending into the North Pacific. Later (in June), these birds seemed to follow the recession of the ice north toward Bering Strait. He described some of these movements in considerable detail. Pigeon Guillemot (Cepphus columba) Two individuals of this species were seen from the "Brown Bear"; one bird just off Cape Lisburne and one at the southern limit of the pack ice, 70°50'N, 165°30'W. An unidentified guillemot was seen near the latter Pigeon Guil- lemot (Fig. 11). Apparently Shuntov (1961: 1059-1061) did not observe this species. Al- though Jacques (1930:357) observed it, ap- parently he saw it only in waters south of the Diomedes. Curiously, Jacques (1930:356-357) observed Black Guillemots (C. grylle) in con- siderable numbers north of Bering Strait even up to Herald Island, while none were observed on the "Brown Bear" cruise. Kittlitz’s Murrelet (Brachyramphus brevirostris) Three sightings of this species, totaling four birds, occurred in the open ocean north of Cape Lisburne. Another bird was seen close to shore in this area at about 69°50'N, 164°33'W (Fig. 11). Neither Shuntov (1961: 1058-1069) nor Jacques (1930:353-366) ob- served this species. Parakeet Anklet (Cyclorrhynchus psittacula) This species is doubtfully recorded from Kotzebue Sound. Several individuals were seen and tentatively identified as Parakeet Auklets (Fig. 11). Both Jacques (1930:356) and Shuntov (1961:1061) cited this species, but it is not clear in either case where the sightings were made. Shuntov (1961:1061) implied that this species, in common with Crested Auklets and Least Auklets, was seen near coast- lines but seldom in the open sea. Crested Auklet (Aethia cristatella) Two sightings of this species, totaling six individuals, were made about 18 miles west 345 of Cape Thompson, but most observations were made farther south. Hundreds were seen in Bering Strait near their breeding sites on the Diomedes. A single Crested Auklet was seen off Port Clarence (Fig. 11). Many auklets were observed on the voyage which could not be identified positively because of poor visibility. This was particularly true in Bering Strait near the Diomedes. Jacques (1930:356) observed this species near the Diomedes but did not definitely identify it farther north. Least Auklet (A. pusilla) Least Auklets were observed on August 26 only in and near Bering Strait, where they occurred in considerable numbers. Visibility was poor at the time, with waves up to 10 ft high and winds gusting up to 40 mph from the northwest, and accurate determination of abundance was not possible (Fig. 11). This is the only area in which Jacques (1930:356) observed them. Horned Puffin (Fratercula corniculata) This species was found almost everywhere that murres were found (Fig. 11) but in much smaller numbers. Horned Puffins did outnum- ber murres in the vicinity of Puffin Island in Kotzebue Sound, where the cliffs apparently support large numbers of puffins but few murres. Data on feeding areas are inconclusive, but it appears likely that puffins resemble murres in this respect. The observations from the "Brown Bear” are at variance with those of Shuntov (1961:1061) and Jacques (1930: 355) in that sightings were common far from shore. Tufted Puffin (Lunda cirrhata) This species was rarely seen except near Cape Thompson, Cape Lisburne, and the Ber- ing Strait area (Fig. 11). It is not an abundant breeder at Cape Thompson (Swartz, 1966). It is more numerous at Cape Lisburne and reached its greatest abundance in the vicinity of the Diomedes. Curiously, Shuntov (1961: 1059-1061) seems not to have observed this species. Jacques (1930:355) often did observe it in the Bering Sea and near the Diomedes, but did not list it north of the Diomedes. 3 46 Yellow Wagtail (MotaciUa flava) This Old World species has become well established as a breeding species in Alaska (Gabrielson and Lincoln, 1959:692) and mi- grates back and forth from the Asian mainland. Pelagic observations are to be expected in the migration season, but it is somewhat startling to make three such observations in early Au- gust (August 7, 10, and 13) and in such a pattern as to imply that the birds make little or no effort to move along shore to the point closest to Siberia before flying out over the sea (Fig. 10). To my knowledge, no other authors have reported this species from offshore. Water Pipit (Anthus spinoktta) The single doubtful pelagic observation of the Water Pipit is difficult to evaluate (Fig. 10). In view of the doubt which exists as to its identity, it is futile to speculate on the significance of the observation. REFERENCES Alexander, H. G, 1935. The movements of sea birds. Brit. Birds 29:298-299. Belopolski, L. O. 1957. Ecology of Sea Col- ony Birds of the Barents Sea. Akad. Nauk USSR. 460 pp. [Translated from Russian by R. Ettinger and C. Salzmann, Israel Prog, for Sci. TransL, Off. Tech. Serv., U. S. Dept. Comm., Washington, D.C.] Creager, J. S., and Dean A. McManus. 1961. Preliminary Investigations of the Ma- rine Geology of the Southeastern Chukchi Sea. Dept. Oceanogr., Univ. Washington, Tech. Rept. 68. 46 pp. Fay, F. EL, and T. J. Cade. 1959. An ecologi- cal analysis of the avifauna of St. Lawrence Island, Alaska. Univ. Calif. Publ. Zool. 63(2) : 7 3 — 1 50. Fleming, R. H. and Staff, i960. Second Oceanographic Survey of the Chukchi Sea, 26 July to 28 August I960. Preliminary Report of Brown Bear cruise 268. 9 pp. and Staff. 1961. Physical and Chemi- cal Data for the Eastern Chukchi and North- ern Bering Seas. Dept. Oceanogr., Univ. Washington, Tech. Rept. 69. 162 pp. PACIFIC SCIENCE, Vol. XXI, July 1967 and D. Heggarty. 1962. Recovery of Drift Bottles Released in the Southeastern Chukchi Sea and Northern Bering Sea. Dept. Oceanogr., Univ. Washington, Tech. Rept. 70. 18 pp. Gabrielson, I. N., and F. C. Lincoln. 1959. Birds of Alaska. The Stackpole Company, Harrisburg, and the Wildlife Management Inst, Washington, D.C. 922 pp. Grinnell, J. 1900. Birds of the Kotzebue Sound region, Alaska. Pacific Coast Avi- j fauna 1:1-80. Harrison, C. J. O. 1955. Ornithological ob- servations from Lista. Sterna 29:101-131. Jacques, F. L. 1930. Water birds observed i on the Arctic Ocean and the Bering Sea in 1928. Auk 47:353-366. Kuroda, N. I960. Analysis of sea bird distri- bution in the northwest Pacific Ocean. Pa- j cific Sci. 14:55-67. Salomonsen, F. 1944. The Atlantic alcidae. The seasonal and geographic variation of : the auks inhabiting the Atlantic Ocean and I the adjacent waters. Goteborgs Kungl. Vetenskaps-och Yitterhets-Samhalles Hand- | linger, Sjatte Foljden, Sen B. 3, 5:1-138. ! [Not seen.] Shuntov, V. P. 1961. Migration and distri- bution of marine birds in southeastern Ber- ing Sea during spring-summer season. [In j Russian.] Zoologicheskii Zurnal 40(7): 1058-1069. Sparks, A. K., and W. T. Pereyra. I960. Annual Report on Benthic Marine Inverte- brates of the Chukchi Sea. Project Chariot, j June 1, I960. Dept. Oceanogr., Univ. ! Washington. 59 pp. Storer, R. W. 1952. A comparison of varia- [ tion, behavior, and evolution in the sea bird j genera Uria and Ceppbus . Univ. Calif. Publ. ' Zool. 52(2) : 121— 222. Swartz, L. G. 1966. Sea-cliff birds. Chap. 23, pp. 611-678. In: Environment of the Cape Thompson Region, Alaska. U. S. ; Atomic Energy Commission, Oak Ridge. (pne-481, Fed. Clearing-house for Sci. and Tech. Inf.) Tuck, L. M. I960. The Murres. Queen’s j Printer, Ottawa. 260 pp. Vaughan, H. R. H. 1937. Flight speed of Birds in Bering and Chukchi Seas — Swartz 347 guillemots, razorbills, and puffins. Brit. Birds 31:123. Wolfe, J. N., Chmn. I960. Bioenvironmental Features of the Ogotoruk Creek Area, Cape Thompson, Alaska. Div. Biol. Med., U. S. Atomic Energy Commission, Washington, D. C. 69 pp. 1962. Bioenvironmental Features of the Ogotoruk Creek Area, Cape Thompson, Alaska. Div. Biol. Med., U. S. Atomic Energy Commission, Washington, D. C. 183 pp. Wynne-Edwards, V. C. 1935. On the habits and distribution of birds on the North At- lantic. Proc. Boston Soc. Nat. Hist. 40:233- 346. Branchial Muscles in Representatives of Five Eel Families1, 2 Gareth J. Nelson1 2 3 During the evolution of eels the gill arch skeleton of some forms was profoundly modi- fied in gross structure. The modifications are adaptive and are associated with body form and feeding habits of eels (Nelson, 1966). The present paper deals with the musculature at- tached to the gill arch skeleton of eels repre- senting five families, primarily of the suborder Anguilloidei. Six genera were chosen for study: Conger (Congridae), Anguilla (Anguillidae) , Morin gu a (Moringuidae) , Kaupichthys (Xeno- congridae), Uropterygius , and Gymnothorax (Muraenidae) . The gill arch skeleton of these forms shows a progressive reduction of ele- ments, showing probable stages in the evolu- tionary development of the specialized "pha- ryngeal jaws” of the morays — eels of the family Muraenidae. Gill arch musculature in eels has been studied previously only in Anguilla by Dietz (1912) and again by Kesteven (1943). Muscle ter- minology in the present work follows Vetter (1878) and Edgeworth (1935) as far as pos- sible. Names of gill arch elements are ab- breviated as follows: B, basibranchial; H, hypobranchial; C, ceratobranchial; E, epi- branchial; I, infrapharyngobranchial; LP, lower pharyngeal tooth plate; UP, upper pharyngeal tooth plate. Gill arches in eels are discussed in detail elsewhere (Nelson, 1966). MATERIAL AND METHODS Muscles were dissected in preserved adult specimens and illustrated for Conger mar- gin at us, Anguilla ro strata, Morin gua javanica, Kaupichthys diodontus, Uropterygius knighti, and Gymnothorax petelli. Observations on re- 1 This paper is part of a thesis submitted to the Graduate Division of the University of Hawaii in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Zoology. 2 Contribution No. 261, Hawaii Institute of Ma- rine Biology. Manuscript received May 6, 1966. 3 Present address: Department of Ichthyology, American Museum of Natural History, New York. lated genera (those listed in Nelson, 1966) suggested that the six genera selected for study are representative of the families or subfamilies to which they belong. All material was ob- tained from the collections of the Department of Zoology, University of Hawaii. With the exception of specimens of Anguilla, study material originally was collected by means of shallow water rotenone poisoning around Oahu and Christmas Island. The illustrations show the muscles in ap- proximately their relative size and positions. Muscles attaching to structures other than gill arches are shown transected. Occasionally other muscles are shown with parts removed to reveal underlying structures. Roots of the branchial arteries are included in the illustra- tions, for they serve as convenient landmarks, separating adjacent muscles. Bertmar (1962) and Petukat (1965) have dealt with the on- togeny and comparative anatomy of these ves- sels in some other teleostean fishes. RESULTS Conger Ventral muscles are shown in Figure 1 and listed in Table 1. Obliqui (01-3) occur on arches 1-3, extending between cerato- and hypobranchials. 03 has its insertion apparently transferred anteriorly to H2. A rectus (R4) is present only between arches 3-4, extending between the proximal ends of C4 and H3. It probably represents part of the oblique of arch 4 with the insertion anteriorly transferred to H3. A rectus communis (RC) extends from the proximal end of C4, with some of its fi- bers inserting on H3, others on H2 in common with 03. An anterior trans versus (AT) ex- tends between the proximal ends of C4 of either side. Extending posteriorly from C5, a single pharyngo-clavicularis (PC) attaches to the cleithrum. A posterior transversus (PT) extends between the distal ends of C5 of either side. An adductor (A5) joins the distal ends 348 Branchial Muscles of Five Eel Families — Nelson 349 TABLE 1 Muscles Attached to the Ventral Parts of the Gill Arches in Some Eels* GENUS Ol 02 03 Rl R2 R3 R 4 RC AT PT PC s VR A5 SP Conger X X X - - - X X X X X X X X - Anguilla X X X - X X X X X X X X X X - Moringua X X X - - - X X X X X X X X X Kaupichthys - - - - - X X X X X X X X X Uropterygius - X X X X - X Gymnothorax - - - - - - - - X X X X X - X * A5, Adductor 5; AT, transversus anterior; Ol-3, obliqui 1-3; PC, pharyngo-clavicularis; PT, transversus posterior; Rl-4, recti 1-4; RC, rectus communis; S, sphincter oesophagi; SP, subpharyngealis; VR, retractor ventralis. X, Muscle present; -, muscle absent. of C4-5. A sphincter (S) encircles the esopha- gus and also interconnects C5 of either side. Internal to the sphincter extend longitudinal fibers tending to separate anteriorly, forming a paired muscle, the ventral retractor (VR), attaching to LP and posteriorly extending some distance in the esophageal wall. Dorsal muscles are shown in Figure 2 and listed in Table 2. External levators (ELl-4) occur on arches 1-4, extending between the cranium and the proximal ends of El-4. In- ternal levators (ILl-2) occur on arches 1-2. ILl extends between the fascia of the trunk and 12. Inferior obliques (102-3) intercon- nect El -3. A small accessory oblique (AO) extends between El and 12. A superior oblique (SO) extends between E3 and 1 3. An adductor (A4) extends between E4 and C4. A posterior oblique (PO) extends between E4 and C5. The sphincter (S) encircles the esophagus and its anterior portion extends between the arches of either side. Internal to the sphincter occurs a longitudinal layer tending to separate an- terior, forming a paired muscle, the dorsal retractor, attaching to UP4 and posteriorly extending some distance in the esophageal wall. Anguilla Muscles are shown in Figures 3 and 4 and listed in Tables 1 and 2. They are rather simi- lar to those of Conger and have been studied by Dietz (1912) and Kesteven (1943), whose terminologies are compared with that used here in Tables 3 and 4. In the second arch, that portion of the oblique (02) inserting on Hi corresponds to a rectus (Table 1, R2). A posterior transversus (Table 1, PT) is repre- sented by the anteroventral portion of the sphincter (S). Morin gua Muscles are shown in Figures 5 and 6 and listed in Tables 1 and 2. A major feature of the musculature is the subpharyngealis (SP), a sheet of longitudinal fibers dorsal to the ven- tral arch elements. A transverse (Fig. 6, TD) is partly distinct from the anterodorsal part of the sphincter. Kaupichthys Muscles are shown in Figures 7 and 8 and listed in Tables 1 and 2. With some minor differences the muscles are most similar to those of Morin gua. TABLE 2 Muscles Attached to the Dorsal Parts of the Gill Arches in Some Eels* GENUS ELI EL2 EL3 EL4 IOl 102 103 AO ILl IL2 so PO A4 DR s MP LP PP Conger X X X X - X X X X X X X X X X - - - Anguilla X X X X X X X X X X X X X X X - - - Moringua X X - X X X X - X X X X X X X - - - Kaupichthys X X X - X X X - X X X X X X X - - - Uropterygius - - Hj - - X X - X - - - X X X X - X Gymnothorax X X X - X X X - X X - - X X X X X - * A4, Adductor 4; AO, obliquus inferior accessorius; DR, retractor dorsalis; ELl-4, levatores externi 1-4; ILl-2, levatores interni 1-2; 101-3, obliqui inferiores 1-3; LP, protractor lateralis; MP, protractor medialis; PO, obliquus pos- terior; PP, protractor posterior; S, sphincter oesophagi; SO, obliquus superior. X, Muscle present; -, muscle absent. Fig. 1. Conger marginatus, ventral gill arch muscles of left side, ventral view, showing roots of afferent branchial arteries. A3, Adductor 5; AT, transversus anterior; 01—3 , obliqui 1-3; PC, pharyngo-clavicularis ; PT, transversus posterior; R4, rectus 4; RC, rectus communis; S, sphincter oesophagi ; VR, retractor ventralis. Branchial Muscles of Five Eel Families — Nelson 351 ELI Fig. 2. Conger marginatus, dorsal gill arch muscles, dorsal view of left side, showing parts of efferent branchial arteries. A4, Adductor 4; AO, obliquus inferior accessorius; ELl-4, levatores externi 1-4; ILl-2, levatores interni 1—2; 102—3, obliqui inferiores 2—3; PO, obliquus posterior; S, sphincter esophagi; SO, obliquus superior. 352 PACIFIC SCIENCE, Vol. XXI, July 1967 03 Fig. 3. Anguilla rostrata, ventral gill arch muscles, ventral view, with some of those of right side omitted, and showing roots of afferent branchial arteries. Muscles as in Figure 1. Branchial Muscles of Five Eel Families — Nelson 353 Fig. 4. branchial 101 ELI SO Anguilla rostrata, dorsal gill arch muscles, dorsal view of left side, showing parts of efferent arteries. Muscles as in Figure 2. 354 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 5. Moringua javanica, ventral gill arch muscles, ventral view, with some of those of right side omitted, showing roots of afferent branchial arteries. SP, Subpharyngealis. Other muscles as in Figure 1. Branchial Muscles of Five Eel Families — Nelson 355 Fig. 6. Morin gua javanica, dorsal gill arch muscles, dorsal view, with some of those of right side omitted, showing parts of the efferent branchial arteries. DR, Retractor dorsalis; TD, transversus dorsalis. Other muscles as in Figure 2. 356 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 7. Kaupichthys diodontus, ventral gill arch muscles, ventral view, with some of those of the right side omitted, and showing roots of afferent branchial arteries. Muscles as in Figures 1 and 5. Branchial Muscles of Five Eel Families — Nelson 357 Fig. 8. Kaupichthys diodontus, dorsal gill arch muscles, dorsal view, with some of those of right side omitted, and showing parts of efferent branchial arteries. Muscles as in Figure 2. 358 PACIFIC SCIENCE, Vol. XXI, July 1967 TABLE 3 Muscle Terminology for Ventral Gill Arch Muscles in Anguilla PRESENT TERMINOLOGY DIETZ (1912) KESTEVEN (1943) Obliquus 1 (01 ) Obi. I Subarc. obi. 1 Obliquus 2 (02) Obi. II Subarc. obi. 2 Obliquus 3 (03) Obi. Ill Subarc. obi. 3 Rectus 2 (R2) Interarc. I/ll none Rectus 3 (R3) Interarc. II/III none Rectus 4 (R4) Interarc. III/IV none Rectus communis (RC) Interarc. I/IV Rectus Trans versus anterior (AT) Trans. IV Transversus Transversus posterior (PT) Trans. V Transversus Pharyngo-clavicularis (PC) Phar.-clav. ex. + int. Clav.-pharyng. ext. Adductor 5 ( A5 ) none none Retractor ventralis (VR) none none Sphincter oesophagi (S) Pharynxmuskulatur Sphincter oesophagi TABLE 4 Muscle Terminology for Dorsal Gill Arch Muscles in Anguilla PRESENT TERMINOLOGY dietz (1912) KESTEVEN (1943) Levator externus 1 (ELI) Lev. ext. I none Levator externus 2 (EL2) Lev. ext. II Lev. ext. 2 Levator externus 3 (EL3) Lev. ext. Ill Lev. ext. 3 Levator externus 4 (EL4) Lev. ext. IV Lev. ext. 4 Levator internus 1 (ILl ) Lev. int. I? Retractor dorsalis Levator internus 2 (IL2) Lev. ext. IV Lev. int. Obliquus inferior 1 (IOl) Obliq. inf. hy/I none Obliquus inferior 2 (102) Obliq. inf. I/II none Obliquus inferior 3 (103) Obliq. inf. II/III none Obliquus inferior accessorius (AO) none none Obliquus superior (SO) Obliq. sup. ant. Epiarc. obi. 1 Obliquus posterior (PO) Obliq. dors. post. Epiarc. obi. 2 Adductor 4 (A4) none none Retractor dorsalis (DR) none none Uropterygius The ventral muscles are similar to those of Kaupichthys but are reduced in number (Ta- ble 1). The dorsal muscles are shown in Fig- ure 9. Levators are lacking, having been replaced it seems by two new muscles, one extending between the cranium and UP3-4, the medial protractor (MP), the other extend- ing between the cranium and E4, the posterior protractor (PP). Gymnothorax The ventral muscles are similar to those of Uropterygius with the exception of the sub- pharyngealis, which appears subdivided into many parts (interbranchial attractors). These are shown in Figure 10. The dorsal muscles (Fig. 11) resemble those of Kaupichthys more than those of Uropterygius. However, they do include a medial protractor. They include in addition a lateral protractor extending between UP3-4 and the ventral part of the hyoid arch, attaching there in common with the ventral muscles LAl, OAl-2. The dorsal retractors attach in common with the ventral retractors to the ventral surfaces of the 1 3 th— 1 5 th vertebrae. Branchial Muscles of Five Eel Families — Nelson 359 Fig. 9. branchial PP U ropterygius knighti, dorsal gill arch muscles, dorsal view of left side, showing parts of efferent arteries. MP, Protractor medialis; PP, protractor posterior. Other muscles as in Figure 2. 360 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 10. Gymnothorax petelli, some of the gill arch muscles of the left side, ventral view, showing sub- divisions (attractores interbranchiales and sphincteres branchiales) of the subpharyngealis. IAl-2, Attractores intermediates 1-3; LAl-4, attractores laterales 1-4; MAl-3, attractores mediates 1-3; OAl-2, attractores obliqui 1-2; SRl-4, sphincteres branchiales 1-4. Branchial Muscles of Five Eel Families — Nelson 361 LP Fig. 11. Gymnotloorax petelli, dorsal gill arch muscles, with some of those of the right side omitted, and showing parts of efferent branchial arteries. LP, Protractor lateralis; MP, protractor medialis. Other muscles as in Figure 2. 362 PACIFIC SCIENCE, Vol. XXI, July 1967 DISCUSSION If they are represented by the rather linear reduction in gill arch elements, relationships among the examined genera may be as dia- grammed in Figure 12. Thus, Conger would be the most primitive and Gymnothorax the most advanced. Gill arch muscles of Conger are not structurally far removed from those of Elops (Nelson, 1967) or those of other gen- eralized lower teleostean fishes (Vetter, 1878; Dietz, 1912, 1914, 1921; Greene and Greene, 1913). On the other hand, the muscles of Uropterygius or Gymnothorax are far removed structurally from those of Conger and conse- quently appear to be advanced rather than primitive. The series of studied forms ranging from Conger to Gymnothorax shows a progres- sive series of muscle modifications, involving the loss of some muscles and the appearance of others (Tables 1 and 2). The series of mus- cle modifications in a general way parallels the linear reduction in gill arch elements of these forms. Particular modifications of gill arch muscles in eels seem correlated with particular modifi- cations of the gill arches themselves. Reduction of ventral musculature (obliqui and recti) parallels reduction in ventral arch elements Fig. 12. Diagram of possible relationships among some eels. (basi- and hypobranchials) . Appearance of protractors and enlargement of the retractors and their attachment to the vertebral column in muraenines parallel the enlargement of the fourth arch and the tooth plates it supports. The appearance of the subpharyngealis is not so easily correlated with any particular modification of the gill arch skeleton. It ap- pears, seemingly, in place of the obliqui and recti. However, nothing is known of its em- bryonic development and it may or may not represent modified obliqui or recti. Its position is distinctive, being internal to the skeletal elements rather than external as are the obliqui and recti. Probably the ventral musculature shifted from a relatively external to a relatively internal position with the reduction and loss of basibranchials. In any event, it assumed a sheetlike form, gradually encroaching upon the gill slits, which in the more advanced eels (e.g., the muraenids) are reduced to small round openings. Dorsal and ventral paired retractor muscles are present in all of the eels examined. In most forms they are only partly distinct sub- divisions of the inner longitudinal muscle layer of the anterior esophagus. In eels of the sub- family Muraeninae, they acquire an attachment to the vertebral column. The taxonomic significance of retractors in other groups of bony fishes has been dealt with by Dietz (1912, 1914, 1921) and Holst- voogd (I960, 1965). According to Nelson (1967), retractors probably have developed an attachment to the vertebral column independently in many evolutionary lineages of bony fishes. Probably in each lineage they are associated with and constitute part of an improvement or specialization in the feeding I mechanism. It is hardly to be doubted that the attach- ment to the vertebral column has been acquired independently among the eels. No other group having retractor muscles has both dorsal and ventral retractors attaching in common to the vertebral column. Indeed, except among eels, ventral retractors seem to be lacking. Thus, the attachment to the vertebral column of muraenines no doubt is another example of independent development. In this case they attach to the tooth plates of the pharyngeal Branchial Muscles of Five Eel Families — Nelson 363 jaws and apparently constitute, with the pharyn- geal jaws, an advancement or specialization in the feeding mechanism. The common course and attachment of the retractors, both ventral and dorsal, to the ver- tebral column in morays are evidence that the muscles act together, simultaneously retracting both upper and lower tooth plates. Indeed, the construction of the plates and their supporting bones prohibits independent movement of the ventral and dorsal plates. Protraction probably occurs through the contraction of the sub- pharyngealis and the dorsal muscles joining the cranium and gill arches. Protraction and retraction probably succeed one another during the swallowing of prey. It is likely that, in the morays, the pharyngeal jaws and the mus- cles attached to them enable these forms to transport relatively large prey from the jaws into the esophagus, a distance which in eels is secondarily lengthened (Nelson, 1966). Thus, these structures appear to be adapted to the known predatory habits of the morays. SUMMARY 1. Branchial muscles are described for six genera representing five eel families: Conger (Congridae), Anguilla (Anguillidae) , Morin- gua (Moringuidae) , Kaupichthys (Xenocon- gridae), Uropterygius and Gymnothorax (Muraenidae) . 2. In the examined forms, muscles as well as gill arches suggest stages in an evolutionary sequence, with the Congridae being the most primitive and the Muraenidae being the most advanced. 3. Dorsal and ventral retractor muscles oc- cur in all species examined. In eels of the sub- family Muraeninae they acquire secondarily an attachment to the vertebral column. 4. Gill arch muscles and pharyngeal jaws of muraenids are adaptations probably enabling these fishes to swallow large prey. REFERENCES Bertmar, G. 1962. On the ontogeny and evo- lution of the arterial vascular system in the head of the African characidean fish Hep- setus odoe. Acta Zool. 43:255-294, 12 figs. Dietz, P. A. 1912. Vergelijkende Anatomie van de Kaak- en Kieuwboogspieren der Teleostei. Leiden. 196 pp., 25 figs. • 1914. Beitrage zur Kenntnis der Kiefer- und Kiemenbogenmuskulatur der Teleostier. I. Die Kiefer- und Kiemen- bogenmuskeln der Acanthopterygier. Mitt. Zool. Stat. Neapel 22:99-162, 45 figs. 1921. Uber die systematische Stellung der Gadidae. Zugleich Nr. 2 der "Beitrage zur Kenntnis der Kiefer- und Kiemenbogen- muskulatur der Teleostier.” Mitt. Zool. Stat. Neapel 22:433-457, 14 figs. Edgeworth, F. H. 1935. The Cranial Muscles of Vertebrates. Cambridge Univ. Press, viii -j- 300 pp., 841 figs. Greene, C. W., and C. H. Greene. 1913. The skeletal musculature of the king salmon. Bull. Bur. Fish. 33:25-59, 14 figs., 2 pis. Kesteven, H. L. 1943. The evolution of the skull and the cephalic muscles. A compara- tive study of their development and adult morphology. Part I. The fishes {continued') . Mem. Austral. Mus. 8:63-132, 69 figs. Holstvoogd, C. I960. The importance of the retractores arcuum branchialium for the classification of teleostean fishes. Bull. Aquat. Biol. 2:49-50. 1965. The pharyngeal bones and muscles in Teleostei, a taxonomic study. Proc. Konikl. Nederl. Akad. Wetens., Ser. C, 68:209-218, 12 figs. Nelson, G. J. 1966. Gill arches of teleostean fishes of the order Anguilliformes. Pacific Sci. 20:391-408, 58 figs. 1967. Branchial muscles in some gen- eralized teleostean fishes. Acta Zool. [In press.] Petukat, S. 1965. Uber die arteriellen Gefass- stamme bei den Teleostiern. Zool. Beitr. 11:437-515, 34 figs. Vetter, B. 1878. Untersuchungen zur ver- gleichenden Anatomie der Kiemen- und Kiefermusculatur der Fische. Jena Z. Natur- wiss. 12:431-550, pis. 12-14. Acoustical Behavior of the Menpachi, Myripristis herndti, in Hawaii1 Michael Salmon2 ABSTRACT : The menpachi ( Myripristis berndti) is found in aggregations inside caves and under ledges during the day in water more than 3 m deep. Diel tape recordings in these areas showed that the fish produced four types of sounds (knocks, growls, grunts, and staccatos), with no crepuscular peaks, from dawn to dusk. At night, when the fish scattered to feed, few sounds were detected. A fifth sound was produced when fish were hand-held. The sound-producing mechanism was determined by a series of ablation experiments on hand-held fish. It consisted of a pair of bilateral muscles attached to the skull anteriorly and the air bladder, the first two dorsal ribs, and the cleithrum bone posteriorly. Populations of 6-7 fish were maintained in the laboratory in large tanks with an artificial cave. They remained inside the cave during the day but swam actively throughout the tank at night. Brief chasing of a small fish by a larger, accompanied by knocking sounds, was frequently observed. Growl sounds were produced during more intense aggressive interactions between two fish of about the same size. There was no evidence of territoriality by members of any population. Few grunt or staccato sounds were produced when various species of nonpreda- tory fish were introduced among laboratory populations. Many of these sounds were elicited when moray eels were introduced. Sound playbacks to four populations from one of two speakers on either side of the cave elicited different responses depending on the sound tested. All fish immediately turned to and moved toward the experimental speaker when grunt or staccato sounds were played. Some fish briefly turned to the experimental speaker when knocks were emitted but none moved to the source. There was no detectable change in behavior when background noise was played back. Three fish tested in an aktograph showed increases in locomotory activity at night which corresponded with periods of nocturnal scattering and feeding in field populations. The acoustical system of the menpachi is compared with that of the longspine squirrelfish, Holocentrus rufus, an Atlantic species. The "Menpachi” consist of four species of economically important fishes in the Hawaiian area. Although their habits are well known to trap- and spearfishermen, there have been few published studies on their ecology and none on their acoustical behavior. In this report the be- havior correlated with or stimuli eliciting four types of sounds (grunts, staccato, knocks, and growls) produced by Myripristis berndti (Jor- dan and Evermann) are described. A fifth 1 Contribution No. 267, Hawaii Institute of Marine Biology. Manuscript received March 7, 1966. 2 Department of Biological Sciences, De Paul Uni- versity, Chicago, Illinois. sound, produced when fish were hand-held, ; was physically analyzed in conjunction with experiments to determine the sound-producing mechanism. Diel patterns of locomotory and feeding activity in nonreproductive groups of M. berndti , and their relationship to sound production were determined by field and lab- oratory observations. Experiments were carried out to determine the response of laboratory populations to playbacks of their own sounds and to other fish species commonly associated with them in their coral reef community. It has been known for many years that several species of squirrelfishes (family Holo- 364 Acoustical Behavior of Myripristis berndti — Salmon 365 centridae) produce sounds. Studies to date have been made on two species in the genus Holo- centrus. Fish (1948) first reported sound pro- duction in the group. Moulton (1958) studied H. ascensionis in Bimini and described two types of sounds produced in the field, a single sound (the grunt) and one composed of several thumplike sounds produced in a series (the staccato). Winn, Marshall, and Hazlett (1964) were the first to study the significance of these sounds experimentally. They found that the nonreproductive social organization of H. rufus, which produced the same types of sounds as H. ascensionis, was territorial. When a conspecific individual entered the territory of another squirrelfish, the resident produced many grunt sounds and rarely staccatos, some- times acompanied by fin erection, nipping, and lateral displays in which the two fish moved parallel to each other. Intruders of other species elicited both staccatos and grunts, but more staccatos were produced toward larger fish or a potential predator, such as a moray eel. Lab- oratory populations were maintained in large tanks and each fish defended a territory con- sisting of the inside of a large can, open at one end, and the area immediately before the open- ing. When staccato sounds were played back to these populations from one of two speakers on each side of the tank, the fish at first retreated into their cans. Some then swam to the sound source, while others turned their heads toward the speaker from just outside the can, indicating that the fish were probably able to localize the source of sound. Diel recording showed that more sounds were produced during the day than at night, when the fish were active and feeding. Peaks in sound production occurred at dawn and dusk. It was hypothesized that the peaks were caused by movements of nocturnal and diurnal species into and out of the reef and through the territories of squirrelfishes under conditions of reduced light intensity. Moulton (1958) stated that contractions of the body wall musculature associated with the first three ribs and the air bladder were re- sponsible for sound production in H. ascen- sionis. In a series of ablation experiments, Winn and Marshall (1963) showed that the muscles involved in sound production were bilateral and attached to the posterior part of the skull, the air bladder, and the first two dorsal ribs in H. rufus. Removal of one muscle reduced the intensity of sounds produced by hand-held specimens, but did not significantly change sound duration or number of pulses per sound, indicating that the two muscles con- tracted simultaneously to produce each sound. Gainer, Kusano, and Mathewson (1965) studied the electrophysiological and mechanical properties of the sound-producing muscle in the same species. The muscle was capable of con- tracting at a frequency of 100/second with no mechanical summation, while fast white muscle from the same fish showed considerable summation at 50/second. Myripristis is the second largest genus in the family. These fish live in schools and move over the reef more than do members of the genus Holocentrus (Herald, 1961), which are soli- tary-territorial. Other reports indicate that the schools remain in caves or under ledges during the day and scatter to feed at night (Hobson, 1965). The presence of sand-dwelling annelids in the stomachs of M. berndti from the Mar- shall Islands indicated that the fish move to open areas, away from the reef during noc- turnal feeding (Hiatt and Strasburg, I960). There have been no published studies on the acoustical behavior of any species in this genus. Nelson (1955) described the antero-bilateral projections of the air bladder which, in M. argyromus, completely covered the auditory bullae and were thus more extensively modified, presumably for an auditory function, than in H. ascensionis and H. rufus. MATERIALS AND METHODS All observations and experiments were carried out at Oahu, Hawaii, from February to July 1965. Most of the field observations were made in Pokai Bay, Waianae, in water 3-9 m deep. The study area spanned a 1-km distance along the coast. Several other schools were observed in similar habitats offshore at Black Point and Ilikai Harbor. The topographic features of the habitats and estimates of school size in number of fish were recorded with the aid of an under- water flashlight and drawing pad or were pho- tographed directly with a Nikonos underwater camera. 366 All tape recordings were made with an Uher 4000-S Report recorder and an Atlantic Re- search Corp. hydrophone (Model LC-57). Field recordings were carried out by securing a boat with three anchors over the reef area containing a school of fish. The hydrophone was placed inside a cave or under a ledge within 1 m of the fish and was secured with a weight. A small air-filled bottle was attached to the hydrophone cable about 1 m from the water surface to keep the cable taut and prevent entanglement in the reef. Field recordings were made at tape speeds of 2.3 cm/sec Q§ i.p.s.); laboratory recordings were made at 9.5 cm/sec (3f i.p.s.). Specimens 12-20 cm in total length were caught by hook and line or in traps and brought into the laboratory for study under more con- trolled conditions. They were established in groups of 6-7 fish in 756-liter fiberglass tanks with a plexiglas front, in which a "cave” was constructed with two building blocks covered with a piece of masonite (Fig. 1). Holes in the blocks allowed the fish to enter and leave through the side as well as through the front of the cave. A continuous flow of fresh sea TOP FRONT Fig. 1. Top and front views of tank in which laboratory populations were maintained, showing out- side dimensions. 1, Roof of cave; 2, underwater speakers used in sound playbacks; 3, building blocks with two holes through which fish could enter and leave the cave through the side as well as by the front opening; X, position of hydrophone. PACIFIC SCIENCE, Vol. XXI, July 1967 water circulated through the tank at tempera- tures between 21° and 23° C. The hydrophone was suspended in front of the cave to record sounds. Behavior correlated with sound produc- tion was described immediately after recording the sounds. Sound playbacks were carried out. One under- water speaker (University MM-2L) was placed on each side of the cave. The sounds used for playbacks were all recorded from previous pop- ulations of M. berndti. They were played through one of the speakers from a continuous loop of tape on a Crown tape recorder (Model CR-25) which repeated the entire playback every 11 seconds. The response of four popu- lations to grunt, staccato, and knock sounds was determined. Each type of sound was played back once in a random order to each group of fish and at levels comparable to those emitted by the fish. The number of fish on the left or right side of the tank was determined every 15 sec- onds of a 5 -minute period with sounds played back during minutes 2 and 4 from one speaker, selected randomly. Recordings were made dur- ing the entire 5 -minute period to monitor play- backs and record any sounds produced by the fish. One observer (the recorder) noted the po- sition of the fish in the tank. Another, shielded from both the fish and the recorder, turned the sound on and off through one of the two speak- ers. The recorder had no prior knowledge of which speaker was being used during the test although the response of fish to certain sounds enabled him to determine the experimental speaker with 100% accuracy. Various species of fish commonly associated i with M. berndti in the field were introduced in a random order to seven individual populations, j These were: Myripristis berndti, M. argyromus, 'I Holocentrus xantherythrus, Priacanthus meeki, Parupeneus porphyreus, and Gymnothorax un- I dulatus. The type and number of sounds pro- !] duced by the populations were recorded for a 1 -minute period before and during the intro- ] duction. Patterns of locomotory activity were deter- j mined for three fish, one for 24, one for 56, and one for 72 hours. The fish were placed , singly in a large doughnut-shaped chamber 7.6 cm wide, 9.5 cm deep, and with a mean swim- j ming circumference of 87.6 cm. Fresh aerated Acoustical Behavior of Myripristis berndti — Salmon 367 sea water circulated through the chamber at all times. Two Pflueger Fish Finders (Enterprise Manufacturing Co.), placed 130° apart and po- sitioned to face toward the center of the cham- ber, were used to detect the movement of the fish. The fish finder emits an 800-kc signal as a narrow beam across the chamber. The re- flected signal is identical to the emitted one when no moving object is present and, when the signals are compared (heterodyned) in the receiver,, they cancel out. Movement of a fish past the fish finder shifted the frequency of the reflected signal and caused a deflection on the chart of a Rustrak event recorder (Model 92). The chamber was placed in a small room within 1 m of a large window, so that the fish was ex- posed to normal changes in the daily light cycle. For further details concerning the apparatus, see Muir et al. (1965). A series of ablation experiments was carried out to determine the sound-producing mecha- nism. All fish produced grunts when hand-held by the caudal peduncle. Sounds of normal hand- held fish were recorded, followed by recordings of the same fish (record level on tape recorder left constant) after removal of the following: one or both sound-producing muscles; other associated muscles and bones; the gas from the swim bladder. All fish were held about 7.5 cm from the hydrophone. Operated fish were anes- thetized with MS-222. A few muscle potentials were recorded from the sound producing mus- cle of two fish with a Tektronix Low Level Amplifier (Type RM-122) and oscilloscope (RM-504) and were photographed with a Grass camera (Model C-4). The sound duration, number of pulses, and interpulse intervals were measured by photographing the recorded sounds from a Fairchild oscilloscope (Model 701) with the Grass camera, at film speeds of 100-500 mm/sec. The effect of operations on the intensity of sounds was determined. A General Radio Co. Impact-Noise Analyzer (Type 1556-B) was connected to the output of the tape recorder and a peak sound pressure value was deter- mined for a normal fish. The peak sound pres- sure of the same fish after the operation was also obtained. The peak value for the normal sound was considered as 0 decibel, while the value for the operated fish was considered as positive db (if the value exceeded that of the normal fish) or negative db (if the value was less). Relative sound pressures at various octave band frequences were also measured. The out- put of a General Radio Co. Octave Band Noise Analyzer (Type 15 58- A) was connected to the imput of the impact analyzer. A sine wave of 400 cps was applied to the imput of the octave band analyzer when set in the "all pass” posi- tion, and with the preamplifier in the 20 Kcs weighting (essentially flat response from 20 cps to 40 Kcs). The impact analyzer was then cali- brated to give a peak value 3 db higher than the root mean square value shown by the octave band analyzer for the sine wave. After cali- bration, the fish sounds from the tape recorder were applied to the imput of the octave band analyzer and readings were determined from the impact analyzer. The loudest of the first five sounds produced by a normal fish was measured and considered as 0 db. All sound pressures in various octave band frequencies of the first five sounds produced before and after operations on this fish were compared with the 0 db value. The sound pressures of all filtered signals were always less than the 0 db value. The reduction was measured and expressed in decibels. All sound pressures obtained from the impact ana- lyzer were relative to 0.0002 microbar. RESULTS The Sound-Producing Mechanism Sounds produced by hand-held specimens were accompanied by vibrations which could be felt along an area extending from the dorso- lateral region of the skull to the side of the body just lateral to the air bladder. The most intense contractions were in the dorsal region behind the eye. Removal of some of the super- ficial muscles, opercula, and part of the supra- scapular bone revealed a band of muscle slightly yellow in appearance, which could be observed to contract synchronously with the production of sound. The muscle was attached to the pos- terior part of the skull, just above the eye, and passed over the anterior lobes of the air bladder to its insertion point above the area where the main body of the air bladder gives rise to the lobes (Fig. 2). At its insertion, the muscle was attached medially to the first two dorsal ribs and 368 PACIFIC SCIENCE, Vol. XXI, July 1967 Fig. 2. Anatomy of the sound-producing mech- anism and surrounding bones in Myripristis berndti. 1 , Main body of swim bladder; 2, sound-producing muscle; 3, dorsal portion of clei thrum bone with tendon attached to sound-producing muscle; 4, ante- rior lobe of swim bladder; 5, preoperculum; 6, scapula. the air bladder. A small tendon connected the muscle to the cleithrum bone laterally. Another small, flat muscle (not shown in the figure) at- tached to the skull and ran between the sound - producing muscle and the anterior lobe of the air bladder, to the operculum. This muscle was routinely cut during ablation experiments, with no apparent effect on sound production. The sound-producing muscle was highly vascularized and appeared to be composed of three distinct myomeres. Removal of one sound-producing muscle resulted in a relative decrease of 2-7 db in operated fish, when compared with their own normal sounds (Table 1). Sound pressures were reduced in all octave bands but were greatest in the 75-150 cps band. Normal sounds contained frequences below 75 cps to under 4,800 cps, with most energy between 300-600 cps. Oscillographs of these sounds are shown in Figure 3. The temporal patterns of the pulses within these sounds are shown in Table 2. Normal fish produced sounds composed of 7-10 pulses (mean, 8.2). Operated fish showed more vari- ability in pulse range (6-11), and a mean value of 9.04 pulses per sound. Increases in number of pulses were correlated with increases in sound duration. Interpulse intervals were variable, but in most sounds the intervals be- tween the penultimate and the last pulse were greater than between other pulses. A few mus- cle potentials recorded from two fish were composed of 6-8 spikes (Fig. 3). The inter- spike intervals and total duration for a series of spikes were comparable to values for inter- pulse and total-duration measurements of sounds with the same number of pulses. The effect of removing the superficial mus- cles and bones near the sound-producing mus- cle is shown in Table 3 and Figure 3. The peak pressure of sounds produced by operated fish, when compared with pressures of their own sounds before the operation, increased in one fish, decreased in two, and remained the same in two fish. Five fish in which both sound-producing muscles were removed produced no audible sounds. The role of the air bladder in sound pro- duction was determined by replacing the gas in the bladder with sea water. Five fish, in which a small hole had been punctured in the lobe of the air bladder with a syringe, con- tinued to produce sounds at intensities com- parable to their own normal sounds (mean peak sound pressure =1.1 db above normal fish). Only a few bubbles of gas escaped through the puncture. When the puncture was held open the intensity of the sounds decreased as gas escaped until finally, when the air blad- der was completely filled with water, no au- dible sounds were produced although the muscles could still be felt to contract. The presence of only a small bubble of gas in the bladder resulted in production of sounds of very low intensity. Field Observations and Diel Recordings At least 20 different schools of menpachi were found in the Pokai Bay area. In all cases, these were mixed assemblages of M. berndti and M. argyromus, from 13 to 23 cm in total length. In shallower waters, M. argyromus pre- dominated. Both species were found to pro- duce the same types of sounds and to have similar nocturnal-diurnal activity patterns in Acoustical Behavior of Myripristis berndti — Salmon 369 W PQ W §S 1 rCN | VCN o m i m 1 ON NO ON m NO CM oo m 00 CN CM CM cO CM m _Q -d r- _Q -a _Q T3 Mi" -O vCN J& no CM I TO -d -d m 0 fN 'TXT OO O »— 1 ^ ^ nt 2 TN 1-1 CM 00 CM ITS l/N CM © \TN 00 © 00 rH © cCN P" Mp © © © CCN cCN 00 1-1 CM CM rCN cCN p- O rH rH rH © oCi © rCN © vo 00 © CM 00 vo © © CM rH CM CM rH rCN CM rH © mt VO CT) o © \r\ 00 © © CN Mr ON l/N rCN © © ON © © m rH rH m rH CM rH i/n VO VO VO ON © © i r\ o 00 MT 00 CM rCN © CM 00 ON PH r-t rH m /N VN IT\ o l/N rCN © © CM CM vo p~ © © vo 1/N rH © 00 rH m ITS rH rCN CM rH cCN m pH MT VO Mr 00 v-\ © © VO CM cCN © VO © © vo © P- CM r-l rH rH m rOi rH CM r_l CCN r_l CM CM 00 00 o ON VO rH ON fH. © © ON CM MT MT 00 © vo rCN © © 1-1 rCi VO fCN CM rCN o VO rTi p- rH rCN 00 © CM 00 l/N © CM VO © PH r_l p- CM rH o o o 00 MT CN rH © © CM VO © rCN P- rH © © :'p’ CM 1-1 cCN T_l r_l 1-1 CM GN © o mt CM o © CN rCN © © rCN rCN © MT © ON 00 P- r-1 CM 1-1 1-1 CM 00 o o CM cCN © CM P- © © cCN rCN © Pr © ON l/N l/N <£> r-t CM 1-1 1-1 1-1 rCN P" o CN o © MT 00 © CM 00 CM NT v/~\ VO PH CM i/N 00 1-1 VO CM rCN CM T4 CM VO o ON © MT p- © O p~ ON l/N i/-\ 00 P^ 00 \r\ Mr Q W Q .2 u d »H t/5 o o >, O • H V a j O u w PS nD d M3 c d o c .c "£ ' 4 a -5 O U Lh / *£ * d IT, d o a, W5 u * ''****' t -d c/3 h-1 Sunrise and sunset, July 1, 1965, at Oahu, Hawaii, were 5:53 AM and 7:18 pm, respectively. 374 PACIFIC SCIENCE, VoL XXI, July 1967 S g 2 3 % 3 3 ^ 3 CQ O 1 CK 2 8^ ■3 Q C .3 tf O I a u tyO“S a Sb/ hO P t V-i t-H Un t— < *h 0 as S s Pi S 3 8i IS * ,Si -2^ s' Cl CL, *5 « 2 CT JX ci ^TJ &0 3 c y «H r^d w ; | 6 ;| 2 i O Ph l (N * Constituents in Skipjack Muscle and Blood — Sather and Rogers 409 The data contained in Tables 2 and 3 were analyzed statistically to determine whether the concentrations differed with muscle type; these results are presented in Table 4. The white muscle contains a greater amount of water, po- tassium, and magnesium which possibly indi- cates that this tissue has a larger intracellular space. Statistical differences in the calcium and chloride contents were not found, which may have been due to the TCA used in the extrac- tion. However, blanks were carried throughout the analysis. Figure 1 illustrates that the extracellular vol- ume of the red muscles is greater than that of the white. The calculated inulin spaces for the red and white muscles are 23.79 and 18.97%, respectively. The higher sodium content of the red muscle verifies the larger extracellular space. It was not possible to determine accurately the intracellular sodium and chloride contents because of their high content in the intravas- cular compartment. The other intracellular cal- culated concentrations (mEq/liter cell water) of the red and white muscles, respectively, were: potassium = 160.88, 189.28; magne- sium = 33.76, 43.68; calcium = 0.12, 6.30. It is well known that the amount of intra- cellular potassium determines the threshold value for any tissue. Thus, the potential pro- duced by this ion for the red muscle was calcu- lated to be 83.77 mv and that of the white muscle was 88.08 mv — a potential difference of 4.31 mv. Therefore, the red muscle would have a lower threshold value, indicating that possibly this muscle would be utilized more than the white. Vernick (1964) reported that the red muscle of four pelagic species had a higher content of thiamine, riboflavin, pantothenic acid, vita- min B12, myoglobin, and cytochrome C. This tissue also had a higher degree of vasculariza- tion and larger mitochondria in the sarcoplasm. These findings led to the suggestion that the red muscle provided energy for the white. Hamamoto and Hohl (personal communica- tion) discovered that the mitochondrial density in the red muscle sarcoplasm of K. pelamis was approximately one magnitude greater than that in the white. Because the mitochondria are the cell’s energy producers, there is a strong correlation between the degree of activity of the muscle and the number and shape of the mitochondria within the muscle cells (Davson, 1964). In addition, if one considers the color of the two muscle types and applies the analogy of the breast muscles of chickens versus those of the pigeon, it becomes apparent that the red muscle of K. pelamis with its abundance of mitochondria is possibly used for swimming and not as an energy producer for the white muscle. The red muscle is, indeed, able to con- tract and is probably used for normal swimming activity (Rayner, personal communication). The white muscle may be used secondarily, e.g., for accelerated and rapid movements seen during avoidance and feeding reactions. Table 5 lists some of the plasma constituents of various fishes. It is well known that the marine cyclostomes are approximately isosmotic to the medium and that the marine cartilagin- ous fishes are hyperosmotic to the environment. However, the sea water-inhabiting teleosts are hyposmotic to their medium. Thus, these ani- mals are threatened by desiccation. To prevent dehydration the animals must drink water and selectively excrete ions. The latter process is generally accomplished extrarenally via the gills. Of the teleosts listed in Tables 5 and 6 only the barracuda and herring can be comparable to the skipjack, and the eels would be inter- mediate in comparison; the other species would be least comparable due to their phylogenetic placement and their relative inactivity as com- pared with the scombroid fishes. The mackerel is a scombroid fish, but it inhabits more inshore waters than does the skipjack. As expected, the electrolyte composition of the skipjack plasma (Table 5) is less than that of the cyclostomes. However, it approximates those of the chondrichthyes. The greater os- molality of the latter is due to a higher urea content of the plasma. Concentrations of 300- 400 mM of urea and trimethylamine oxide/ liter are essential for elasmobranch osmoregu- lation (Urist, 1962). The plasma calcium and magnesium in the skipjack are much less than those in the chondrichthyes. This can be attrib- uted to the apatite, which allows the teleost to maintain ionic concentrations independent of the external medium, and to the greater effi- ciency of the kidney and possibly the gills. 410 PACIFIC SCIENCE, VoL XXI, July 1967 TABLE 6 Comparison of Some Muscle Constituents of Some Marine Teleosts1 fish Na K Cl Ca Mg REFERENCE Muraena helena (eel) 25.0 165 23.7 18.7 14.9 Robertson (i960) Mycteroperca bonaci (grouper) 51.5 125.5 26.7 Becker et al. (1958) Scomberomorus maculatus (mackerel) 71.7 153.5 53.8 Becker et al. (1958) Clupea pilchardus (herring) Katsuwonus pelamis (skipjack) 53.8 170.6 65.6 152.1 51.3 Carteni and Aloj (1934) Red muscle 35.46 136.36 67.06 3.15 28.92 authors (1967) White muscle 20.77 191.61 67.71 8.88 44.43 1 Values in mEq/kg water. The plasma sodium content of the skipjack is greater than those of the Atlantic eel, the goosefish, and the kelp bass, but is lower than those of the Roman eel, the barracuda, and the grouper. The same order is found for the skip- jack when the chloride values are compared. In comparing the potassium values, the Atlantic eel and the barracuda have similar concentra- tions of plasma chlorides. The Roman eel, the goosefish, and the kelp bass have lower plasma chlorides than the skipjack. In K. pel amis, lesser concentrations of plasma sodium, potas- sium, and chloride are probably due to dif- ferences in the osmoregulatory mechanism and the type of integument. Excluding the mackerel and possibly the barracuda, the listed teleosts are not true pelagic species and may be sub- jected to some degree of salinity fluctuations. In Hawaii, the barracuda is frequently seen in shallow lagoons which are subjected to dilu- tions during heavy rains. Comparison of the plasma calcium of the listed teleosts shows that only the goosefish (Brull and Cuypers, 1955) had a higher plasma content. The value reported by Robertson (1954) appears to be in agreement with those reported by the other investigators. It is also apparent that the plasma magnesium of Lophius is greater than those reported by the other authors. It is known that temperature plays an important role in the solubility product constant of compounds and, thus, the rate of ionic exchange in apatite. A higher body tem- perature coupled with high serum alkaline phos- phatase activity and other factors would favor a decrease in blood calcium, phosphate, and magnesium. It has been reported that tunas and skipjacks have body temperatures 6°— 12°C higher than their environment (Kishinouye, 1923; Berg, 1940; Morrow and Mauro, 1950; and Van Oosten, 1957). It would be expected, then, that the blood calcium and magnesium content of the skipjack would be less than that found in the colder poikilothermic fishes, but greater than that found in mammals. This is apparent for the magnesium values but not for those of the plasma calcium. Also, there is a correlation between the activity of species and the amount of plasma magnesium. The more active forms generally have lesser concentra- tions of plasma magnesium. The greater blood calcium level of the skipjack, excluding the value of Brull and Cuypers, may be due to the intrinsic factors controlling osteogenesis and the amount of apatite coupled with the effi- ciency of the kidney, the ionic strength of the serum, and the amount of vitamins A and D stored in the liver. In brief, the differences in blood ionic con- centrations of various fishes is greatly influenced |i by the type and composition of the skeleton. Apatite not only stores Ca+2 and P04~2 but also Na+, Mg+2, and C03-2. The regulation of K+ and Cl- is influenced not by the skele- ton but by the gills and kidneys. Osmoregu- lation is delicately controlled by enzymes, hor- mones, and vitamins. The amounts and activities of these complexes are influenced by intrinsic and extrinsic factors which affect cell perme- ability and metabolism in such a way that each organism is unique in its electrolyte compo- sition. Comparative values of some muscle electro- lytes of marine teleosts are presented in Table 6. The ionic composition of the plasma defi- nitely influences that of the surrounding tissues. Constituents in Skipjack Muscle and Blood — Sather and Rogers 411 Fish with high plasma electrolyte values usually have tissues with relatively high electrolyte values. The grouper, mackerel, and herring have greater amounts of muscle sodium than does the skipjack. The sodium content of the muscles of the eel is intermediate between the skipjack’s red and white muscle content. This is to be expected because the plasma sodium of K. pelamis closely aproximates that of the eel. However, it is also apparent that the blood sodium of the mackerel and the skipjack are present in nearly equal concentrations. At first glance the difference in the muscle sodium con- tent of the two species is obscure, but it will be recalled that the body temperature of tuna is 6°-12°C higher than their environment and, therefore, the muscles of the skipjack would probably be more active metabolically than those of the mackerel. If this is truly the case, the sodium pump of the skipjack would be much more efficient, thus producing a lesser intracellular sodium content than that present in the mackerel and possibly in the other higher teleosts. Upon comparing the potassium content of the various muscles, it becomes apparent that the plasma content does not necessarily influ- ence the muscle content. This is obvious on examining the values for Muraena and Scom- beromorus. The plasma potassium content of the former animal is 1.95 mEq/liter in con- trast to a muscle content of 165 mEq/kg water. The potassium content of the plasma for Scom- ber omor us is 10.3 mEq/liter as compared with a muscle content of 153 mEq/kg water. Also, on examination of the blood and muscle con- centration of the teleosts, no obvious order is evident, e.g., the blood potassium order is: mackerel > grouper > skipjack > eel, and the muscle order is: skipjack white > eel> mack- erel > skipjack red > grouper. It appears, therefore, that the difference in muscle potas- sium may be under greater metabolic control than is sodium. Thus, the extracellular potas- sium may be entirely under the influence of the hormonal and genetic composition of the animal. The chloride content of a tissue, like the sodium content, is greater extracellularly than intracellularly. It would then be expected that the chloride content of the muscles would par- allel that of the plasma. However, the data in Tables 5 and 6 do not support this hypothesis. The importance of chloride in a tissue is to maintain electrochemical neutrality. Thus, the chloride content of a tissue is maintained pas- sively as a result of the Na+ and K+ distri- bution. As noted above, this ionic distribution is genetically influenced and thus the Cl- dis- tribution would subsequently be controlled but in a more subtle manner. Further examination of Table 6 reveals that the muscle and blood chlorides of the grouper are greater than those of the eel, and also that the chloride values for the skipjack are greater than those of the mackerel. However, the relationship between blood and muscle chlorides terminates at this, point, because the skipjack muscles have the greatest chloride content, but the plasma chloride content is intermediate between those of the eel and the mackerel. It is not possible to make similar compari- sons with the herring because the blood values of this fish could not be located. Such data would be informative because the herring is more closely related systematically and ecologically to the skipjack than to the eel and grouper. The data for muscle calcium and magnesium of marine teleosts are very meager. In Table 6 only one direct comparison can be made, that between M. helena and K. pelamis. The values for the herring cannot be considered because, as was noted by Robertson (I960), the muscle samples were contaminated with bone frag- ments. The calcium content of both muscle types of the skipjack is less than that of the eel, although the blood calcium levels of both species are approximately the same. The results of the comparison of the mag- nesium contents are opposite to those of the calcium comparison. The eel has about twice the amount of blood magnesium that the skip- jack does. The differences in the muscle con- tent are that the red muscle of K. pelamis has about twice the amount, and the white muscle has about three times the amount found in the eel. This may be due to a greater pre- ponderance of myosin and adenosine triphos- phate (ATP) in the muscles of the skipjack. It is known that magnesium serves as a co- factor for bridging ATP and creatine to the creatine kinase molecule during transphospho- rylation (White et al., 1964). It is quite pos- 412 PACIFIC SCIENCE, Vol. XXI, July 1967 sible that the muscle ATP content of fast swimming fish is greater than that in less active forms. Studies on the phosphorus com- pounds of fish muscle may produce a strong correlation between magnesium and ATP- creatine phosphate contents. SUMMARY 1. The major electrolyte constituents of the plasma, red muscle, and white muscle of the oceanic skipjack, Katsuwonus pelamis, were determined. The potassium content and the greater mitochondrial density of the red muscle suggest that this muscle is utilized for normal swimming activity rather than being an energy source for the white muscle. 2. The plasma electrolytes were compared with those of other marine fishes. In general, the sodium content of the skipjack plasma is less than that found in the cyclostomes, the skate and the shark, but is slighly greater than that found in the majority of other teleosts. The plasma potassium is less than that in the cyclostomes and elasmobranchs and greater than that in other teleosts. The plasma chloride content of the skipjack, as well as the calcium and magnesium, is less than that of the other investigated species. 3. Comparison of the differences in the elec- trolyte composition of the red and white mus- cles reveals that the white tissue contains a larger amount of water, potassium, and mag- nesium. However, the red muscle contains a greater amount of sodium. 4. Using C14-inulin, the extracellular space of the red and white muscles was determined to be approximately 0.24 1/kg muscle and 0.19 1/kg muscle, respectively. 5. The muscle electrolyte content of K. pelamis was contrasted with the muscle con- tents of other teleosts. The order of decreasing composition is as follows. For Na+: mack- erel > herring > grouper > skipjack red > eel > skipjack white; for K+ : skipjack white > herring > eel > mackerel > skipjack red > grouper; for Cl- : skipjack red > skipjack white > herring > mackerel > grouper > eel. Both muscle types of the skipjack contained less calcium and more magnesium than did the muscle of the eel. REFERENCES Becker, E. L., R. Bird, J. W. Kelly, S. S. Schilling, and N. Young. 1958. Physiol- ogy of marine teleosts. I. Ionic composition of the tissue. Physiol. Zool. 31:224. Berg, L. S. 1940. Classification of fishes both recent and fossil. Trav. Inst. Zool. Acad. Sci. USSR 5:517. Brull, L., and Y. Cuypers. 1955. Blood per- fusion of the kidney of Lophius piscatorius L. IV. Magnesium excretion. J. Mar. Biol. Assn. U. K. 34:637. Carteni, A., and G. Aloj. 1934. Composi- tion chimica de animali marini del golfo de Napoli. I. Teleostei. Arch. Internat. Physiol. 42:398. Davson, H. 1964. A Textbook of General Physiology. 3rd ed. Little, Brown and Co., Boston. 1166 pp. Ferro, P. V., and A. M. Ham. 1957^. A simple spectrophotometric method for the determination of calcium. Am. J. Clin. Path. 28:208. ■ 1951b. A simple spectrophoto- metric method for the determination of calcium. II. A semimicro method with re- duced precipitation time. Am. J. Clin. Path. 28:689. Geyer, R. P., and E. J. Bowie. 1961. The di- rect determination of tissue calcium by flame photometry. Anal. Biochem. 2:360. Hartman, F. A., L. A. Lewis, K. A. Brown- ell, F. F. Sheldon, and R. F. Walther. 1941. Some blood constituents of the nor- mal skate. Physiol. Zool. 14:476. - — — — - - C. A. Angerer, and F. A. Sheldon. 19 44. Effect of interrenalec- tomy on some blood constituents in the skate. Physiol. Zool. 17:228. Holliday, F. G. T., and J. H. S. Blaxter. 1961. The effects of salinity on the herring after metamorphosis. J. Mar. Biol. Assn. U. K. 41:37. Kishinouye, K. 1923. Contributions to the comparative study of the so-called scombroid fishes. J. Coll. Agr. Imp. Univ., Tokyo 8:293. Love, R. M., 1957. The biochemical composi- tion of fish, Chapt. 10, pp. 401-418. In: M. E. Brown, ed., The Physiology of Fishes, Vol. 1. Academic Press, N. Y. Constituents in Skipjack Muscle and Blood — Sather and Rogers 413 Morrow, J. E., Jr., and A. Mauro. 1950. Body temperatures of some marine fishes. Copeia 2:108. Potts, W. T. W., and G. Parry. 1964. Os- motic and Ionic Regulation in Animals. Macmillan Co., N.Y. 423 pp. Robertson, J. D. 1954. The chemical com- position of the blood of some aquatic chor- dates, including members of the Tunicata, Cyclostomata and Osteichthyes. J. Exptl. Biol. 31:424. I960. Studies on the chemical compo- sition of muscle tissue. I. The muscle of the hagfish, Myxine glutinosa, and the Roman eel, Muraena helena. J. Exptl. Biol. 37:879. Salome Pereira, R., and P. Sawaya. 1957. Ions in the bloods of elasmobranchs. Bol. Fac. Fil. Cien Univ., Sao Paulo, Zool. 21. Smith, H. W. 1931. The absorption and ex- cretion of water and salts by the elasmo- branch fishes. II. Marine elasmobranchs. Am. J. Physiol. 98:269. Urist, M. R. 1962. The bone-body fluid con- tinuum: calcium and phosphorus in the skeleton and blood of extinct and living vertebrates. Perspect. Biol. Med. 6:75. 1963. The regulation of calcium and other ions in the serums of hagfish and lampreys. Ann. N.Y. Acad. Sci. 109:294. Van Fossan, D. D., E. E. Baird, and G. S. Tekell. 1959. A simplified flame spectro- photometric method for estimation of mag- nesium in serum. Am. J. Clin. Path. 31:368. Van Oosten, J. 1957. The skin and scales, Chapt. 5, pp. 207-244. In: M. E. Brown, ed., The Physiology of Fishes, Vol. 1. Aca- demic Press, N.Y. Vernick, S. H. 1964. Histology of the red muscle in four teleosts. Copeia 4 (44th Ann. Meet. Am. Soc. Ichthyol. Herpetol.) :738. Vinogradov, A. P. 1953. Elementary compo- sition of Pisces, Chapt. 21, pp. 463-566. In: The Elementary Composition of Marine Or- ganisms. Yale Univ. Press, New Haven. White, A. B., P. Handler, and E. L. Smith. 1964. Principles of Biochemistry. McGraw- Hill, N. Y., p. 261. The Systematics of the Prickly Sculpin, Cottus asper Richardson, a Polytypic Species Part II. Studies on the Life History, with Especial Reference to Migration1 Richard J. Krejsa2 ABSTRACT : The occurrence of a downstream spring migration of weakly-prickled Cottus asper in coastal streams is confirmed and documented. Successful intertidal spawning and incubation is followed by a pelagic larval stage of about one month. Metamorphosis occurs and the pre juveniles settle to the bottom to feed in the estuarine portion of the river. An upstream migration of adults precedes that of the young-of-the-year in late summer. During the non-migratory phase, prickly sculpins are located in the low gradient, low velocity portions of coastal streams. Densely-prickled Cottus asper living in distant inland waters, where access to the sea is almost impossible, undertake only local migratory movements. Densely- prickled forms living in some inland lakes and streams relatively close to the sea, where access to the sea is open and relatively easy, do not migrate seaward but undertake only local movements to spawn in fresh water. The present study dis- cusses differences in migratory behavior between "coastal” and "inland” prickly sculpins. The existence of a seaward spawning migra- tion of prickly sculpins in coastal streams has been suggested, or implied, for at least 30 years (Taft, 1934; Pritchard, 1936; Sumner, 1953; Shapovalov and Taft, 1954; Hunter, 1959; and McAllister and Lindsey, 1959). Although some of these authors have observed the presence of reproductively mature Cottus asper in the inter- tidal areas of coastal streams, none has demon- strated that intertidal spawning actually occurs. Populations of C. asper occurring in lakes and streams far enough inland to preclude the possibility of an annual seaward spawning mi- gration are presumed to spawn in fresh water. The approximate or exact spawning sites of some of these populations have now been deter- mined from the presence of larvae (Nicola Lake, British Columbia), and egg clusters or gravid females (Pothole Lake, near Merritt, British 1 With data taken from a thesis submitted in par- tial fulfillment of the requirements for the degree of Doctor of Philosophy at the University of British Columbia. Manuscript received March 23, 1966. 2 Institute of Fisheries, University of British Colum- bia, Vancouver, Canada. Present address: Department of Anatomy, College of Physicians and Surgeons, Columbia University, New York, N.Y. Columbia). Other localities are close enough to the sea to imply the existence of a short seaward migration on the part of the C. asper popula- tions living therein, but access to the sea is pre- vented by natural or man-made barriers, e.g., at Buttle Lake and Horne Lake, on Vancouver Island, British Columbia. Spawning of prickly sculpins in these areas is necessarily restricted to fresh water. Still other localities, frequented by migratory salmonids, are close enough to the sea to permit a seaward migration on the part of C. asper living there, but it does not occur. For example, these spawning sites of the following prickly sculpin populations in the lower Fraser Valley in British Columbia are known from capture of gravid fish and/or egg masses: South Alouette River; Kenworthy Creek and Chilqua Slough (both are inlet streams to Hatzic Lake) ; Squa- kum Lake (Lake Erroch) ; and Cultus Lake. In addition, spawning fish have been captured in inlet streams of Skidegate Lake, in the Queen Charlotte Islands, along with migratory juvenile salmonids. The outlet of Skidegate Lake is only about 13 miles from the sea. A newly hatched larva of C. asper (?) has been taken in a 414 Systematics of Prickly Sculpin, II — Krejsa plankton net in the Second Narrows region of Owikeno Lake (about 30 miles from the sea), on the coast of central British Columbia. It is presumed that the parents spawned in fresh water. MATERIALS AND METHODS To document the supposed occurrence and to determine the success of intertidal spawning of prickly sculpins in the coastal streams, the Little Campbell River (Fig. 1, site 2) was chosen as a study stream. In 1960-1961, a series of 18 permanent collecting sites (Fig. 2) was sampled at biweekly intervals for a period of one year, and at monthly intervals for an additional six months. The lower reaches of the river, stations C-l to C-3, were also sampled several times in late winter and early spring of 1962 and 1963, to obtain live specimens for laboratory studies. Additional live specimens for use in laboratory studies were collected from the following local- ities (Fig. 1) : site 1, Nile Creek and Big Qual- icum River, Vancouver Island; site 3, South 415 Alouette River; site 4, Kenworthy and Edwards creeks (Hatzic Lake); site 5, Sweltzer Creek (outlet of Cultus Lake) ; and site 6, Squakum Lake (Lake Erroch). A 3 mm-mesh, woven-nylon seine, 3 m wide X 2 m deep, was mounted on collapsible tele- scoping aluminum poles and used for all field collections. Salinities were measured with den- sity hydrometers. SAMPLING LOCALITIES AND STUDY STREAM The primary study area was the Little Camp- bell River (Campbell Creek), which is ap- proximately 15 miles long and empties into Semiahmoo Bay between White Rock, British Columbia and Blaine, Washington (site 2, and inset of Fig. 1). The stream’s drainage area is approximately 28 square miles. Collection sites are shown in Figure 2. Sta- tion 0-1 is located on a sand-mud flat outside the main river channel. Station C-0 is located below the railroad trestle at the mouth of the river in the main channel, station C-l about 50 Fig. 1. Localities for spawning populations of Cottus asper used in life history studies. Site 2 (inset) is expanded in Figure 2. Other site localities are listed in text. 416 PACIFIC SCIENCE, Vol. XXI, July 1967 yards inside the mouth, and station C-lA about 75 yards inside the mouth. Station C-2 is lo- cated \ mile, and C-3 ^ mile, from the mouth. Figure 3 shows the collection sites in relation to stream gradients, and maximum upstream effects of tidal fluctuations in salinity and depth. Stations C-0 through C-9 are all subject to tidal fluctuations in depth, whereas stations C-0 through C-3 are tidally inundated with mixo- haline waters. Barnacles ( Balanus sp.) are found at stations C-0 through C-3, and permanent beds of the oyster Crassostrea gigas are located between sta- tions C-l and C-lA, and at C-2. Typical fish associates in areas C-0 through C-3 are Lepto- cottus armatus, Platichthys stellatus, and, throughout the summer, young-of-the-year Cot- tus aleuticus. Oligocottus maculosus and Clino- cottus actiticeps are commonly found upstream as far as station C-2. RESULTS OF FIELD STUDIES The prickly sculpin is distributed primarily in the lower 4 miles of the Little Campbell River. Especially in spring, 1961, an increased number of C. asper were present in the lower reaches of the river, around the spawning site (station C-2). Over the first 9-month sampling period, no C. asper were captured in stations upstream of C-2 2. With three exceptions, none was taken in the fast-flowing, high gradient area of the stream below C-20 and above C-ll (Fig. 3). This area is densely populated with the coast range sculpin, Cottus aleuticus. Figure 4 illus- trates the disjunct distribution of yearlings, subadult, and adult prickly sculpins. From late February to early March the prickly sculpin undertakes a migration downstream to the estuary. The only area in the lower 4 miles of stream in which suitable spawning substrate (large cobbles, flat rocks) occurs is a stretch about 100 yards long lying J mile upstream from the mouth (station C-2, Figs. 2 and 3). The males, which come into spawning condi- tion earlier in the season than the females (see below), select nesting sites under large cobbles or flat rocks in areas of the stream bed with current velocities equal to or less than 1 cubic ft/second (at low tide). Apparently it is im- Systematics of Prickly Sculpin, II — Krejsa 417 MILES FROM SEA Fig. 3. Little Campbell River collection sites in relation to stream gradients and tidal influence. See text for explanation. \ \ M Intertidal Zone; £22125221 Spawning Zone; al Spawning Season. Fig. 4. Monthly distribution of yearling, subadult, and adult Cottus asper within the Little Campbell River. Data taken from pooled biweekly samples representing results of 700 seine hauls. 418 PACIFIC SCIENCE, VoL XXI, July 1967 portant that the substrate surface be relatively rough in texture, since the adhesive eggs adhere only temporarily to a smooth surface such as glass or plastic. Old automobile exhaust pipes, or muffler tubes, are "preferred” nesting sites when available in the environment (as they are in the Little Campbell River). Females aggregate upstream (about station C-3) above the main spawning area and then move individually onto the spawning beds where they display to, and are courted by, males both outside and inside their nests. After a male selects a female to occupy his nest, further courtship and prespawning behavior occurs within the nest. The adhesive eggs are laid in a jelly-enclosed cluster on the ceiling of the spawn- ing chamber. Ovariotomy of preserved gravid females from throughout the distributional range of C. asper yielded counts of 336 mature oocytes in a 49. 5 -mm S.L. female, to 5,652 mature oocytes in a female of 119.5 mm S.L. The largest female examined was 192 mm S.L., but she was spent. A conservative estimate of the number of mature oocytes would be about 10,000 for this female. Numbers of viable eggs, in masses collected in the field, varied from 700 to 4,000 per cluster. However, one male may court and successfully mate with as many as 10 different females (personal observation). As many as 10 egg masses, in varying stages of development from newly-fertilized to near- hatching, have been found in the nest of a sin- gle brooding male. An estimated 25,000-30,000 eggs were present in this one nest. After spawning, the spent females leave, or are chased from, the nests and they again aggre- gate above the spawning areas and begin feed- ing. The males remain in the nests, fanning and protecting the eggs, and do not eat until hatch- ing of all egg clusters is completed. Much of the courtship and prespawning behavior, as well as most of the paternal brooding behavior through hatching, has been documented and will be reported elsewhere. Laboratory studies on the behavior of C. asper larvae, done in extension of salinity-tolerance experiments (also to be reported elsewhere), indicate that at 12° C the larvae 5-7 mm in total length begin swimming immediately upon hatching. They remain pelagic, as lightly- pigmented transparent larvae, for a period of 30-35 days before metamorphosing and settling to the bottom. Figure 5 Illustrates numbers and distribution of C. asper young-of-the-year, 12-25 mm S.L., taken in a total of 700 seine hauls. In late spring and throughout the summer, the newly meta- morphosed young-of-the-year are found in great numbers around and below the spawning site. The concentration is greatest around station C- 1A, where there is a bed of fine, pea-size gravel adjacent to a large oyster bed. In mid-summer, there is a definite upstream migration of the young-of-the-year. In both I960 and 1961, the increasing abundance of young-of-the-year at stations C-4 and C-5 was correlated with the decreasing abundance of specimens In the estu- arine areas of the river (Fig. 5). Spawning Period and Temperature Relationships Egg clusters were collected from several lo- calities in the lower Fraser Valley (cf. Fig. 1) and in the Little Campbell River. Gonads were examined in over 1,100 preserved museum spec- imens from all latitudes within the distribu- tional range of C. asper. These data indicate that egg deposition begins in the south of the distribution range (low latitudes) in February, and progresses northward until late July. Males usually attain full reproductive maturity about a month before, and remain in spawning condi- tion for almost a month after, the period of oviposition in females. Gravid females have been found over a 4- week period in Squakum Lake, and a 6-week period in the Little Camp- bell River. Ripe males have been taken over an 8-12 week period, respectively, in these same localities. The earliest date on which a ripe male, in nuptial dress and with flowing sperm, was col- lected is February 6, in San Francisco Bay. The earliest collection of gravid females was In Waddell Creek, California, on February 24. In the north end of the range, gravid females were taken as late as June 20 in Petersen Creek, near Juneau, Alaska, and on July 22 in streams entering Juskatla Inlet, Queen Charlotte Is- lands. Gravid females have also been collected from Middle River, near Takla Lake, on June 28, and from Meziadin Lake, B. C, on July 25. Field records and personal observations indi- Systematics of Prickly Sculpin, II — Krejsa 419 C/)23 Ilf £ 19 "if II - 9- O UJ - o X o X o X 0 o o X o X X X X X X 9 - o o o o o 0 o X o o o X X X o X o " 8 7 - o o o o o o o o o o o X o X o o X - 6 5 - o o o o o o o o o o o o o X o o o " - o o o o o o o o o Q o o X X X o o - 4 - o o o X • X o o X o X X X X o o o _ 3 - 2 - o • • • • • • # o o o o • X • • • m \\ \ i \ . , ■ * \. \ \ \ i\ \ \ i i J J A S 0 N D ! t 1 1 1 £ ! 1 1 ! JFMAMJJASO I960 1961 m co co m o cQ cn 0 O o CD LEGEND: x=No collection made; a O=€ollection, but no fish; •=l»6 fish; • =7-18; •=19-36; W= 37-60; W=6l-90. f\ \ \ \ \ \l Intertidal Zone; IMSniOnnSI Spawning Zone; Spawning Season. Fig. 5. Monthly distribution of young-of-the-year Cottus asper within the Little Campbell River, taken from pooled biweekly samples representing results of 700 seine hauls. Data cate that natural spawning temperatures range from 8° to 13° C. Egg-rearing experiments at various temperatures resulted in complete mor- tality at 18° C, less than 50% survival at 15° C, and greater than 85% survival at 12° C. Control of lower temperatures was beyond the limitations of available equipment. The annual mean range of ambient tempera- tures experienced by inland populations of C. asper is twice as great as that experienced by coastal populations at the same latitude (Krejsa, 1965:109). The monthly mean temperature range between northern and southern localities is from 8.8° to 19-5° C on the coast, and from 9.6° to 29.6° C inland. A priori , one might expect that inland populations would experience a greater range of developmental temperatures than do coastal populations. Apparently, how- ever, they do not. When monthly mean temper- atures, representing inland and coastal localities encompassing the distributional range of C. asper, are plotted against latitude, the mean temperature differences between inland and coastal localities during the spawning period are almost negligible (Fig. 6). In fact, the empiri- cally determined, average spawning temperature range of 8° to 13° C (shaded bar, Fig. 6) can be followed as a thermal "wave” progressing through inland and coastal localities from the south in February, to the north in June. Ap- proximate spawning times, determined from examinations of gonadal condition in more than 1,100 specimens from all latitudes, are in gen- eral agreement with this south-north progres- sion although the latitudinal range over which spawning occurs in March, April, and May is remarkably consistent (Fig. 6). Theoretically, inland forms have a shorter period of exposure to spawning temperatures of 8° to 13° C than do coastal forms (Fig. 6). This supposition has been borne out by field data from the two most frequently collected spawning sites, Squakum Lake and Little Camp- bell River. According to Figure 6, the inception of spawning in inland streams should lag behind that of coastal streams at similar latitudes. This is because upstream or inland areas remain colder for a longer period than do coastal areas. This is apparently true in the lower Fraser Val- ley. For example, the following localities are all within 15' of 49° N, and gravid females and/ 420 PACIFIC SCIENCE, VoL XXI, July 1967 Latitudinal Range of Probable Spawning in: Coastal l\ \ \ \ \\: Inland K// / \: and Both IX yVSAl Forms. State of Maturation: VII = Near Spawning, VIII- Spawning Imminent, Eggs and Sperm Flowing,' IX= Spent. Fig. 6. Monthly mean temperatures in °C arranged by latitude for coastal and inland localities encompass- ing the distributional range of Cottus asper. Solid or unfilled vertical bars, left side of each panel, indicate actual ranges of latitude over which specimens have been found in a given state of maturity. Temperature values after Krejsa, (1965:109). or eggs have been collected from them during the following dates: March 7 to May 10 in Little Campbell River (122° 46' W) ; March 25 to April 15 in South Alouette River (122° 35' W) ; and April 30 to May 27 in Squakum Lake (122° 00' W). DISCUSSION The present study confirms the existence of a downstream spawning migration of the prickly sculpin in the Little Campbell River. It not only documents the occurrence of intertidal spawn- ing in this coastal population of C. asper, but indicates that this spawning is successful. By extrapolation, the success of intertidal spawning in other coastal streams of the Pacific coast, where catadromous populations of C. asper have been reported, is indicated. The lack of a seaward migration in "inland” populations with or without immediate access to the sea is also documented. In these popula- tions, fresh-water spawning is the invariable norm. Krejsa (1965) has discussed the evidence for recognizing a genetic distinction between "coastal” and "inland” populations of prickly sculpins based on morphology (prickling pat- terns) and geographic distribution. Bohn and Hoar (1965) have presented physiological evi- dence based on a comparison of iodine metab- olism in "coastal” (Little Campbell River) and "inland” (Lagace Creek, Hatzic Lake) forms of C. asper. They concluded that the two popu- lations studied have diverged genetically in their physiological capacities to deal with water of different electrolyte content. If the above evidence is considered in light of the present study, the following picture emerges. Weakly -prickled C. asper living in coastal streams having open access to the sea undertake a downstream migration to the estu- arine regions where eggs are spawned, hatched, and the young are reared successfully. Densely- prickled C. asper living in distant inland streams, where access to the sea is almost impossible, undertake only local migratory movements. But densely-prickled forms living in some inland Systematics of Prickly Sculpin, II — Krejsa lakes and streams relatively close to the sea, where access to the sea is open and relatively easy to achieve, do not migrate seaward. They also undertake only local movements and spawn within the fresh-water system in which they are found. Such primary differences in behavior, correlated with distinct differences in prickling patterns, geographical distribution, and iodine metabolism, further corroborate the contention that "coastal” and "inland” forms of C. asper are genetically distinct. Figure 5 shows a lag of about two months in the appearance of young-of-the-year C. asper, 12-25 mm S.L., after the first recorded spawn- ing in March. Water temperatures in March are normally from 8° to 10° C. Eggs spawned early in March probably have an incubation period several days longer than the 15-16 day period found to be typical at 12° C in the laboratory. The lag shown in Figure 5 probably is due to an incubation period of 3 weeks followed by a pelagic larval stage of 4-5 weeks. Plankton tows taken during late April in the shallow back-eddies of the stream have captured a few larvae 9-10 mm S.L. (not recorded in Kg- 5). An upstream migration of adults precedes that of the young-of-the-year in late summer. This is probably related to the food habits of the two groups and also to the fact that a later return of the young-of-the-year coincides with lower water levels in the stream, when re- duced water velocity facilitates access upstream. The usual spawning temperature range for C. asper is from 8° to 13° C. It is assumed that in most populations spawning is initiated within this 5° range of temperature, which pro- ceeds in somewhat of a thermal "wave” from south to north in both inland and coastal locali- ties (Fig. 6). This is not to say that they all spawn at the same temperature within the range. Furthermore, because the monthly rate of in- crease of ambient temperature is greater in the north, the duration of exposure of prickly scul- pin eggs to any given temperature within the 5° C temperature range may be shorter. Eggs subjected to these conditions presumably would have a faster development than eggs which de- veloped under relatively more thermostable con- ditions, such as occur in the south of the dis- tributional range. Low meristic counts are gen- 421 erally (although not invariably) associated with faster rates of development. An experimental analysis of temperature-determined morpholog- ical differences is needed, especially of meristic differences, between "coastal” and "inland” populations of C. asper at the same latitude and between fishes from the north and south ends of their respective ranges. The implications and the desirability of such studies in determining the validity of the proposed genetic distinction between "inland” and "coastal” populations are obvious. The existing meristic evidence (Krejsa, 1965) is equivocal. SUMMARY 1. The stream studied, the Little Campbell River, is a small coastal stream, the lower mile of which is subject to tidal inundation of mixohaline waters. 2. In late winter and early spring, the adults and juvenile prickly sculpins migrate down- stream to the estuarine region of the Little Campbell River, the only region in the lower 4 miles in which suitable spawning substrate is available. 3. Males set up nesting sites under large cobbles and rocks, and courtship occurs both outside and within the nest. 4. Spawning occurs from March throughout early May. 5. Newly-hatched larvae begin swimming immediately and remain pelagic for a period of 30—35 days before metamorphosing and settling to the bottom. 6. In May, metamorphosed young-of-the- year (approximately 12 mm S.L.) begin appear- ing only in those collections taken in the estu- arine portion of the river. They occur in great abundance until September, when the numbers decrease in the estaury and increase in upstream, nones tuarine waters. 7. During the nonmigratory phase of its life history, the prickly sculpin population in the Little Campbell River is distributed primarily in the low gradient, low velocity, portions of the stream. 8. Within any given population of prickly sculpins, the males are reproductively active longer, in a given season, than the females. The period of reproductive activity of both sexes is 422 PACIFIC SCIENCE, Vol. XXI, July 1967 more extensive in "coastal” populations than in "inland” populations. 9. The empirically determined, average nat- ural spawning temperature of C. asper is from 8° to 13° C for both "coastal” and "inland” populations. Within this range of temperature, egg deposition begins in February in the south of the distributional range and progresses north- ward until July. 10. The existence of catadromous "coastal” populations and nonmigratory "inland” popu- lations is indicative of genetic distinction be- tween them. This contention is further strength- ened by the existence of parallel differences in morphology, physiology, and geographic dis- tribution. ACKNOWLEDGMENTS The Research Division of the British Colum- bia Fish and Game Branch provided financial support during the summer of I960 and also the much appreciated and entertaining assistance of Dr. G. F. Hartman and Mr. C. A. Gill in the field studies. The Vancouver Public Aquar- ium provided research space for life history studies and I wish to thank the Curator and staff for their help and kindness. Almost every graduate student enrolled in the Institute of Fisheries volunteered assistance in field collec- tions at one time or another during the period 1960-1 963. I offer collective thanks to all. REFERENCES Bohn, A., and W. S. Hoar. 1965. The effect of salinity on the iodine metabolism of coastal and inland prickly sculpins, Cottus asper Richardson. Can. J. Zool. 43(1965) :977- 985. Hunter, J. G. 1959. Survival and production of pink and chum salmon in a coastal stream. J. Fish. Res. Bd. Can. 16(6) :835-886. Krejsa, R. J. 1965. The Systematics of the Prickly Sculpin, Cottus asper : An Investiga- tion of Genetic and Non-genetic Variation within a Polytypic Species. Unpublished Ph.D. thesis, Univ. British Columbia, June, 1965. 109 pp. McAllister, D. E., and C. C. Lindsey. 1959. Systematics of the freshwater sculpins ( Coi- tus) of British Columbia. Nat. Mus. Can. Contr. Zool., Bull. 172:66-89. Pritchard, A. L. 1936. Stomach content anal- yses of fishes preying upon the young of Pacific salmon during the fry migration at McClinton Creek, Masset Inlet, British Co- lumbia. Can. Field-Naturalist 50(6) : 104— 105. Shapovalov, L., and A. C. Taft. 1954. The life histories of the steelhead rainbow trout (Salma gairdneri gairdneri) and silver sal- mon (Oncorhynchus kisutch ) . Calif. Dept. Fish and Game, Fish Bull. 98:1-375. Sumner, F. H. 1953. Migrations of salmonids in Sand Creek, Oregon. Trans. Am. Fish. Soc. 82(1952) :139-150. Taft, A. C. 1934. California steelhead experi- ments. Trans. Am. Fish. Soc. 64(1933) :248- 251. ADDENDUM Since this manuscript was submitted for pub- lication, Mr. Gerald D. Taylor has presented an excellent statistical analysis of experimental field and laboratory studies on the interrelation- ships of Cottus asper and C. aleuticus in the Little Campbell River.1 His thesis forms a valuable contribution to our present knowledge of sculpin ecology. Most of his conclusions ex- pand and elaborate upon many of the observa- tions reported above. However, his study opens up the possibility that some of the prickly sculpins in the Little Campbell River do not spawn in the lower estuarine portion of the stream but upstream "in close proximity to spawning C. aleuticus In other nearby coastal streams where stream profiles are different, I have collected reproduc- tively active C. asper and C. aleuticus in close proximity, but never have I done so in the Little Campbell River. I am skeptical of Taylor’s statement primarily because his collections were carried out only during 5 months (of a 7 -month period: August to November, 1965, and Febru- ary, 1966), all of which are outside the normal spawning season of most C. asper as documented above. Taylor’s study will be given more con- sideration in a future report on the behavior of C. asper and C. aleuticus. 1 Taylor, G. D. 1966. Distribution and Habitat Responses of the Coastrange Sculpin ( Cottus aleu- ticus) and Prickly Sculpin ( Cottus asper ) in the Little Campbell River, British Columbia. Unpublished M.S. thesis, Department of Zoology, University of British Columbia (December, 1966). Herpetofauna of the Hawaiian Islands1 Don Hunsaker II2 and Paul Breese3 This study was undertaken to determine the changes that have occurred in the herpetofauna of the Hawaiian Islands since the work in the early forties by Oliver and Shaw (1953). The work is a result of a field survey of most of the islands conducted by the authors during the summer of 1962. Since the purpose of the sur- vey was to observe gecko vocalization, more data are available on this group than on the others. The islands afford a unique opportunity to observe a dynamic fauna since there are con- stant introductions from foreign sources. Peri- odical surveys have been made which furnish an investigator a well annotated history. Changes have been observed during each of the major surveys, made by Stejneger (1899), Sny- der (1917), and Oliver and Shaw (1953). In an accurate analysis of the faunal charac- teristics of the islands during 1943 when their study was conducted, Oliver and Shaw (1953), listed 8 species of amphibians and 15 species of reptiles. Since that time the house gecko, Hemi- dactylus jrenatus , has been introduced. Two iguanid lizards, the Cuban anole, Anolis por- catus, and the horned lizard, Phrynosoma cor- nutum, have become established and are now considered to be permanent residents of the islands. Anolis was established by 1951 (Shaw and Breese, 1951) and Phrynosoma since that time. Hawaii’s laws prohibit introduction of any snakes, but occasional specimens have been col- lected. Gopher snakes, Pituophis catenifer, and garter snakes, Thamnophis elegans, have been collected on Oahu, but are not considered to be established. These probably represent pets that have escaped and at the present time do not constitute a significant part of the fauna. 1 This research was supported in part by the Na- tional Institutes of Health, Grant No. M-4996. Man- uscript received April 18, 1966. 2 San Diego State College, San Diego, California. 3 Formerly Director of Kapiolani Park Zoo, Hono- lulu, Hawaii. REPTILIA SERPENTES (SNAKES) Blind Snake (Typhlops braminus) About 1930 this species was accidentally in- troduced on the island of Oahu with a shipment of palm trees from the Philippines which were planted around the new Kamehameha Schools in Kapalama, Honolulu. It became established quickly in the vicinity of the schools, and dur- ing the following decade spread into the resi- dential areas of Honolulu. By 1947, it had been collected in an area several miles from the original locality (Fisher, 1948), and it now appears to occupy the lowland area over the entire island. It is unknown on the higher mountains, but this may be due to inadequate collecting efforts. The large amounts of top soil that are transported from one part of the island to another probably are responsible for the spread of this fossorial animal. Recently, Typh- lops braminus has been collected in Kahului, the major port city of the island of Maui. While there are some restrictions on the transportation of soil around vegetation or domestic plants between the islands, certain people manage to evade them. Such shifting of top soil, potted plants, etc., as well as large-scale freight ship- ping by sea, probably is responsible for the in- troduction of the blind snake on Maui. Its establishment on the other islands is to be ex- pected. SAURIA (LIZARDS) GEKKONIDAE Mourning Gecko (Lepidodactylus lugubris) The first Hawaiians probably introduced this species during early population invasions ap- proximately 1,000 years ago. The eggs are highly adhesive and can be seen clinging to mats and other household articles such as early immigrants probably brought with them. They have been recorded in the Panama Canal Zone 423 424 PACIFIC SCIENCE, Vol. XXI, July 1967 by Smith and Grant (1961), undoubtedly in- troduced by this method of egg transport. Today the species is found throughout the islands, from remote forests to the downtown areas of the largest cities. It is the most frequently ob- served lizard and is well represented in collec- tions because of its gregarious habits, little fear of humans, and population concentration in areas inhabited by people. This common gecko apparently has adapted itself with great success to living in close association with humans. It is a common observation that these geckos occur in greater numbers in well populated areas than in more remote sections. Of a series of 21 indi- viduals collected in a transect from an uninhab- ited area into a city, 10 were collected in a pop- ulous section, 5 from the fringe area, and 6 from the uninhabited area. In making this tran- sect an attempt was made to maintain a con- stant unit of effort in each of the three habitats during the collections. The great number of numerous species of nocturnal insects attracted by the electric lights of the city probably is re- sponsible, in large part, for the large gecko population in the city. This does not explain, however, the abundance of geckos in or near man-made structures that were remote from any lighted areas. In nonurban areas, they are defi- nitely associated with open forests rather than with densely forested areas. Observations of several individuals indicate that L. lugubris is rather sedentary and that the home range usually does not exceed an area of 6 or 8 ft in diameter. However, these lizards periodically migrate from one area to another. These movements are not a coordinated group effort, but appear to be a simultaneous evacua- tion of the normal home ranges of many indi- viduals. During these periods, individual geckos have been seen moving across walls, down tree trunks, over sidewalks, etc. Such periods of ex- cessive activity have not been correlated with season, rainfall, temperature, or other factors. L. lugubris is active from shortly before dark until sunrise. Although highly gregarious, they show some aggression toward each other. A chirping noise is utilized in social behavior and a squeaking occurs during painful or aggressive encounters. Tail-waving has been observed in social interactions. The females are slightly larger than the males. The mean snout-vent length of 26 females was 41.69 mm, ranging from 33.0 to 46.7 mm. The mean snout-vent length of 7 males was 35.96 mm, ranging from 32.2 to 42.4 mm. Tree Gecko (Hemiphyllodactylus typus typus) No significant changes have appeared in the density and distribution of this species since 1943; it remains rather rare. Of 161 geckos collected in Kailua on Oahu, only 5 were of this species; 4 of them were collected on the sides of buildings in the city and 1 from under the bark on a tree. They are extremely agile and wary lizards and it is much easier to collect them from the sides of buildings than from tree trunks. Hence we do not believe that our larger collection from the city buildings neces- sarily implies a larger population there. H. typus typus is not a gregarious species; only 1 individual was collected from a well- lighted building which supported over 80 other geckos. On a darkened building about 30 ft away, 2 other individuals were collected about 10 ft apart. The only lizard found in close association with the tree gecko was the house gecko. House Gecko (Hemidactylus frenatus) This species is the latest addition to the her- petofauna of the Hawaiian Islands (Hunsaker, 1966). It was first observed in June 1951 in the city of Kailua, about 20 miles north of Honolulu. It is well established at the present time and appears to be rapidly replacing both Hemidactylus garnoti and Lepidodactylus lugu- bris in the cities on Oahu. H. frenatus can be identified easily by the series of enlarged scales which encircle the tail, exhibiting concentric circles of short spines (Fig. 1). These circles are separated by normal scales. This species has a cylindrical tail and lacks the lateral folds and | loose femoral skin of H. garnoti. H. frenatus is very similar to the fox gecko in size and color, but it is much more aggressive. Not uncom- monly it attempted to bite the collector. Mixed colonies of H. garnoti and H. frenatus are rare. Apparently the new immigrant is much more successful a competitor than are the other species of geckos. This factor, and its greater aggressiveness, apparently are responsible for its rapid replacement of the fox gecko in urban Herpetofauna of Hawaii — Hunsaker and Breese 425 Fig. 1. Hemidactylus frenatus. areas. H. frenatus has not been collected in areas removed from human habitation. In a transect which extended from an uninhabited area into the city of Kailua, 5 H. garnoti were collected in the uninhabited area and only 1 in the city proper. Conversely, 40 H. frenatus were collected in the city and none in the peripheral areas. Very few mourning geckos were found to be associated with H. frenatus. In Kuala Lumpur, Malaya, H. frenatus is quite common in the inhabited areas and L. lugubris is much more difficult to collect in the fringe areas. In the past, the well-lighted residential areas of Honolulu have been occupied by L. lugubris and the invasion of this habitat corresponds with findings of Church and Lim (1961), who stated that in Bandung, Java, H. frenatus preferred residential areas which were well-lighted and damp. In an area that has been under observa- tion for the past few years, the disappearance of the fox gecko and the mourning gecko popula- tion coincided with the appearance of this spe- cies. These large geckos appear to have a home range of an area about 12-15 ft in diameter. They are highly vocal and a distinctive series of five or six call notes can be heard up to 100 ft away. Aggressive or painful situations may pro- voke a prolonged squeak. This species occurred in close association with the stump-toed gecko (Per op us mutilatus ). H. frenatus were observed in Kailua, in many parts of Honolulu, around the International Airport, and in the wharf area. Four individuals of this species were col- lected only in the harbor areas of Kahului on Maui, while H. garnoti and L. lugubris were collected in other sections of the city. This dis- tribution indicates that H. frenatus has arrived only recently on Maui. None of these lizards has been collected on any other island. This is a large species; the largest individual collected was a male 58 mm in snout-vent length. The current distribution of H. frenatus indi- 426 PACIFIC SCIENCE, Vol. XXI, July 1967 cates that it is quite successful in establishing itself in new areas. The species is very wide- spread in the Orient and in the Malay States. Grant (1957) recorded it from Acapulco, Mex- ico; Tanner on Saipan in 1948; in 1950 it was recorded from Morotai; Church and Lim (1961) recorded it in Java; Larry Richards col- lected it from Guam in 1947. In all probability representatives of the Mexico population were transported by early traders, and the recent range expansion is due to equipment and ma- terial shipped during World War II. Fox Gecko (Hemidactylus garnoti) The future of this species in the islands will be interesting to follow. At the present time it is uncertain whether H. frenatus is replacing this species in unpopulated areas. If the fox gecko is better adapted to living in forested areas, it will probably continue its existence in this habitat. If the house gecko is as efficient in displacing the fox gecko in remote areas as it is in the cities, the future of this long-term resident is questionable. The fox geckos have been established for many years and probably are one of the earliest inhabitants of the islands. In Malaya, there are areas where both H. gar- noti and H. frenatus live in close association, and so it is possible that the two species will continue to be sympatric in Hawaii’s fauna. Of 12 eggs from the island of Hawaii which were laid during the last week of June, the largest was 17.0 X 8.9 mm, the smallest 9.4 X 8.7 mm. These measurements compare well with those of eggs of the same species re- corded by Cagle (1946) for eggs from a popu- lation on Tinian (which had a mean of 12 X 7 mm). Measurements made 2 weeks later did not indicate a significant change in size. Of the 12 eggs 7 hatched, and the mean snout- vent length of the newborn lizards was 24.08 mm, the mean total length, 48.0 mm. The range of snout-vent length was 26.0-22.5 mm; the range of total length was 51.0-45.5 mm. These measurements are well within the range quoted by Snyder (1917) for Hawaiian popu- lations: 39.5-56.0 mm total length. The maximum incubation period for eggs laid in Hawaii and hatched at room tempera- ture (74°F) was 64 days. Cagle (1946) hatched a series of eggs of this species in a 45 -day incubation period. Stump -Toed Gecko (Peropus mutilatus) There appears to be no significant change in the distribution of this gecko. It occurs in rather dense populations in some areas and is scarce in others. It is found away from the city of Honolulu in the back country and is quite common at the International Airport on Oahu. Females are slightly larger than the males. In a series of 17 females measured, the mean snout-vent length was 45.60 mm, with a range of 39.5-55.5 mm; for 34 males, the mean snout-vent length was 43.65 mm, with a range of 29.0-57.5 mm. IGUANIDAE Gray Cuban Anole (Anolis porcatus) This species is well established on Oahu at the present time. Both large adults and im- mature forms are commonly seen in Honolulu. The original site of collection was in the Kai- muki section of Honolulu. It has spread to other nearby sections of the city and to Manoa Valley, and is presently established on the north side of the island, at Kailua. The first intro- ductions were probably imported pet lizards which escaped. We can consider this species to be a permanent member of the fauna of Oahu. It has not been collected from the other islands. Texas Horned Lizard (Phynosoma cornutum) An increasing number of reports and of specimens collected indicate that this species is probably established as a permanent resident of the Hawaiian fauna on the island of Oahu. These animals undoubtedly originated from escaped pets. They have been found from the slopes of Diamond Head throughout Honolulu to the xeric areas above Pearl Harbor. No con- crete evidence has been obtained to indicate that a reproducing colony has been established, with eggs and hatchling lizards. Probably this lizard is reproducing, since immature speci- mens have been obtained from the islands, and the 10 or 15 reported is an unusually high Herpetofauna of Hawaii — Hunsaker and Breese 427 number to be accounted for by the escape of pet lizards. SCINCIDAE Snake-Eyed Skink (Ablephams boutoni poecilopleurus) No important changes have occurred in the population of this species. It is still rather com- mon in some of the more arid sections, but it occurs in definitely localized populations. In some areas It is absent, although the environ- ment is similar to that of other areas where the skink is common. Moth Skink (Lygosoma noctua noctua) At the present time, the moth skink is found only in a small area on the northern coast of Oahu, near Kahuku Point. There seems to be little doubt that the rapid expansion of Lygo- soma metallicum is responsible for the decrease in the once large populations of the moth skink. The area in which it now occurs is similar to large areas of Oahu which are now occupied by the metallic skink and at one time were occupied by L. noctua. The increase in amount of land under cultivation has not been great enough to account for the reduction that has been observed. No data are available for popu- lations once observed on Hawaii, Kauai, and Maui. Metallic Skink (Lygosoma metallicum) Few animals have been so successfully intro- duced as was the metallic skink on the island of Oahu. Its rapid multiplication on this island has produced remarkable numbers in the lower areas. It is aggressive with individuals of the moth skink, and it is certainly unafraid of humans. The success of the species probably is due in part to its apparent lack of fear of humans. It is easy to approach these lizards, and they can be found close to human habi- tations. The species is still known only from Oahu. AMPHIBIA Gold and Black Poison Frog (Dendrobates auratus) This frog has been limited in its distribu- tion only to the sites where it was released. It was originally introduced in upper Manoa Valley in 1932. This site now has a well-estab- lished population of frogs which extends to lower parts of the valley during the rainy season. Additional plantings with subsequent establishment have been made in Waiahole Valley, and the population has been observed to fluctuate in size at this locality, again accord- ing to the amount of water available. Bull Frog (Rana catesbeiana) This frog is extremely prolific and is well established on all major islands. No major changes in the populations have appeared since the 1940s. Green Frog (Rana clamitans) Since its introduction on Oahu in 1935, no great expansion of the population has been evident. It is not common anywhere on Oahu and has not been reported on other islands. Wrinkled Frog (Rana rugosa) This species is well established and is quite common in some areas. The population size varies with the amount of rainfall available. During 1962 the populations of most amphib- ians were reduced to half the numbers found in 1961, when surface water was much more plentiful. Marine Toad (Bufo marinus) This species is found on all the major islands and is the commonest species of amphibian. SUMMARY A survey of the herpetofauna of the Ha- waiian Islands was conducted during 1962 to determine any changes that might have oc- curred in the previous 20 years. New faunal species which have become established are Anolis pore at us, introduced in the late 1940s; Phrynosoma coronutum, introduced about 1955; and Hemidactylus frenatus, first observed in July 1961. Other species have extended or contracted their ranges, but no other significant changes were observed. It was noted that populational variations in amphibians could be attributed to annual changes in available surface moisture. 4 28 PACIFIC SCIENCE, Vol. XXI, July 1967 The house gecko, Hemidactylus frenatus, is rapidly increasing in numbers and is appar- ently being introduced into the other islands from Oahu. The most important means of intro- duction appears to be by the movement of boat cargoes from one harbor to another. REFERENCES Cagle, Fred. 1946. A lizard population on Tinian. Copeia 19 46:4-9. Church, G., and L. C. Lim. 1961. The distri- bution of three species of house gecko in Bandung (Java). Herpetologica 17:1 99— 201. Fisher, H. 1948. Locality records of Pacific island reptiles and amphibians. Copeia 1948: 69. Grant, C. 1957. The gecko Hemidactylus frenatus in Acapulco, Mexico. Herpetologica 13:153. Hunsaker, D. 1966. Population expansion of the house gecko, Hemidactylus frenatus. Philippine J. Sci. 95. Oliver, J., and C. Shaw. 1953. The amphib- ians and reptiles of the Hawaiian Islands. Zoologica 38:65-95. Shaw, C., and P. Breeze. 1951. An addition to the herpetofauna of Hawaii. Herpeto- logica 7:68. Smith, H. M., and C. Grant. 1961. The mourning gecko in the Americas. Herpeto- logica 17:68. Snyder, J. O. 1917. Notes on Hawaiian liz- ards. Proc. U. S. Natl. Mus. 54:19-25. Stejneger, L. 1899. The land reptiles of Hawaiian Islands. Proc. U. S. Natl. Mus. 21:783-813. NOTES Note on the Distribution of Euphausia eximia and E. gibboides in the Equatorial Pacific Claude Roger1 The material here considered was collected during the Alize cruise of the R.V. "Corio- lis” from the Centre O.R.S.T.O.M. Noumea. The Alize collections extended from 92°20'W to 162°45'E along the equator. The samples were taken with a 5-ft Isaacs-Kidd midwater trawl, towed obliquely from a depth of 300 m to the surface. GENERAL RESULTS The distributions of the whole euphausiid fauna will be discussed in detail in a further publication. At present they appear to be not far different from those described by Brinton (1962) ; however, two features become evident: 1. There is an evolution of the specific com- position of the euphausiid fauna from east to west. 2. Two species among the most important ones, Euphausia eximia Hansen and E. gibboides Ortmann, have been caught not only in the east- ern equatorial Pacific as previously recorded (Brinton, 1962), but also in the Central Pacific, as far westward as 164°15'W and 148°07'W, respectively. The present note deals with the oc- currence of these two species in Central Pacific waters. distribution of Euphausia eximia Table 1 lists the stations at which E. eximia were taken. According to Brinton (1962), the farthest westward record for this species is 118°W in the South Equatorial Current (2°N-2°S) and 145°W at 10°N. During the Alize expedition, 1 Section Oceanographie, Centre O.R.S.T.O.M. de Noumea, New Caledonia. Manuscript received May 18, 1966. TABLE 1 Quantitative Distribution of E. eximia and E. gibboides NUMBER PER STANDARD HAUL* STATIONS E. eximia E. gibboides 0.50S, 92.20W 752 316 0.49S, 95.28W 416 58 0.53S, 98.18W 6,224 272 1.00S, 101. 14W 780 36 0.16S, 103.48W 2,856 256 0.05S, 106.45W 2,136 96 0.40S, 109. 10W 1,233 33 0.20S, 115.40W 896 7 0.03N, 118. 27W 585 24 0.00 120.45W 933 45 0.40S, 123. 35W 1,330 0 0.40S, 125. 53W 558 16 0.33S, 128.26W 183 5 0.19S ,131.42W 40 5 0.33S, 134.46W 277 4 0.17S, 137.45W 17 3 0.01N, 145.06W 0 0 0.14S, 148.07W 0 1 0.27S, 151. 15W 0 0 0.28S, 154. 38W 0 0 0.38S, 158. 10W 0 0 0.22S, 161.06W 1 0 0.20S, 164. 15W 2 0 0.23S, 167.30W 0 0 0.28S, 170. 30W 0 0 0.23S, 174. 10W 0 0 0.20S, 177. 30W 0 0 0.23S, 179.00E 0 0 0.27S, 176.05E 0 0 0.12S, 172. 30E 0 0 0.18S, 169.00E 0 0 0.30S, 166.00E 0 0 0.38S, 162.45E 0 0 * Length of the column of water filtered: 5000 m. three specimens were caught at 164°15'W and l6l°06'W, about 2,700 miles farther west. E. eximia seems very common at 135°00'W, and very abundant east of 126°W. From 92° 20'W (beginning of the cruise) to 137°45'W .429 430 PACIFIC SCIENCE, VoL XXI, July 1967 this species accounts for 50-90% of the whole euphausiid material. On the other hand, it must be pointed out that, in a number of individuals, the inner pro- tuberance of the anterior margin of the second segment of the first antennal peduncle is trifur- cate (Fig. 1) and not simple or bifurcate as usually described (Hansen, 1912; Boden, John- son, and Brinton, 1955). In some specimens, this protuberance presents four spines (Fig. 2). distribution OF Euphausia gibboides This species was present more in the west than was previously known (see Table 1). The farthest westward that a specimen of E. gibboides was collected during the Alize cruise was 148°07/W. This record extends the west- ward limit of distribution, recorded previously as 132°W (Brinton, 1962). The species is present between 148°07/W and 126°W, rather common between 126°W and 109°10/W, and common between 109° 10'W and 92°20'W (beginning of the cruise). REFERENCES Boden, B. P., M. W. Johnson, and E. Brin- ton. 1955. The Euphausiacea (Crustacea) of the North Pacific. Bull. Scripps Inst. Oceanog., Univ. Calif., 6(8) :287-400. Brinton, E. 1962. The distribution of Pacific euphausiids. Bull. Scripps Inst. Oceanog., Univ. Calif. 8(2) :51— 270. Hansen, H. J. 1912. The Schizopoda. Repts. Sci. Res. Exped. Tropical Pacific. . . U. S. Fish. Comm. Steamer "Albatross.” Mem. Mus. Comp. Zook, Harvard College 35:177- 296. Fig. 1. E. eximia. Protuberances of the distal end of the second segment of the first antennal peduncle. Foreground', outer protuberance (simple). Back- ground’. inner protuberance (trifurcate) ; on the right , beginning of the dorsal keel of the third seg- ment. Fig. 2. E. eximia . Inner protuberance of the distal end of the second segment of the first antennal pe- duncle, showing four spine-shaped denticules. Monobrachium parasitum, a One-Tentacled Hydroid, Collected at Vancouver Island Richard D. Campbell1 The colonial hydroid Monobrachium para- situm (Mereschowsky) is of interest to system- atic and developmental biologists because of its peculiar form and its commensal habitat. Each feeding polyp has only a single tentacle, and the colony grows and develops a characteristic polymorphic pattern on living bivalves (Wag- ner, 1890; Hand, 1957). This note reports the rapid collection of this unusual hydroid with- out heavy equipment. On August 16, 1964, 13 colonies of M. para- situm were collected at Fraser’s (1918) original locality in Nanoose Bay, Vancouver Island, British Columbia. Nanoose Bay is about 1 km wide and 3 km long, and opens eastward. Water turnover is extensive during tidal changes. Tidal currents are located predominantly along the northern shore. Accordingly, the bottom of the northern third of the bay is composed of sand and light gravel, with remains of broken shells. The southern two-thirds presents a graded bot- tom from fine sand centrally to mud along the shallow southern shore. The bivalve Axinopsis was collected with a 1-mm mesh conical net 2 inches in diameter, towed for 10 minutes be- hind a small boat. Monobrachium colonies were found in three out of seven tows made in the central third of the bay (124° 9.2' W, 49° 1 Friday Harbor Laboratories, San Juan Island, Washington. Present address: Department of Orga- nismic Biology, University of California, Irvine, Cali- fornia 92664. Manuscript received July 23, 1966. 15.8' N, 16-17 fathoms), but not in either the deeper northern or the shallower southern sides. Axinopsis with Monobrachium which were brought into the laboratory showed rapid and extensive burrowing through fine sand. Gen- erally the hydroids were completely under the surface of the sand. No colonies showed any evidence of reproductive polyps or of medusa buds. Hand (1957) found Monobrachium in a number of dredge loads from the California and Baja California coast. Besides Hand, only Fraser (1918) had reported Monobrachium from the West Coast of North America. Re- peated dredgings in the San Juan Archipelago, Washington, failed to reveal any specimens, even in habitats similar to those prevailing on the east coast of Vancouver Island. REFERENCES Fraser, C. McLean. 1918. Monobrachium parasitum and other west coast hydroids. Trans. Roy. Soc. Can. 12:131-138. Hand, Cadet. 1957. The systematics, affinities, and hosts of the one-tentacled commensal hydroid Monobrachium, with new distribu- tional records. J. Washington Acad. Sci. 47(3) : 84-88. Wagner, Jules. 1890. Recherches sur l’organi- sation de Monobrachium parasiticum. Arch. Biol. 10:273-309. 431 Manuscript Form. Manuscripts should be typed on one side of standard-size, white bond paper and double-spaced throughout. Pages should be consecu- tively numbered in upper right-hand corner. Sheets should not be fastened together in any way, and should be mailed flat. Inserts should be either typed on separate sheets or pasted on proper page, and point of insertion should be clearly indicated. Original copy and one carbon copy of manuscript should be submitted. The author should retain a car- bon copy. Although due care will be taken, the editors cannot be responsible for loss of manuscripts. Introduction and Summary. It is desirable to state the purpose and scope of the paper in an introductory paragraph and to give a summary of results at the end of the paper. Dictionary Style. It is recommended that authors fol- low capitalization, spelling, compounding, abbrevia- tions, etc., given in Webster’s New International Dic- tionary (unabridged), second edition; or, if desired, the Oxford Dictionary. Abbreviations of titles of pub- lications should, if possible, follow those given in World List of Scientific Periodicals. Footnotes. Footnotes should be ' used sparingly and never for citing references (see later). When used, footnotes should be consecutively numbered by supe- rior figures throughout the body of the paper. Foot- notes should be typed in the body of the manuscript on a line immediately below the citation, and sepa- rated from the text by lines running across the page. Citations of Printed Sources. All references cited should be listed alphabetically by author at the end of the paper, typed double-spaced. References to books and to papers in periodicals should conform to the following models: Batzo, Roderick L., and J. K. Ripkin. 1849. A Treatise on Pacific Gastropods. Rice and Shipley, Boston, vii + 326 pp., 8 figs., 1 map. Crawford, David L. 1920*. New or interesting Psyllidae of the Pacific Coast (Homop.). Proc. Hawaii. Ent. Soc. 4(l):12-l4. 1920&. The sandalwoods of Hawaii. Proc. Hawaii. Ent. Soc. 4(2) :374-375, 13 pis. In the text, sources should be referred to by author, date, and page, as follows: "It was noted (Rock, 1916:18) that . . ” or "Rock (1916:21-24) says ..." Quotations. Quoted matter of fewer than five printed lines (about 200 characters) should be given in the text in the usual form, using double quote marks. Longer quotations should be set flush with left mar- gin. The author is responsible for the accuracy of quoted material. Numbers. Decimals, measurements, money, percent- ages, time; enumerations in which any figure is 10 or over; and isolated enumerations of 10 and over should be given in Arabic figures, rather than spelled out, except when the number begins a sentence. ILLUSTRATIVE MATTER Only the minimum number of illustrations required to supplement the text will be accepted by the editors. Reproduction costs of illustrations in excess of the number allowed by the editors will be paid by the author. Artwork for illustrations should be 8V2 x 11 inches or smaller, and it should accompany manuscript, on separate sheets. Often more valuable than a photo- graph is a good line drawing. Figures and Graphs. Copy for figures and graphs should always be drawn large enough to allow for at least one-third reduction by the engraver. Copy should consist of carefully prepared line drawings in one color only, drawn in India ink on plain white draw- ing paper or tracing cloth. Co-ordinate paper with lines in light blue (a color which will not show in a photograph) may be used; but co-ordinates which should appear in the finished graph must be drawn in India ink. If original figures may not be conven- iently submitted with manuscript, duplicate rough sketches or photographic prints may be furnished to aid the editors in their decisions. It is strongly urged that an indication of scale be incorporated as a part of all drawings in which mag- nification and size are critical considerations. Photographs. Photographs should be chosen for clarity in portraying essential information. They should be printed for contrast, on glossy paper, and should be sent unmounted. They should be identified with serial number written in soft pencil on the back to corre- spond with list of captions. Illustrations will be returned to the author. Tables. Tabular matter should be kept to a minimum. Each table, prepared to conform with Pacific Science style, should be typed on a separate page, and its posi- tion indicated on the manuscript. Mathematical Formulas. Complicated formulas cannot be set by the printers. Authors should submit them as illustrations. Captions. Readily identifiable captions for figures, graphs, photographs, and other illustrative matter should be supplied on a separate page. PROOF Proof should be corrected immediately and returned at once to John S. Howe, Jr., assistant to the editors. Authors are reminded that the editors will allow only a minimum number of corrections on galley proof. Additions to the printed text and changes in style and content are not permitted. All queries on proof should be answered. If cor- rected proof is not received within four weeks after being sent to the author, author’s changes cannot be accepted. REPRINTS Reprints or separates should be ordered on the form provided and returned with author’s proof. All correspondence concerning separates must be directed to the printer, Heffernan Press Inc., 35 New Street, Worcester, Massachusetts 01605. The publisher does not keep a reprint file. VOL. XXI OCTOBER 1967 NO. 4 PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION YUZO KOMAKI Surface Swarming of Euphausiid Crustaceans in Japan GARY J. BRUSCA Ecology of Pelagic Amphipoda, II ANTHONY J. PROVENZANO, JR. Zoeal Stages and Glaucothoe of Pacific Hermit Crab ANGELES ALVARINO Bathymetric Distribution of Chaetognatha, Siphonophorae, Medusae, and Ctenophorae off California REGINALD M. GOODING and JOHN J. MAGNUSON Significance of a Drifting Object to Pelagic Fishes W. J. R. LANZING The Dendritic Organ and the Plotosidae JOHN Q. BURCH and ROSE L. BURCH The Family Olividae HAROLD ST. JOHN Revision of the Genus Pandanus, Parts 23, 24, and 25 | HANS R. HOHL and SUSAN T. HAMAMOTO Reversal of Ethionine Inhibition by Methionine in Slime Molds LYNETTE D. OSBORNE Accelerated Laboratory Tests of Fijian Timber Species' Resistance to Decay NOTES INDEX UNIVERSITY OF HAWAII PRESS BOARD OF EDITORS O. A. Bushnell, Editor-in-Chief Department of Microbiology, University of Hawaii John S. Howe, Jr., Assistant to the Editors University of Hawaii Press Gladys E. Baker Department of Botany University of Hawaii Sidney C. Hsiao Department of Zoology University of Hawaii George W. Chu Departments of Microbiology and Zoology University of Hawaii Charles H. Lamoureux Department of Botany University of Hawaii Maxwell S. Doty Department of Botany University of Hawaii John J. Naughton Department of Chemistry University of Hawaii Augustine S. Furumoto Hawaii Institute of Geophysics University of Hawaii Walter R. Steiger Department of Physics University of Hawaii Philip Helfrich Hawaii Institute of Marine Biology University of Hawaii Goro Uehara Department of Agronomy and Soil Science University of Hawaii Robert W. Sparks, Director University of Hawaii Press INFORMATION Contributions to Pacific biological and physical science will be welcomed from authors in all parts of the world. (The fields of anthropology, agriculture, engineering, and medicine are not included.) Manu- scripts may be addressed to the Editor-in-Chief, PACIFIC SCIENCE, University of Hawaii, Honolulu, Hawaii 96822, or to individual members of the Board of Editors. Use of air mail is recommended for all communications. Manuscripts will be acknowledged when received and will be read promptly by members of the Board of Editors or other competent critics. Authors will be notified as soon as possible of the decision reached. FOR AUTHORS Manuscripts of any length may be submitted, but it is suggested that authors inquire concerning possi- bilities of publication of papers of over 30 printed pages before sending their manuscripts. Authors should not overlook the need for good brief papers, presenting results of studies, notes and queries, com- munications to the editor, or other commentary. PREPARATION OF MANUSCRIPT It is requested that authors following the style of Pacific Science described herein and exemplified in the journal. Authors should attempt to conform with the Style Manual for Biological Journals , Am. Inst. Biol. Sci. Washington. (Continued on inside back cover) PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION VOL. XXI OCTOBER 1967 NO. 4 Previous issue published July 10, 1967 CONTENTS PAGE On the Surface Swarming of Euphausiid Crustaceans. Yuzo Komaki 433 The Ecology of Pelagic Amphipoda, IL Observations on the Reproductive Cycles of Several Pelagic Amphipods from the Waters off Southern California. Gary ]. Brusca 449 The Zoeal Stages and Glaucothoe of the Tropical Eastern Pacific Hermit Crab Trizopagurus magnihcus ( Bouvier , 1898) ( Decapoda ; Diogenidae) , Reared in the Laboratory. Anthony J. Provenzano, Jr. 457 Bathymetric Distribution of Chaetognatha, Siphonophorae, Medusae, and Cteno- phorae off San Diego , California. Angeles Alvarino 474 Ecological Significance of a Drifting Object to Pelagic Pishes. Reginald M. Gooding and John J. Magnuson 486 A Possible Relation between the Occurrence of a Dendritic Organ and the Dis- tribution of the Plotosidae (Cypriniformes) . W.f. R. Lanzing 498 The Family Olividae. John Q. Burch and Rose L. Burch 503 Revision of the Genus Pandanus Stickman, Part 23. Three Australian Species of Pandanus. Harold St. John 523 Revision of the Genus Pandanus Stickman, Part 24. Seychellea, a New Section from the Seychelles Islands. Harold St. John 531 Pacific Science is published quarterly by the University of Hawaii Press, in January, April, July, and October. Subscription prices: institutional, $10.00 a year, single copy, $3.00; individual, $5.00 a year, single copy, $1.25. Check or money order payable to University of Hawaii should be sent to University of Hawaii Press, 535 Ward Avenue, Honolulu, Hawaii 96814, U. S. A. Printed by Heffernan Press Inc., 35 New Street, Worcester, Massachusetts 01605. CONTENTS ( continued ) PAGE Revision of the genus Pandanus Stickman, Part 25. Pandanus tectorius var. sinen- sis Warburg . Harold St. John 533 Reversal of Ethionine Inhibition by Methionine during Slime Mold Development. Hans R. Ho hi and Susan T . Hamamoto 534 Comparative Decay Resistance of Twenty- five Fijian Timber Species in Acceler- ated Laboratory Tests. Lynette D. Osborne 539 NOTES Notes on the Hawaiian Flora. Benjamin C- Stone 550 Notes on the Ecology of the Pogonophoran Genus Galathealinum Kirke- gaard, 1956. Oluwafeyisola S. Adegoke 558 Notes on the Systematic Status of the Eels Neenchelys and Myroconger, Gareth ]. Nelson 562 Record of a Lane el et from Hawaii. L. G. Eldredge 564 News Note 565 INDEX 567 On the Surface Swarming of Euphausiid Crustaceans1 Yuzo Komaki2 ABSTRACT: A general aspect of the daytime surface swarming of Euphausia pact pea in Japanese nearshore waters is described in connection with the water temperature. Swarming usually starts with a local minimum temperature around 7°C and terminates with a temperature just below 16°C. The swarming season is essentially in spring, from February through May, with little difference among regions. The main swarming areas are on the Pacific coast around Kinkazan, and on the coast of the Sea of Japan around Sadogashima, in Wakasawan and its vicinity, around Oki, and on the east side of the Tsushima Gunto. It is shown that the swarming is closely related to cold water masses, and that the approach of offshore cold water masses to the nearshore areas and the mixing process in the coastal areas may provide favorable conditions for swarming. Swarming of E. pacipca is a phenomenon that occurs at the margins of the cold water bodies, and is related to the seasonal change in the geographical distribution of those euphausiids. A uniformly low water temperature profile must be the necessary condition for swarming, but other possible factors stimulating euphausiids to swarm are enumerated. It has been reported frequently from vari- ous parts of the world that conspicuous day- time aggregation of euphausiids takes place at the surface in rather nearshore waters (Table 1). The animals swarm in such large numbers that the sea surface turns red or brownish-red from their red and/or orange pigments. This phenomenon, as Bigelow (1926) pointed out, differs from their usual vertical migratory be- havior because it occurs in the daytime, inde- pendent of the light intensity. Most previous reports on this peculiar be- havior of euphausiids have been descriptive, and very few observations have been made on its relationship to environmental factors. This may be attributed not only to the complexity of animal behavior in general, but also to the capriciousness of such phenomena. As is the 1 Contribution No. 400 from the Department of Oceanography, University of Washington, Seattle. The manuscript was prepared with support from National Science Foundation Grant GB-3360. Manu- script received July 23, 1966. 2 Present Address: Department of Fisheries, Fac- ulty of Agriculture, University of Tokyo, Tokyo, Japan. case with red tides, the surface swarming of euphausiids is not predictable from physical measurements. Such a swarming phenomenon with Euphausia pacipca occurs in the nearshore waters sur- rounding Japan. A fishery based on E. pacipca is maintained in certain areas — around Kin- kazan (Komaki, 1957), in Wakasawan Bay and vicinity, and along the coast of northwest- ern Kyushu. The local fishermen scoop E. pacipca swarming at the surface with a pyra- mid-shaped net. The surface swarming of euphausiids is regarded as sporadic, but the existence of the euphausiid fishery means that the surface swarming of E. pacipca must be at least an ordinary annual phenomenon. While participating in the Survey of the Warm Tsushima Current and Related Waters during the period from 1953 to 1958, the author carried out ecological studies on eu- phausiids and reported briefly on the relation- ship between water temperature and the swarm- ing of E. pacipca (Komaki and Matsue, 1958). The present paper emphasizes this relationship more extensively, employing additional data, information, and references. 433 434 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE l Representative Reports on Euphausiid Swarming REFERENCE SPECIES* REGION SEASON Smith (1879) M. norvegica Eastport area autumn Murray (1888) M. norvegica Loch Fyne, Scotland H. F. Moore (1898) T hysanopoda spp. Eastport area summer, fall Lo Bianco (1902) M. norvegica around Capri I. Feb., June, July 1901 Bigelow (1926) M. norvegica Gulf of Maine spring-fall T. rascloii T. inermis MacDonald (1927) M. norvegica Oslo Fjord fall, early win- ter Hjord and Ruud (1929) M. norvegica off Mpre, Norway spring-fall Hardy and Gunther (1935) E. superba S. Georgia waters Nov.-Feb. Fish and Johnson (1937) Thysanoessa spp. Bay of Fundy Manteufel (1938, 1941) T. inermis Barents Sea winter-spring Mossentzova (1939) Barents Sea spring Dakin and Colefax (1940) N. australis Sydney area Sept. 1938 Einarsson (1945) M. norvegica N. Atlantic spring-summer Thysanoessa spp. Gunther (1949) E. superba S. Georgia waters Jan. 1937 Uda (1952) E. pacifica southern part of Sea Feb -May 1948 of Japan Sheard (1953) N. australis S. Australia, Bass breeding sea- T. gregaria Strait, S. Victoria, son N. Tasmania wa- ters Fisher et al. (1953) M. norvegica Monaco coasts Aug. 1951 Jan. 1952 Boden et al. (1955) T. spinifera La Jolla coasts June 1948 Peters (1955) E. superba Bouvet area in Ant- December arctic waters Komaki (1957) E. pacifica around Kinkazan Feb -May Komaki and Matsue (1958) E. pacifica Japanese waters Feb -May Ponomareva (1955, 1959, T. longipes northern part of Sea Mar.-June 1963) T. inermis of Japan, N. Pa- T. raschii cific E. pacifica Zelickman (1961) T. inermis Murman coasts Mar.-July T. raschii Brinton (1962^) E. pacifica north of Pt. Con- Apr. 1956 ception Marr (1962)t E. superba S. Georgia waters 1 * Abbreviations of genus names: E. = Euphausia, M. = Me ganyctip banes, N. = Nyctiphanes, T. i= Thysanoessa. f Review of surface swarming of E. superba in Antarctic waters. i SOURCES OF INFORMATION swarming of euphausiids was obtained from fishermen. Fishermen are the best, most fre- Swarming quent, and most experienced observers of phe- Because of the infrequency of swarming of nomena occurring in their favorite fishing euphausiids in the Sea of Japan since 1953, grounds. From 1954 to 1956, a questionnaire direct observation of the ph« momenon was not was sent three times to more than 800 local ( possible, but valuable information on the local fishermen’s unions, which are scattered along Surface Swarming of Euphausiids — Komaki 435 Fig. 1. (a), Map showing partition of coastal areas of the Sea of Japan into 27 zones in order to observe the regional differences in euphausiid swarming. Zones 15 and 16 cover the Wakasawan area. (b), Fishing ground (shaded area) of Euphausia pacifica in the vicinity of Kinkazan. the coasts of Honshu, Kyushu, and adjoining small islands facing the Sea of Japan. The rate of response to the questionnaire was approximately 30%. To facilitate analysis of the replies, the coastline of the Sea of Japan was partitioned into 27 zones of 0.5° latitude or according to geographical features of the coasts (Fig. la) . The fishermen’s unions were divided into 27 groups corresponding to these zones, so that geographical differences in the swarming, if any, could be detected. There were many responses suggesting interesting relationships between the swarming and en- vironmental factors (hydrographical, meteoro- logical, biological, and so on) . The author interviewed fishermen from important regions, on the basis of the results of the questionnaire, such as the Wakasawan area. Information that could be treated numerically (at least to some extent) was employed in the present paper. It was learned that, along the Pacific coast, the area around Kinkazan (Fig. lb) is the only place where the relatively stationary swarming is observed every year and the euphausiid fish- ery is maintained. The author visited this area in every swarming season and participated in the euphausiid fishing operation in order to make direct observations. No written question- naires were employed there. Reliable quanti- tative records of euphausiids fished from the Kinkazan district since 1953 were obtained from the fish market in Onagawa where al- most all the euphausiids from this fishing ground were landed, but no similar numerical data on the yield of euphausiids were available from the Sea of Japan. Water Temperature Among the important environmental factors governing distribution and behavior of the organisms, water temperature is not only the most important factor, but also is the one for which data can be obtained most easily. Ac- cordingly, these data were sought from the files of the hydrometeorological observations made by the various maritime and fisheries agencies listed below: Federation of Fishermen’s Unions of Miyagi Pre- fecture, Onagawa Branch. 1953-1959. Fisheries Statistics. Fukushima Prefecture, Fisheries Experimental Sta- tion. 1956. Data on oceanographic observations off Fukushima Prefecture, 1914-1939- Imperial Fisheries Research Office, Central Labora- tory. 1915-1950. Data on oceanographic surveys in the surrounding areas of Japan. Iwate Prefecture, Fisheries Experimental Station. 436 PACIFIC SCIENCE, Vol. XXI, October 1967 1954-1956. Data on oceanographic observations off Iwate Prefecture. Japan Sea Regional Fisheries Research Laboratory. 1953-1959- Monthly report on the out-lined sea conditions of the Sea of Japan. Ministry of Agriculture and Forestry, Section of Statistics. 1946-1956. Fisheries Statistics. Miyagi Prefecture, Fisheries Experimental Station. 1910-1959- Records of hydrometeorological obser- vations at Enoshima, Miyagi Prefecture. Plankton Samples Collected in the Sea of Japan Thousands of plankton samples were col- lected during the period 1953-1958 by various institutions that participated in the Survey of the Warm Tsushima Current and Related Wa- ters. Most samples were collected in a strip within 100 miles of the coastline. The selected samples were examined for the purpose of eco- logical studies on euphausiids, and they were very useful in the confirmation of euphausiid species composing surface swarms. RESULTS The Species Examination of specimens landed at the Onagawa fish market revealed that the eu- phausiids swarming in the Kinkazan waters were large specimens of Euphausia pacifica (longer than 20 mm from the tip of the ros- trum to the end of the telson). Throughout four fishing seasons, only one stray specimen of Nematoscelis diffcilis was found among the catches of huge quantities of E. pacifica. Com- bined swarmings of more than two species, such as those reported by H. F. Moore (1898), Fish and Johnson (1937), and Zelickman (1961), were not encountered in the Kinkazan waters. The following five species of euphausiids inhabit the Sea of Japan (Ponomareva, 1955; Komaki and Matsue, 1958) : E. pacifca, Thysanoessa raschii, T. inermis, T. longipes, and Pseudeuphausia latifrons. Plankton samples collected during the sur- vey showed that E. pacifca and P. latifrons could be obtained from the areas within 100 miles from the coast. Thysanoessa spp. never have been collected from that area, although Ponomareva (1959, 1963) has reported that the three Thysanoessa species mentioned ap- pear at the surface, forming remarkably dense swarms in the northernmost part of the Sea of Japan. P. latifrons is a small, warm-water form, shorter than 10 mm in total length (Hansen, 1916; Brinton, 1962^), and appar- ently it penetrates into the southernmost part of the Sea of Japan from the Tsushima Kaikyo only during summer and fall (Komaki and Matsue, 1958). Samples from surface swarms of a few occasions revealed that the swarms were composed of large E. pacifca only. Thus, E. pacifca must be the species composing the surface swarms in the nearshore waters of the Sea of Japan. Features of Swarming Hardy and Gunther (1935) have described beautifully the swarming behavior of E. su- perha at the surface, and their descriptions can be exactly applied to the surface patches of E. pacifca. The animals swim as close as 1-2 cm from each other, orienting themselves in one direction as if commanded by a leader. The swarms look like formless clouds, and fre- quently change in shape. They are red, brown- ish-red, or pale brown in color, depending upon their distance from the surface. When participating in euphausiid fishing operations in the Kinkazan waters, the author found that swarming took place intensively at intervals of a few days. As shown in Figure 2, the daily landings of E. pacifca at the Ona- gawa fish market fluctuate with a pulselike rhythm. Inasmuch as only surface patches of euphausiids are harvested because of the fish- ing method, the landing on a given day is probably related to the standing crop of euphausiids swarming at the surface on that day. Daily catch per unit of fishing effort is the most suitable term to use in discussing such a fluctuation of standing crop, but data indicating fishing effort were not available. Fishing boats unload their catches at the mar- ket two or three times a day while euphausiids are abundant. Accordingly, the aggregate num- bers of fishing boats in Figure 2 change ap- proximately in parallel with the landings of euphausiids. Efforts were made to relate this pulselike occurrence of swarming to the environmental factors that vary within short periods, e.g., irradiation, wind direction and intensity, and Surface Swarming of Euphausiids — Komaki 437 Fig. 2. Daily landings of Euphausia pacifica ( solid line ) from Kinkazan waters and daily change of aggregate number of fishing boats ( dashed line) yielding euphausiids (1959). the like, but there were no apparent relation- ships. The yield on a calm day is likely to be more abundant than that on a rough day, but this is caused essentially by the relative difficulty of the fishing operation. The time of swarming throughout a given day is not definite. It occurs in early morning on some days, while on other days it takes place in the afternoon even in bright sunshine. Swarming Season The fishing season of euphausiids corre- sponds with the swarming season in the area where the euphausiid fishery is carried out. As shown in Figure 2, the swarming season during 1959 in the Kinkazan waters started in the middle of March and terminated at the end of May. The season, however, changes slightly from year to year. The swarming season in the Kinkazan waters ranges between late February and late May in maximum extent. Figure 3 shows the general aspect of the swarming season along the coasts of the Sea of Japan. The rectangles in Figure 3 show the range of the swarming season in the zones indi- cated by the numbers in Figure la. Figure 3 was derived from replies of fishermen to the question, "In what month (s) do you usually observe the euphausiid swarming in your favor- ite fishing grounds?" The three degrees of swarming intensity of euphausiids were fixed as follows: 1-3 affirmative response (s) from a certain zone in a given month were expressed by a white area, 4-6 affirmative responses by a shaded area, and more than 7 by a black area. Actually, each month expressed by a black area represents more than 15 affirmative responses. Figure 3 also indicates that the major swarm- ing regions in the coastal areas of the Sea of Japan are around Sadogashima (zone 8), in Wakasawan and vicinity (zones 15 and 16), off Sanin district including Oki (zones 18-21), and on the east side of the Tsushima Gunto (zone 24). There seems to be a tendency for swarming to take place earlier in the year in the southern part of the Sea of Japan. Thus, in the north of Kyushu (zones 23-25), the most intensive swarming is observed in February and March; in the Wakasawan and Sanin districts (zones 15-21), the most notable swarming season is in March and April (also in May in some zones) ; and in the area surrounding Sadoga- shima (zone 8), the euphausiids aggregate most actively in April through June. According to the fishermen’s information, the swarming season in each zone changes from year to year, as in the Kinkazan waters. The beginning, peak, and terminating times of the swarming are different from year to year. The swarming season along the coasts of the Sea of Japan occurs between January and June. Seasonal Change of Water Temperature Figure 4 shows the mean surface tempera- ture and salinity cycles throughout a year at three representatives coastal points in the Sea of Japan. All temperature records for a given month, taken over a period of 26 years, were averaged. The same was done with salinity determinations. Upon comparing Figure 4 with Figure 3, it may be seen that the occurrence of swarming is closely associated with colder water tempera- tures. Swarming starts at a slightly higher tem- perature than the local minimum, continues with increasing temperature, and then termi- nates as the temperature exceeds 16°C. The 438 PACIFIC SCIENCE, Vol. XXI, October 1967 ZONE RANGE (°N) OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV ZD X CD 2 o X 3 X CO ZD 1 2 3 4 5 6 7 8 9 10 I I 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 41.0 -40.5 40.5- 40.0 40.0- 39.5 39.5- 39.0 39.0- 38.5 38.5- 38.0 38.0- 37.5 38.0- 37.5 37.5- 37.0 37.0- 36.5 37.5 - 37.0 37.5 -37.0 37.0- 36.5 36.5- 36.0 36.0- 35.5 36.0-35.5 36.0-35.5 36.0- 35.5 35.5- 35.0 35.0- 34.5 36.5- 36.0 34.5- 34.0 34.5-34.0 34.5- 34.0 34.0- 33.5 33.5- 33.0 33.0- 32.5 Fig. 3. Swarming season of Euphausia pac/fica in the coastal areas of the Sea of Japan (see text). Zones are shown in Figure la. temperature during the swarming season ranges between 7° and 16°C. Figure 5 a shows the average change of sur- face water temperature during the first six months at Enoshima, a tiny island located in the middle of the fishing ground for euphau- siids around Kinkazan (Fig. lb). Daily hydro- meterological observations have been made at this island since 1910, with an interruption during the period from 1945 to 1953. In this Pacific coastal region, swarming starts with the minimum surface water temperature (in Feb- ruary and March), continues as the temperature rises, and ends when it reaches about 12°C, thus demonstrating again a relationship between swarming and low temperature. The vertical distribution of temperature and its annual change was also examined for the Kinkazan waters. Figure 6 shows the mean annual changes of temperature profiles down to 200 m at two points 10 miles off Ozaki ( a ) and Shioyazaki (b) , respectively. It was not possible to learn the results of long-term oceano- graphic observations carried out at a definite station adequately close to Kinkazan, which would have been an ideal station. Therefore these two stations were substituted. Monthly observations have been made along the west- east lines, including the above two stations as the nearest ones to the coast, for 20 years off Ozaki and for 24 years off Shioyazaki. Figure 6 shows that sea water is completely mixed from the top down to a depth of 200 m, and that a i low temperature prevails during the months corresponding to the swarming season of [ euphausiids. It seems probable that the water temperature ' profile in the euphausiid fishing ground near Kinkazan is more nearly similar to that off j Ozaki than to that off Shioyazaki, which is located in the south where the sea conditions are more directly influenced by the warm Surface Swarming of Euphausiids — Komaki 439 Fig. 4. Annual changes of surface water temperature and salinity at Nyudozaki (a), Kyogasaki ( b ), and Okinoshima (c). Locations are shown in Figure la. Measurements were made from 1915 to 1950 with interruptions in some years. Kuroshio. Such a condition of low and uniform temperature as prevails from the surface down to 200 m in the vicinity of Kinkazan provides no temperature barriers to euphausiids that may be dwelling in the depths (Boden and Kampa, 1958). In other words, euphausiids, during their vertical migration, do not encounter the temperature difference between day and night levels that has been discussed by H. B. Moore (1952). Annual Variations in Swarming The statistics for annual yields of euphausiids recorded at the Onagawa fish market (Table 2) indicate remarkable annual changes in euphausiid abundance in the vicinity of Kin- kazan. Each value in Table 2 can be inter- preted as an index of the standing crop of swarming euphausiids in each swarming sea- son, because there were no notable changes in fishing effort (i.e., the number of fishing boats) throughout these seven years, and almost all euphausiids fished in the area were landed at the Onagawa fish market. Similarly, in the nearshore waters of the Sea of Japan, it was also observed that the extent of swarming fluctuates widely from year to year. Although profuse swarming oc- curred during the period from 1943 through 1949 (Uda, 1952), it has been observed only rarely since 1953. The relative abundance of euphausiid swarming during the period 1945- 1956 in the nearshore waters of the Sea of Japan is shown in Figure 7. Each symbol repre- sents an answer from an individual fisherman. Figure 7 illustrates that intensive swarming in the Sea of Japan took place until 1953 but ceased thereafter. Although no numerical data 440 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE 2 Annual Yields of Euphausia pacifica from the Kinkazan Area* YEAR YIELD (TONS) 1953 431.8 1954 0.0 1955 887.1 1956 1,029.2 1957 274.0 1958 404.1 1959 1,419-1 * From the statistics of the Onagawa fish market. are available, the information at hand indicates that the disappearance of swarms from near- shore waters continued at least until 1958. In order to relate the annual change in the euphausiid swarming to the surface water tem- perature, examination of water temperatures during the first six months of the year at Enoshima (Fig. 5a) are plotted in Figures 5b (1954) to 5 g (1959) with 10-day inter- vals. Each point represents the mean values of 10 daily measurements. Upon relating Figure 5 to Table 2, it is quite obvious that the temperature was abnor- mally higher than the mean throughout the winter and spring of 1954, when absolutely no euphausiids were caught. On the other hand, in 1956 and 1959, when more than 1,000 tons of euphausiids were captured, the temperature was a little lower than the mean in February and March 1956, and it was a little higher than, but close to, the mean in April and May 1959. Euphausiids were fished mainly in the early spring of 1956, while they were caught more abundantly late in the spring of 1959 than earlier (Fig. 2). The peculiarity of the water temperature in the winter and spring months of 1954 also can be learned from Figure 8. Figure 8 is based on the monthly observations at the sta- tion 10 miles off Tsubakishima (Fig. la), which is located closer to Kinkazan than is Ozaki. Figure 8a shows the water temperature profile down to a depth of 200 m during the period from November 1953 to November 1954. Figure 8b shows the temperature profile at the same station for the following year. It is obvious that a remarkable temperature gra- dient was present during the winter and spring months of 1954, when no euphausiids were Fig. 5. Surface water temperatures at Enoshima. (a), Average temperatures for 1910-1944 and 1953- 1959; {b)-{g), deviations from the mean for 1954- 1959. captured. Considerable warm water occupied the surface layers in the spring of 1954, espe- cially in April, which corresponds to the middle of the swarming season in the Kinkazan wa- ters. The temperature profile for the next year (Fig. 8b) shows normal seasonal change in the area, as is indicated by the temperature profiles at the stations off Ozaki and Shioyazaki (Fig. 6). The records of water temperatures observed Surface Swarming of Euphausiids — Komaki 441 NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV Fig. 6. Average annual changes of water temperature profile 10 miles off (a) Ozaki during 1919-1939, and (b) Shioyazaki during 1914-1939. Locations are shown in Figure la. at Kyogasaki were the only data available for relating temperature to the long-term change in swarming in the Sea of Japan shown in Figure 7. The monthly mean values of water temperatures in a given month from 1943 to 1957 are plotted in Figure 9. It appears that all water temperatures in March, from 1943 to 1952, were lower than the accumulated av- erage value calculated over the years beginning with 1915 (dashed lines in Fig. 9). March is the month when swarming takes place exten- sively in the Wakasawan area. Also, the least squares regressions (solid lines in Fig. 9) ob- tained from values during these 14 years sug- gest an explanation for the disappearance of euphausiids from the neritic areas. All regres- 442 PACIFIC SCIENCE, Vol XXI, October 1967 ZONE YEAR 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 2 • * © O 4 • * X 0 © ©o ©o ©O 5 • * © * © ©O © ©oo ©©O 6 0 © • -* 0 o 7 • © o II O o o © © HONSHU 12 © * © o o 15 •• * • *** ••• **•* •© ** ©©©© * e®©©© * ©0©O * ©oo ooooo OOOO 16 • ** • V V V Y V AA A /v7v © V V V V V ATUVAA © *** * — ©0© ©00© •#©0© ©0©©Q OGOOO GQOQO 0000 0000 17 * • * oo OO O ©0©O ©0OO 18 © ©© •© ** * * © © ©O OO •OOO ©OOO 19 * * •• *** • •0 *-** © * ©0 * ©©©(X) ©oco (XXX) •©OOO ©eooo 21 * * * © © o ©©©©© oo •©€©© ©OO 22 • ** * * ©0OOO ©0OOO ©0000 ©OOOO 3 23 • « ©© o o ©ooo ©ooo X CD 3 24 • © * • © © * © * © • © * ©0®©© * ©©©0© oo ©00©© ooo >- 26 ** © * © © OO oo (I) O (2) © (3) © (4) © 15) • (6) * Fig. 7. Fluctuation in relative abundance of Euphausia pacifica swarming at the surface in the coastal areas of the Sea of Japan, (l) a few, (2) less than usual, (3) usual, (4) more abundant than usual, (5) much more abundant than usual, (6) extraordinarily abundant. sions have a positive slope, which means that successive years tended to be warmer during this period, and this tendency is greatest in March and April (Fig. 9 b and c) when the most extensive swarmings take place. Predators It was learned from fishermen that there is a close relationship between the abundance of euphausiids at the sea surface and fishes in the swarming areas of euphausiids. Various pre- vious workers — Smith (1879), Lebour (1924), MacDonald (1927), Hjord and Ruud (1929), Sheard (1953), Zelickman (1961), and Marr (1962) among others — also reported the eu- phausiid swarms accompanied by predators such as fishes, whales, and birds. In order to take this relationship into account, an investi- gation was made on representative predators. Figure 10 shows the change in annual yield of Japanese mackerel, Scomber japonic pis, from the Wakasawan area. This was obtained from the statistics of the Ministry of Agriculture and Forestry. Spring mackerel fishing in the area usually is carried out during the period approximately corresponding to the swarming season of euphausiids, and examination of stomach contents of mackerel has shown that the mackerel is predatory on euphausiids. A conspicuous increase in the yield from 1945 to 1946 (Fig. 10) probably was caused by resto- ration of the local fishing fleet from the war- time decline, i.e., the increase of fishing effort. In 1949, when an extraordinary abundance of Surface Swarming of Euphausiids — Komaki 443 CO QC 100 200 Fig. 8. Water temperature 10 miles off Tsubakishima. fall 1954; (b), fall 1954 to fall 1955. mackerel was fished, abnormally conspicuous swarming of euphausiids was observed in this area (Fig. 7). Rapid decrease in mackerel yield occurring in 1952 and thereafter may be related to the disappearance of the spring swarms of euphausiids from the coastal area after that year. Location shown in Figure la. ( a ), Fall 1953 to Table 3 shows the change in annual yield of sand eel, Ammodytes personatus, from the Kinkazan waters. In general, the fishing season of sand eel in this area starts in January and terminates by the end of July, and they are taken from almost the same area as the euphausiid fishing ground shown in Figure lb. 444 PACIFIC SCIENCE, VoL XXI, October 1967 Fig. 9. Monthly mean temperature in each of four spring months from 1943 through 1957 at Kyogasaki. Dashed lines show mean temperatures, 1915-1957; solid lines show least squares regressions for 1943- 1957. Comparing Table 3 with Table 2, it is obvious that annual yields of both euphausiids and sand eel from the area fluctuate in parallel. In addition to the sand eel, the sea gull, Larus crassirostris, and a small blackish bird, Cerorhinca monocerata, can be considered pred- ators on euphausiids in the Kinkazan waters. According to local fishermen, C. monocerata can dive into the deep, and it probably attacks submerged euphausiids from the bottom and drives them up to the surface. DISCUSSION It has been clearly demonstrated here that the swarming of Euphausia pacifica in the nearshore waters of Japan is closely related to Fig. 10. Annual yield of Japanese mackerel. Scomber japonicus, from the Wakasawan area. cold water temperatures. This may be expected, since the species is a boreal form, occurring commonly north of the subarctic convergence, as shown by Banner (1949), Boden et al. (1955), Nemoto (1957), Brinton (1962^), and Ponomareva (1963). The habitat is cres- cent in shape, covering the northern part of the North Pacific. E. pacifica occurs in the cold water lying under the upper strata in the Sea of Japan (Komaki and Matsue, 1958). Inasmuch as the branches of the warm Kuroshio wash the Japanese coasts, E. pacifica may be excluded from the coastal areas except during the coldest season, while it is more commonly concentrated in the nearshore and inshore waters of the eastern North Pacific (Banner, 1949; Boden et al., 1955; Brinton, 1962^, b\ Banse and Semon, 1963; Regan, 1963). Except during its swarming season. TABLE 3 Annual Yield of Sand Eel, Ammodytes personaius, from the Kinkazan Area* YEAR YIELD (TONS) 1953 741.7 1954 349.6 1955 580.8 1956 4,029.1 * From the statistics of the Onagawa fish market. Surface Swarming of Euphausiids — Komaki E. pacifica is seldom found in the stomachs of various fishes caught on the continental shelves around Japan. Therefore, the approach of the offshore populations of E. pacifca into near- shore waters must be the first step of swarm- ing in the coastal area around Japan. It may be considered as a seasonal expansion of their distribution to the coastal areas, inasmuch as Beklemishev and Semina (1956) and Semina (1958) demonstrated that the seasonal change of zoo- and phytogeographical boundaries ac- companied the seasonal shift of the conver- gence between Kuroshio and Oyashio. The size of the approaching offshore popu- lations may be strongly influenced by tempera- ture conditions from year to year. Uda (1964) illustrated and discussed the meanderings of the Kuroshio. Masuzawa (I960) discussed the annual variation of the Kuroshio axis, and showed that the north-south swing of the axis around the point of 36°N, 144°E was much greater than that of the other portions, and also that there is a possibility that the isolation of the water bodies may take place at the top of the conspicuous current axis curvature. The point mentioned above is located in the south- east off Kinkazan, and therefore it is quite possible that an isolated warm water mass from the top of the northward curvature of the Kuroshio approaches the coastal region around Kinkazan in some cases. The extraordinarily high temperature observed in the winter and spring months of 1954 possibly was caused by such an approach of the isolated water mass, although Masuzawa’s data of 1954 through 1959 showed that the northernmost meander- ing of the current axis occurred in 1955 and not in 1954. Miyata and Shimomura (1959) and Miyata (I960) classified several cold water masses in the offshore areas of the Sea of Japan and dis- cussed their location and transference. Uda (1952, 1958) discussed the variation of the conditions in this sea, describing the change of the position of the polar front from year to year. It cannot be doubted that the offshore cold water masses come close to the Honshu coasts in winter and spring when the monsoon blowing from Siberia prevails. Actually, the cold water masses defined by the above workers are located off the zones 445 where the swarming of euphausiids takes place most actively, e.g., around Sadogashima (zone 8), in the Wakasawan area (zones 15 and 16), and around Oki (zone 21) (Fig. la). In winter and spring these cold water masses may protrude against the coastal areas in the man- ner of a tongue. The warm water of the Tsushima current is very shallow because of the shallowness (about 100 m) of the Tsushima Kaikyo. Accordingly, it is likely that the originally warm Tsushima current water will be mixed with cold water masses through relatively simple processes. The stronger the approach of offshore cold water masses to the coast and the stronger the mixing in the coastal areas, the more profuse will be the s warmings of euphasiids along the coasts on the Sea of Japan. As a matter of fact, Uda (1958) showed that the polar front came very close to the Honshu coast during the period from 1946 to 1949, while it was away from the coast after 1952. It appears that such approach and recession of the polar front corre- sponds in time to the change in swarming (Fig. 7). Thus, low temperature induces the coastward approach of the offshore stocks of E. pacifca, and it must be one of the indispensable con- ditions for surface swarming in Japanese near- shore areas. However, it would be premature to conclude that it is the only sufficient con- dition. One is still unable to explain why euphausiids do not swarm in early and middle winter months when the temperature is as low as it is in spring and vertical mixing is actively taking place, and why they come to the surface in the daytime when the light intensity may be harmful to them. There have been several different opinions as to the cause of the daytime surface-swarming of euphausiids: (a) Predators may drive euphausiids to the surface, as previously men- tioned. (b) Euphausiids come to the surface to search for food in the upper strata where phytoplankton is abundant (Paulsen, 1909; Manteufel, 1938, 1941). (c) Current condi- tions may accumulate euphausiids, or stimulate them to swarm at the surface (Fish and John- son, 1937; Einarsson, 1945; Peters, 1955). (d) Some internal demands related to matu- ration or reproduction may drive euphausiids 446 PACIFIC SCIENCE, Vol. XXI, October 1967 to the sea surface (Sheard, 1953; Ponomareva, 1959, 1963; Zelickman, 1961). Although these theories are still controversial, the last can prob- ably be applied to the swarming of E. pacifica in Japanese nearshore waters. The author observed that more than 50% of the females in a few swarms in the Sea of Japan had spermatophores in the thelycum. No specimens from the Kinkazan area were carry- ing spermatophores. However, the females were full-grown and the degree of maturation of the ovaries corresponded to the stage 3 de- fined by Ruud (1932) or to the stage 5 or 6 established by Bargmann (1945) for Euphausia superb a. As is shown in Figure 2, offshore populations of E. pacifica seem to come close to the coast around Kinkazan with a pulselike rhythm. This may suggest that the population of E. pacifica can be divided into several stocks in accordance with the phase of maturation, and that, as stocks reach a certain degree of maturity, they approach the coast in succession. The tendency for swarming to occur earlier in the southern part of the Sea of Japan than in the northern part, as shown in Figure 3, may be understood in relation to the geo- graphical difference between the maturing or reproduction phases of euphausiids and to sea- sonal differences in temperature of the two areas. Kun’s opinion (1955) on the daytime ascent of Calanus tonsus, which may be related to a biochemical process (e.g., transformation of vegetable carotenoids into vitamin A by ultra- violet radiation) may be applied to the swarm- ing of euphausiids. During certain periods of their maturing or reproductive process, eu- phausiids might need rather strong daylight, regardless of the usual daytime level of their vertical distribution. As listed in Table 1, nearly all species whose swarming has been reported hitherto are cold water forms. This suggests that their physiology should be ana- lyzed, with the objective of solving the mech- anism of this peculiar behavior of euphausiids. ACKNOWLEDGMENTS The author is indebted to Dr. Yoshiyuki Matsue, former professor at the University of Tokyo, who kindly oriented and supervised the initial phase of this work. Special appreciation is due to Dr. Joyce C. Lewin and Dr. Karl Banse, Department of Oceanography, Univer- sity of Washington, and to Dr. Michael Mul- lin, Institute of Marine Resources, University of California, for their critical reading of the manuscript. REFERENCES Banner, A. H. 1949. A taxonomic study of the Mysidacea and Euphausiacea (Crusta- cea) of the northeastern Pacific, Part III. Euphausiacea. Trans. Roy. Can. Inst. 28: 1-63. Banse, K., and D. Semon. 1963. On the ef- fective cross-section of the Isaacs-Kidd mid- water trawl. Univ. Washington Dept. Oceanogr. Tech. Rept. 88:1-9. Bargmann, H. E. 1945. The development and life history of adolescent and adult krill, Euphausia superba. Discovery Rept. 23: 103-176. Beklemishev, K. V., and G. I. Semina. 1956. On the structure of the biogeographical boundary between the boreal and tropical regions of the pelagial of the northwestern Pacific Ocean. [In Russian.] Dokl. Acad. Nauk SSSR 108(6) :1057-1060. Bigelow, H. B. 1926. Plankton of the off- shore water of the Gulf of Maine. Bull. Bur. Fish. Washington 40(2):l-486. Boden, B. P., M. W. Johnson, and E. Brinton. 1955. The Euphausiacea (Crus- tacea) of the North Pacific. Bull. Scripps Inst. Oceanogr. 6(8) :287-400. and E. M. Kampa. 1958. Lumiere bioluminescence et migrations de la couche diffusante profonde en Mediterranee occi- dentale. Vie Milieu 9(1) :1-10. Brinton, E. 1962a. The distribution of Pa- cific euphausiids. Bull. Scripps Inst. Ocean- ogr. New Ser. 8(2) :51-218. 1962 A Variable factors affecting the apparent range and estimated concentration of euphausiids in the North Pacific. Pacific Sci. 16(4) :374-408. Dakin, W. J., and A. N. Colefax. 1940. The plankton of the Australian coastal wa- ters off New South Wales, Part I. Publi- cations of University of Sydney, Department of Zoology, Monograph 1:1-215. Surface Swarming of Euphausiids — Komaki 447 Einarsson, H. 1945. Euphausiacea, I. North Atlantic species. Dana Rept. 27:1-185. Fish, C. J., and M. W. Johnson. 1937. The biology and zooplankton population in the Bay of Fundy and Gulf of Maine with spe- cial reference to production and distribu- tion. J. Biol. Bd. Canada 3(3) :189— 322. Fisher, F. S., S. K. Kon, and S. Y. Thomp- son. 1953. Vitamin A and carotenoids in some Mediterranean Crustacea with a note on the swarming of Meganyctiphanes. Bull. Inst. Oceanogr. Monaco 1021:1-19. Fraser, F. S. 1936. On the development and distribution of the young stages of krill ( Euphausia superba). Discovery Rept. 14:1- 192. Gunther, E. R. 1949. The habit of fin whales. Discovery Rept. 25:113-142. Hansen, H. J. 1916. The euphausiacean crus- taceans of the Albatross expeditions to the Philippines. Proc. U. S. Natl. Mus. 49: 635-654. Hardy, A. C., and E. R. Gunther. 1935. The plankton of the South Georgia whaling grounds and adjacent waters, 1926-1927. Discovery Rept. 11:1-456. Hjord, J., and J. T. Ruud. 1929. Whaling and fishing in the North Atlantic. Rapp. P.-v. Reun. Cons. Perm. Intern. Explor. Mer 56:1-123. Komaki, Y. 1957. On the "esada” ( Euphausia pac/fca H. J. Hansen) fishery in the vicinity of the Ojika Peninsula. [In Japanese.] In- formation Bull. Whales Research Inst. 60: 10-17. and Y. Matsue. 1958. Ecological studies on the Euphausiacea distributing in the Sea of Japan. [In Japanese.] Report of Cooperative Survey of the Warm Tsushima Current and Related Waters 2:146-162. Kun, M. S. 1955. About reasons of various behaviour of mackerel. [In Russian.] Trudy Tikhookean. Inst. Morsk. Ryb. Khoz. okeanogr. 43:206-208. Lebour, M. V. 1924. The Euphausiidae in the neighbourhood of Plymouth and their importance as herring food. J. Mar. Biol. Assoc. U.K. 13(2) :402-4l8. Lo Bianco, S. 1902. Le pesche pelagiche eseguite dal Maia nelle vicinanze di Capri. Mitt. Zook Stn. Neapel 15:413-482. 1903. Le pesche abissali da F. A. Krupp col yacht Puritan nelle adiacenze di Capri ed in altra localita del Medditerraneo. Mitt. Zook Stn. Neapel 16:109-279. MacDonald, R. 1927. Food and habits of Meganyctiphanes norvegica. J. Mar. Biol. Assoc. U.K. 14(3) :753-784. Manteufel, B. P. 1938. A short essay on changes of plankton in the Barents Sea. [In Russian with English summary.] Trudy Polar. Inst. Morsk. Ryb. Khoz. Okeanogr. 1:134-148. 1941. Plankton and herring in the Barents Sea. [In Russian with English sum- mary.] Ibid. 7:125-21 6. Marr, J. W. S. 1962. The natural history and geography of the Antarctic krill ( Euphausia superba Dana). Discovery Rept. 32:33 -464. Masuzawa, J. I960. Statistical characteristics of the Kuroshio current. Oceanogr. Mag. 12(1) :7— 15. Miyata, K. I960. Hydrographic characteristics of the polar front fishing grounds in the western part of the Sea of Japan (August- September, 1959). [In Japanese.] Report on the Polar Front Fishing Ground in the Sea of Japan, 1959, pp. 135-151. and T. Shimomura. 1959. On the transfers of low temperature regions in the north Japan Sea in spring. [In Japanese with English summary.] Bulk Japan Sea Regional Fish. Res. Lab. 7:1-16. Moore, H. B. 1952. Physical factors affecting the distribution of euphausiids in the North Atlantic. Bulk Mar. Sci. Gulf Caribb. 1 (4) :278-305. Moore, H. F. 1898. Observations on the her- ring and herring fisheries of the coast, with special reference to the vicinity of Passama- quoddy Bay. Rept. U. S. Fish. Comm. 1896, pp. 387-442. Mossentzova, T. N. 1939. The seasonal changes in the macroplankton in Barents Sea (based on materials collected in 1938). [In Russian with English summary.] Trudy Po- lar. Inst. Morsk. Ryb. Khoz. Okeanogr. 4:129-147. Murray, J. 1888. On the effects of winds on the distribution of temperature in the sea- and fresh-water lochs of the west of Scotland. Scot. Geogr. Mag. 4:345-365. 448 PACIFIC SCIENCE, Voi. XXI, October 1967 Nemoto, T. 1957. Foods of baleen whales in the northern Pacific. Sci. Rept. Whales Res. Inst., Tokyo 12:33-89. Paulsen, O. 1909. Plankton investigations in the waters around Iceland and in the North Atlantic in 1904. Meddr. Kommn Havun- ders., Serie: Plankton 1(8): 1-57. Peters, H. 1955. Uber das Vorkommen des Walkrebschens Euphausia superba Dana und seine Bedeutung fur die Ernahrung des siid- lichen Baren Wale. [English summary.] Arch. Fisch.-wiss. 6(5,6) : 288-304. Ponomareva, L. A. 1955. Nutrition and dis- tribution of euphausiids in the Sea of Japan. [In Russian.] Zool. Z. 34:85-98. 1959. Reproduction of Euphausiidae of the Sea of Japan and development of their early larval stages. [In Russian.] Ibid. 38(11) :1643-1662. 1963. The euphausiids of the North Pacific, their distribution and mass species. [In Russian with English summary.] Inst. Oceanol. Acad. Sci. U.S.S.R., Moscow, Monogr. : 1-142. Regan, L. 1963. Field trials with the Clarke- Bumpus plankton sampler. Inst. Oceanogr., Univ. British Columbia, Manuscript Rept. 16:1-28. Ruud, J. T. 1932. On the biology of southern Euphausiidae. Hvalrad. Skr. 2:1-105. Semina, G. I. 1958. Relation between phyto- geographic zones in the pelagial of the north- western part of the Pacific Ocean with the distribution of water masses in the region. [In Russian.] Trudy Inst. Okeanol. 27:66- 76. Sheard, K. 1953. Taxonomy, distribution and development of the Euphausiacea (Crus- tacea). British- Australian-New Zealand Re- search Exped. Rept., Ser. B, 8(l):l-72. Smith, S. I. 1879. The stalk-eyed crustaceans of the Atlantic coast of North America, north of Cape Cod. Trans. Connecticut Acad. Arts Sci. 5:27-138. Tattersall, W. M. 1936. Mysidacea and Eu- phausiacea. Sci. Rept. Barrier Reef Exped. 5(4):143-176. Uda, M. 1952. On the hydrographical fluctua- tion in the Japan Sea (preliminary report). Appendix: The extraordinary abundant catch of Euphausia pacifica Hansen in winter and spring of 1948 along the coast of the Japan Sea. [In Japanese with English summary.] Japan Sea Regional Fish. Res. Lab., Spec. Publ. for its Third Anniversary of Founda- tion, pp. 291-300. 1958. On the out-lined sea conditions of the Sea of Japan. [In Japanese.] Reports of the Cooperative Survey of the Warm Tsu- shima Current and Related Waters 1:31-45. 1964. On the nature of the Kuroshio, its origin and meanders, pp. 98-107. In: K. Yoshida, ed., Studies on Oceanography. To- kyo. Zelickman, E. A. 1961. The behavior pattern of the Barents Sea Euphausiacea and possible causes of seasonal vertical migrations. Intern. Revue Ges. Hydrobiol. 46(2) :276-281. The Ecology of Pelagic Amphipoda, II Observations on the Reproductive Cycles of Several Pelagic Amphipods from the Waters off Southern California Gary J. Brusca1 The purpose of this study was to utilize mid- water samples from off the coast of southern California to determine the general patterns of the reproductive cycles of the local pelagic Amphipoda. All of the materials analyzed dur- ing this project were collected in the area of the Outer Santa Barbara Passage at approxi- mately 33°20'N, 118°4(/W. All samples were collected with an Issacs-Kidd Midwater Trawl (Issacs and Kidd, 1953) from the R/V "Velero IV” of the Allan Hancock Foundation, Univer- sity of Southern California. Partial support for this work was furnished by National Science Foundation Grants (G10691 and G23467). Complete information on the reproductive cycles of amphipods is scarce and, in the case of pelagic species, difficult to obtain. Some data can be gained from various expedition reports in which the presence of ova and young in the brood pouches is recorded. Some of these past records are mentioned in this paper. METHODS AND MATERIALS Samples were taken at various depths and at different times of the day and night from the summer of 1962 through the spring of 1963 using a 10 X 1 0-ft Issacs-Kidd Midwater Trawl. A few samples were used from later in 1963 for qualitative confirmation of the data gathered earlier. Species accounts and analyses of vertical distributions and migrations are pre- sented by Brusca (1967). For comparative information on fluctuations in population densities throughout the year, counts made from pint aliquots were converted to the number of individuals captured per hour trawling time. Such values are only approximate 1 Assistant Director, Pacific Marine Station, De- partment of Biology, University of the Pacific, Stock- ton, California. Manuscript received October 13, 1966. and do not account for the suspected gregarious nature of these pelagic amphipods, but they probably do reflect the general trends in the population size. Complete raw data are on file with the author. OBSERVATIONS Presented in this section are only those data which specifically pertain to the reproductive cycles of these animals. For a more detailed species account and reference lists see Brusca (1967). Suborder gammaridea Family eusiridae Rhachotropis natator (Holmes) A total of 77 specimens was taken from pint aliquots. None of these individuals was noted to be carrying ova or young in the brood pouches, but there was some variation in the size range and in the density of the population. Table 1 illustrates these changes. Since a rise in population density accom- panied an extension of the lower limit of the size range during the winter months, this time probably represents the entrance of young into the adult or "catchable” population. The data TABLE 1 Seasonal Changes in Size Range (SR), and Average Number per Trawl Hour (n/th) Based on Positive Samples for Rhachotropis natator SEASON SR (mm) n/th Summer 20 (l spec) 4 Fall 13-17 22 Winter 9-16 32 Spring 9-17 24 449 450 PACIFIC SCIENCE, Vol. XXI, October 1967 are too scant, however, to draw any definite conclusions. Family lysianassidae Paracallisoma coecus (Holmes) During this study 56 specimens were re- covered from pint aliquots. The presence of ovigerous females was noted in October and November and a few were carrying young in May. Seasonal variations in number and size offer no conclusive data due to the small quan- tity of individuals collected. Cyphocaris anonyx Boeck A total of 1 1 5 individuals was collected from pint aliquots. Ovigerous females were noted during the months of May, June, July, and August, indicating that this season is a time of high reproductive activity. No females were found which were carrying young. As shown in Table 2, there was an extension of the lower limit of the size range during the summer months and an increase in population density, probably indicating the entrance of young into the mature population. The drop in mean size in the fall may reflect the death of the older portion of the population. Cyphocaris richardi Chevreux A total of 193 specimens was sorted from pint aliquots. Table 3 illustrates the pertinent reproductive information. As can be seen, the greatest production of eggs occurred during the fall months, and young were present in the brood pouches only in the winter. Males were most abundant at the times of high ova production. It appears that embryonic development took place through- TABLE 2 Seasonal Variation in Mean Size (MS), Size Range (SR), and Average Number per Trawl Hour (n/th) for Cyphocaris anonyx SEASON MS (mm) SR (mm) n/th Summer 11.5 4-14 37 Fall 8 6-12 4 Winter 13-5 5-15 13 Spring 10 7.5-13.5 30 out the winter months, with the release of young completed by early spring. There was a slight drop in mean size and in the lower limit of the size range in the spring, followed by a more drastic decrease in these measurements during the summer. These periods indicate the time of the entrance of young into the adult population. Suborder hyperiidea Family platyscelidae Platyscelus serratulus Stebbing Only 8 specimens of this species were taken, all of which were females. Although Hurley (1956) collected P. serratulus from southern California waters during the months of January and February, the individuals captured during this present study were taken in September and November. Ovigerous females were noted in November. Sizes ranged from 3 to 7 mm with a mean of 5.1 mm for the 8 specimens col- lected. Family pronoidea Eupronoe minuta Claus In all, 322 individuals were collected from pint aliquots. From the data presented in Table 4 it can be seen that some of the reproductive activity of E. minuta is not clear, in that the greatest production of ova occurred at the same time as the highest incidence of young in the brood pouches. Apparently young were released from the females early in the fall. This release is indicated by a drop in the percentage of females carrying young, a drop in mean size, and an extension of the lower limit of the size range. Males were present only during the fall months, suggesting that that was the period of fertilization. Family phrosinidae Primno macropa Guerin A total of 315 individuals was taken from pint aliquots. The pertinent reproductive data are recorded in Table 5. Maximum production of eggs occurred in the Ecology of Pelagic Amphipoda, II — Brusca 451 TABLE 3 Seasonal Variations in the Percentages of Mature Females Carrying Ova (O), Carrying Young (Y), Mean Size (MS), Size Range (SR), Female/Male Ratio (f/m), and the Average Number per Trawl Hour (n/th) Based on Positive Samples for Cyphocaris richardi SEASON O Y MS (mm) SR(mm) f/m n/th Summer 10% 0% 20 11-29 1.5/1 33 Fall 43% 0% 22 17-28 0.7/1 56 Winter 10% 13% 26 18-33 1/1 51 Spring 14% 0% 24 15-30 1.6/1 48 TABLE 4 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O) and Young (Y), Mean Size (MS), and Size Range (SR), and Average Number per Trawl Hour (n/th) for Eupronoe minuta SEASON O Y MS (mm) SR (mm) n/th Summer 35% 15% 6 4. 2-7.8 4 Fall 26% 0% 5.5 3-7.5 26 Winter 10% 0% 5.5 3. 5-7.5 8 Spring 25% 15% 5.7 5-7 5 TABLE 5 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O), and Young (Y), Mean Size of Females (MSf), and Males (MSm), Size Range of Females (SRf), and Males (SRm), and Average Number per Trawl Hour (n/th) for Primno macropa SEASON O Y MSf MSm SRf (mm) SRm (mm) n/th Summer 50% 5% 7 5 4-12 4-9 5 Fall 50% 3% 8.5 4.5 4-12 4-5 26 Winter 33% 0% 7 none 5-13 none 4 Spring 27% 25% 12 none 6-14 none 4 summer and fall months, with a high percentage of the females carrying young by spring. Males were captured only during times of highest ova production. The young were released into the adult population during the summer, as indi- cated by a drop in mean size, an extension of the lower limit of the size range, and an in- crease in the average number per trawl hour by the fall months. Family cystisomidae Cystisoma fabricii Stebbing Only 31 specimens were collected. Ovigerous females were noted in October and December, and some were carrying young during January and August. There was an increase in the num- ber of males collected at times of ova produc- tion. Because of the paucity of specimens no definite conclusions can be drawn. Cystisoma pellucidum (Suhn) Only 4 individuals were collected during the entire study period, including a single ovigerous female in July (64 mm). Other specimens were taken in August (male, 70 mm), January (male, 116 mm), and May (damaged, 70 mm). Family oxycephalidae Calamorhynchus pellucidus Streets A single male specimen (19 mm) was col- lected during September. Fage (I960) gave an account of the known reproductive biology of 452 PACIFIC SCIENCE, Vol. XXI, October 1967 this species in his monograph on the oxy- cephalids. Oxycephalus clausi Bovallius Only two individuals of this species were taken (September: female, 30 mm; and Jan- uary: female, 25 mm). Neither of these speci- mens was carrying ova or young. Again, Fage (I960) discussed the reproductive biology of O. clausi. Streetsia challengeri Stebbing This species was by far the most common oxycephalid taken during this study; 67 speci- mens were collected from pint aliquots. Table 6 gives information regarding breeding activity. Egg production was highest in the spring, at which time the ratio of females to males was lowest. Young were most prevalent in the brood pouches during the fall. They were probably released from the parents in late fall and early winter, as indicated by a drop in mean size during the fall. The winter size data do not correlate with the suggested time of the en- trance of young into the adult population. Fage (I960) reviewed some of the reproductive biology of this species. Family hyperiidae Hyperia spinigera Bovallius It has been suggested by Shoemaker (1945) that Hyperia spinigera may be conspecific with H. galba. Only six specimens of this species, all males, were collected throughout this study. The months of capture and sizes of individuals were as follows: August: 1 male (5 mm); Septem- ber: 1 male (4 mm); October: 3 males (4,5, 15 mm); November: 1 male (4 mm). The presence of this species during only the late summer and early fall may have some bear- ing on the reproductive activity, but paucity of individuals prevents speculation. Hyperia bengalensis (Giles) A total of 52 individuals was sorted from pint aliquots. Specimens were captured only during the months of August through Novem- ber. During this time the mean size of the female population increased from 2 mm to 2.8 mm, and the size range for females increased from 2-2.4 mm to 1.7— 3-5 mm. Males were taken only in September, October, and Novem- ber, and were consistently larger than the females displaying a mean size of 33 mm and a size range of from 3 to 4 mm. The ratios of females to males for these three months were 5/1, 15/1, and 3/1, respectively. Egg production was highest in September, with about 40% of the females being gravid; values of less than 20% were noted for the other three months. The percentages of females carrying young in the brood pouches increased throughout the four months during which H. bengalensis was captured (August, 0%; September, 10%; October, 15%; November, 45%). The reasons underlying the odd and sudden appearance and disappearance of this species from the local population are unclear and con- sequently the breeding cycle is incomplete. Hyperia galba (Montague) A total of 178 specimens was collected from pint aliquots. Table 7 summarizes the reproduc- TABLE 6 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O) and Young (Y), Mean Size of Females (MSf) and Males (MSm), Size Range of Females (SRf), and Males (SRm), Female/Male Ratio (f/m), and Average Number per Trawl Hour (n/th) for Streetsia challengeri MSf MSm SRf SRm season O Y (mm) (mm) (mm) (mm) f/m n/th Summer 20% 30% 21.5 22 13-27 20-24 8/1 2 Fall 0% 40% 14 15.5 7.5-26 14-17 11/1 2 Winter 0% 0% 22 13 21-22.5 11.5-14 2/1 less than 1 Spring 51% 0% 19 14 12-23 11.5-16 3/1 2 Ecology of Pelagic Amphipoda, II — Brusca 453 TABLE 7 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O) and Young (Y), Mean Size of Females (MSf) and Males (MSm), Size Range of Females (SRf) and Males (SRm), and the Female/Male Ratio (f/m) of Hyperia galba SEASON O Y MSf (mm) MSm (mm) SRf (mm) SRm (mm) f/m Summer 0% 45% 9 12 8-12.5 11-13 4.5/1 Fall 10% 17% 9 11.5 6-19.5 9-15 4.6/1 Winter 50% 18% 10.5 10.5 7.5-12.5 6-15 1.5/1 Spring 98% 0% 9 11.5 8.5-10 8.5-17 1.5/1 tive information. Egg production was highest in the summer months, and the highest inci- dence of young in the brood pouches was in the spring. The abundance of males does not appear to correlate with the presence of ova. Young were released from the parents by sum- mer, but their entrance into the mature popula- tion was not reflected in the size data until fall, at which time there was an extension of the lower limit of the size ranges for both males and females. This suggestion is also supported by an increase in population density from an average of 3 individuals per trawl hour in the summer to 10 per trawl hour during the fall months. There appears to be about a three- months’ lag between the release of young from the brood pouches and their entrance into the catchable population. The whereabouts of the newborn amphipods is unknown. Analyses of local plankton samples taken during the sus- pected time of release offered no positive in- formation. Family vibiliidae Vibilia armata Bovallius A total of 2,742 individuals was sorted from pint aliquots. Some of the reproductive data for this species are difficult to interpret. Stephensen (1918) reported on collections from the Mediterranean in which he found breeding females in January and February and from June through September and young at all times of the year. In this present study no ovigerous females were noted during the winter and only a low percentage was recorded for the rest of the year (Table 8). The presence of young in the brood pouches suggests that the release from the parents took place in the late spring and early summer, at which time an ex- tension of the lower limit of the size range and a drop in mean size were noted. The popula- tion density, however, did not show a signifi- cant increase until early fall. Vibilia viatrix Bovallius In all 658 individuals were recorded from the sorting of pint aliquots. Stephensen (1918) reported specimens of V. viatrix in breeding condition during March and October in the North Atlantic. The information gathered in this present study is given in Table 9. Ovigerous females were most abundant in the fall and a high percentage was carrying young by winter. Apparently juveniles were re- leased from the brood pouches during the TABLE 8 Seasonal Variation in the Percentages of Mature Individuals Carrying Ova (O) and Young (Y), Mean Size (MS), Size Range (SR), and Average Number per Trawl Hour (n/th) for Vibilia armata SEASON O Y MS(mm) SR(mm) n/th Summer 3% 10% 6.5 3-10 14 Fall 8% 35% 7.5 6-10 251 Winter 0% 45% 7 6-9 71 Spring 8% 51% 8.5 6-10 30 454 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE 9 Seasonal Variation in the Percentages of Mature Individuals Carrying Ova (O) and Young (Y), Mean Size (MS) and Size Range (SR), and Average Number per Trawl Hour (n/th) for Vibilia viatrix SEASON O Y MS(mm) SR (mm) n/th Summer 12% 25% 9 4-15.5 38 Fall 20% 40% 9 7-12 8 Winter 9% 60% 10.5 7-14 5 Spring 3% 21% 11.5 8.5-15 3 spring and summer months, as is shown by a drop in mean size and an extension of the lower limit of the size range along with an increase in the average number per trawl hour. All of these indications became obvious during the summer. Family phronimidae Phronima sedentaria (Forskal) A total of 575 individuals was sorted from pint aliquots. Previously reported reproductive data on P. sedentaria were reviewed by K. H. Barnard (1932), who indicated that, in the more northern regions of its distribution, this species has its highest period of reproductive activity during the summer and fall months. Table 10 gives the breeding information gathered in this present study. The peak of egg production occurred in the summer. Since the developing young of this species are carried for some time on the inner walls of the salp "barrels” in which P. seden- taria is known to live, and only a few of these "barrels” were collected, probably the per- centages of females with young are inaccurate. Because of the confusion regarding the release of young from the parents (or from the bar- rels), it is difficult to analyze the size variations as related to breeding season. In addition to this problem, one can see from Table 10 that the extensions of the lower limits of the size range for males and females do not coincide. In spite of these difficulties, however, males were most prevalent during the times of high egg production and the other data do indicate greatest reproductive activity during the sum- mer and fall months, concurring with Barnard’s 1932 report. Family paraphronimidae Paraphronima gracilus Claus A total of 472 specimens was collected from pint aliquots. Hurley (1956) reported a single female with young in the brood pouches during August. His work was conducted in the local southern California area near the collection sites of this present study. As indicated in Table 11, the highest per- centages of ovigerous females were noted in the fall and females carrying young were most TABLE 10 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O) and Young (Y), the Mean Size of Females (MSf) and Males (MSm), Size Range of Females (SRf) and Males (SRm), the Female/Male Ratio (f/m), and the Average Number per Trawl Hour (n/th) for Phronima sedentaria SEASON O Y MSf (mm) MSm (mm) SRf (mm) SRm (mm) f/m n/th Summer 53% 20% 25 16 11-35 11-17 2.4/1 20 Fall 20% 3% 22 14 14-33 9-18.5 2.2/1 15 Winter 7% 22% 27 13 13-35 11-15 10/1 8 Spring 25% 19% 28 none 18-37 none no males 3 Ecology of Pelagic Amphipoda, II — Brusca 455 TABLE 11 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O) and Young (Y), Mean Size of Females (MSf) and Males (MSm), Size Range of Females (SRf) and Males (SRm), and the Average Number per Trawl Hour (n/th) for Paraphronima gracilus season O Y MSf (mm) MSm (mm) SRf (mm) SRm (mm) n/th Summer 8% 50% 13.5 10.5 9-16.5 10-14 8 Fall 43% 47% 10 10.5 9.5-13 10-11 33 Winter 24% 55% 11 10.5 10-12 10-11 18 Spring 2% 75% 11 11 10-12 10-12 20 TABLE 12 Seasonal Variation in the Percentages of Mature Females Carrying Ova (O) and Young (Y), Mean Size of Females (MSf) and Males (MSm), Size Range of Females (SRf) and Males (SRm), and the Average Number per Trawl Hour (n/th) for Paraphronima. crassipes season O Y MSf (mm) MSm (mm) SRf (mm) SRm (mm) n/th Summer 7% 50% 20 16.5 14-28 13.5-23 16 Fall 49% 49% 19 18.5 13-26.5 14-23 40 Winter 24% 56% 22 20 20-28 18-22 52 Spring 2% 75% 24 18.5 13.5-31 13-24 41 abundant in the summer. The female/male ratio was relatively low at the suggested time of ova production (2/1), and there was an increase in the average number per trawl hour from 8 per trawl hour in the summer to 33 per trawl hour in the fall, indicating the release of young into the mature population. There was also a drop in the mean size of the female population during the fall, but other size data offer no correlations. Paraphronima crassipes Claus A total of 922 individuals were recovered from pint aliquots. Hurley (1956) reported specimens collected in local waters during July and August to be carrying ova or young in their brood pouches. Table 12 gives the reproductive information regarding P. crassipes gathered in this present study. The highest percentage of females with ova was noted in the fall, followed by a high inci- dence of young in the brood pouches in the winter and spring. Apparently young were re- leased from most of the females in the summer and fall, during which time there was a drop in mean size and, by fall, an increase in the population density. The ratio of females/males was lowest during the fall (1/1), correspond- ing with the production of eggs. CONCLUSIONS During this study certain general trends were noted regarding the reproductive activity of the total amphipod population. As illustrated in Table 13, the highest production of ova occurred during the summer and fall months, with de- velopment of young continuing throughout the following spring and summer. Most species re- leased their young into the adult population by early fall, as indicated by the percentages listed in Table 13 along with the increase in the total amphipod population density. It can be seen that, although the above men- tioned trends are observed, precise analyses of the activities of individual species are complex and difficult to make. A sampling program de- TABLE 13 Seasonal Variation in the Percentages of Mature Females with Ova (O) and Young (Y) in the Brood Pouches, and the Number per Trawl Hour (n/th) for the Total Amphipod Population SEASON O Y n/th Summer 24% 40% 49 Fall 20% 32% 135 Winter 10% 35% 50 Spring 14% 43% 40 456 PACIFIC SCIENCE, Vol. XXI, October 1967 signed specifically to obtain large numbers of individuals throughout the year is needed in order to eliminate density errors due to vertical migrations and gregarious activity. The fact that amphipods brood their young makes the group an ideal one for embryological studies if enough material can be obtained. The development and ecology of newborn amphipods could provide excellent problems for future research. REFERENCES Barnard, K. H. 1932. Amphipoda. Discovery Rept. 5:283-284. Brusca, G. J. 1967. The ecology of pelagic Amphipoda, I. Species accounts, vertical zona- tion and migration of Amphipoda from the waters off Southern California. Pacific Sci. 21(3) :382-393. Fage, L. I960. Oxycephalidae, amphipodes pelagiques. The Carlsberg Foundation’s Oceanographical Expedition Round the World 1928-30 and previous "Dana” ex- peditions. Dana Rept. 52:24-26, 33-37, 54-63. Hurley, D. E. 1956. Bathypelagic and other Hyperiidea from California waters. Allan Hancock Foundation, Occas. Papers 18:12- 13, 21-22. Issacs, J. D., and L. W. Kidd. 1953. Issacs- Kidd Midwater Trawl. Scripps Inst. Ocean og., Ref. 53-3:1-21. Shoemaker, C. 1945. The Amphipoda of the Bermuda Oceanographic Expeditions, 1929- 1931. Zoologica 30:185-266. Stephensen, K. 1918. Hyperiidea- Amphipoda (part 1), D,2. Report on the Danish Ocean- ographical Expeditions 1908-1910 to the Mediterranean and Adjacent Seas. Vol. II, Biology, C-D, 3, pp. 41-43. The Zoeal Stages and Glaucothoe of the Tropical Eastern Pacific Hermit Crab Trizopagurns magnificus (Bouvier, 1898) (Decapoda; Diogenidae), Reared in the Laboratory1 Anthony J. Provenzano, Jr. ABSTRACT: Larvae were reared under various temperature conditions. Those maintained at 15°C were unable to moult to the second instar although some in- dividuals lived as long as 35 days after hatching. At 20°C some individuals were able to reach fifth instar, but glaucothoes were obtained only at 25 °C, 33-52 days after hatching. Effects of starvation and temperature on larval survival are dis- cussed. The number of zoeal stages in the development of this species is variable, as it is in other diogenids which have been studied in the laboratory, glaucothoes of this species being obtained after four or five zoeal instars. Descriptions and illustrations of the zoeal stages and the glaucothoe are presented. No other larvae of this genus have been described and intra-generic comparisons of larval mor- phology were not possible, but a comparison was made of the zoeal and glaucothoe stages of this species with those of others in the family. The eastern Pacific contains a relatively rich hermit crab fauna but one which is still rather poorly known systematically. In spite of the great number of species which occur even in shallow waters from Alaska to the Equator, very little is known concerning the life histories or larval development of eastern Pacific hermit crabs. One of the first successful attempts to rear anomuran larvae in the laboratory was made by Hart (1937), who described larval stages of two species of Pagurus and one of Paguristes and of the mud shrimp Upogebia, all from British Columbian waters. Coffin (I960) studied another species of Pagurus. No papers describing development of any tropical eastern Pacific hermit crab have been published. The genus Trizopagurus occurs in tropical seas around the world with the exception of the Caribbean (Forest, 1952). In the eastern Pa- cific, the genus is represented by T. magnificus, a not uncommon hermit crab of moderate size, 1 Contribution No. 830 from the Institute of Marine Science, University of Miami, Miami, Florida. This work was supported by research grants Nos. 16298 and GB-4305 from the National Science Foundation and grant No. GM-11244 from the Insti- tute of General Medical Sciences, U. S. Department of Health, Education and Welfare. Manuscript re- ceived September 14, 1966. black with orange spots and orange antennae and antennules. The species was first described by Bouvier (1898) as Clibanarius magnificus, and was redescribed and illustrated by Boone (1932) as Clibanarius chetyrkini. Forest (1952) recognized it as belonging to his newly estab- lished genus. The species is distributed from the Gulf of California southward at least as far as La Plata Island, Ecuador, and occurs also in the Galapagos Islands, but nothing is known of its ecology. The limited data available from various systematic papers which have dealt with T. magnificus and the data accompanying specimens in various collections are sufficient to indicate that this species seems to prefer rocky areas, from the intertidal zone down to a few tens of meters. The female from which the larvae were obtained for the present study was collected in an area where the substrates con- sisted of rocky patches surrounded by mud. Species collected with Trizopagurus in this lo- cality included Dardanus s inis tripes Stimpson and Clibanarius panamensis Stimpson, both typical of inshore waters along the major part of the range of T. magnificus. C. panamensis is most often found in brackish water and muddy areas, often close to mangrove shores. Also collected with adults of T. magnificus were a species of Isocloeles, about which virtually 457 4 58 PACIFIC SCIENCE, Vol. XXI, October 1967 nothing is known, and a specimen of an un- described species of Clibanarius. In recent years there has been an increasing amount of effort to study the larvae of the several families of hermit crabs especially to obtain ontogenetic information useful in clas- sification and phylogeny of the group. No lar- vae of the seven currently recognized species of Trizopagurus have been studied previously. The purpose of the present work is to provide de- scriptions of the zoeae and glaucothoe of this tropical eastern Pacific hermit crab based on laboratory reared specimens, and to make avail- able the limited ecological data obtained inci- dental to the rearing experiments. ACKNOWLEDGMENTS I am indebted to the National Science Foundation and to the Institute of General Medical Sciences, U. S. Public Health Service for their support of this work. Dr. F. M. Bayer of this Institute initiated the study by bringing an ovigerous female to Miami by air from Panama, thus providing the opportunity for study at this laboratory of a genus not found in the West Indian faunal region. Dr. A. L. Rice and C. Edith Marks helped with the rearing. Osvaldo Moran-Ribeau did many dis- sections and made preliminary study sketches. Barbara Stolen made the illustrations. METHODS Several females were collected in January 1964 by F. M. Bayer and R. Chesher at sev- eral intertidal localities near Venado Island in the Bay of Panama off the Canal Zone. One female Trizopagurus magni ficus retained her eggs during passage back to Miami and yielded larvae during 16-18 January 1964. The tem- perature of the running sea water in which the female was kept was less than 20 °C for one week prior to hatching and was 18°C during the hatching period. More than 800 larvae were obtained from this hatching and were placed in plastic compartmented trays as de- scribed in previous papers (Provenzano, 1962a; Provenzano and Rice, 1964), 1-10 larvae per compartment. Trays to which no food was added were placed in four experimental tem- peratures (10°, 15°, 20°, and 25 °C) to de- termine survival time of starved larvae. Addi- tional trays, to which Artemia nauplii were added as food for the zoeae, were placed at the same temperatures. The three lower tempera- tures were maintained by thermostatically con- trolled refrigerators with fans to circulate air within the cabinets so that temperatures during the experimental periods did not vary more than db0.1°C. For the highest temperature an air-conditioning unit in the culture room kept air temperature at 25°C=b 1.5°C. Two lots of filtered sea water were used dur- ing the experiments, 32.6 parts per thousand from 16 January-3 February and 35.7 parts per thousand from 3 February to termination of the experiments. Specimens and exuvia were preserved in al- cohol or formalin. Specimens were cleared in 3-5% KOH and whole specimens and exuvia were dissected after staining with Mallory’s acid fuchsin red, lignin pink, or chlorozol black, and were mounted. Study sketches were made with a Bausch & Lomb microprojector, and details were checked under higher magni- fication using a Tasco compound microscope. Final drawings were made with the aid of a Wild binocular M-5 dissecting scope equipped with a camera lucida. The term stage is used herein in the sense of instar or intermoult. All scales in the illustrations represent 0.5 mm. Carapace length of zoeae was measured from the tip of the rostrum to the most pos- terior lateral margin of the carapace, not to the dorsal posterior margin. Total length was measured from tip of rostrum to the median posterior margin of the telson exclusive of tel- son spines. Because of the flexible nature of the abdomen, the total length is less reliable a measure than the carapace length, which is based on a rigid structure. In the glaucothoes, shield length was measured from the tip of the rostrum to the cardiac suture. Carapace length was taken from the tip of the rostrum to the dorsal posterior transverse margin, and total length was measured from the tip of the ros- trum to the posterior margin of the telson exclusive of setae. The female from which the larvae were Zoeal Stages and Glaucothoe of Trizopagurus magni ficus — Provenzano 459 hatched has been deposited in the U. S. Na- tional Museum (Catalog No. 113559). EXPERIMENTAL RESULTS Effect of Starvation at Various Temperatures In order to determine the maximum survival time for unfed animals, several trays of larvae without food were placed in each experimental temperature. At 25°, 20°, and 15°C, 7-8 days were required for 50% mortality of the 54 starved larvae in each temperature, but at 10°C only 3 days were required for 50% mortality of 36 larvae. At 25 °C, total mortality of the starved group required 10-12 days; at 20 °C, 9-11 days; at 15°C, 8-10 days; and at 10°C all larvae were dead by the sixth day after hatching. Survival at Various Temperatures of Larvae Fed with Artemia At 10°C, 36 larvae were placed two per compartment. They began dying on the third day and by the seventh day all were dead. At 15°C, 145 larvae were placed one, two, or five per compartment. None moulted to stage II. By 21 days after hatching, approximately one- half had died, but a few survived as long as 35 days, then died in stage I. At 20°C, 329 larvae were placed in trays, one, two, four, or more per compartment. Most moulted to stage II within 13-18 days after hatching, but a few lived to stage III. None became glaucothoe, but three specimens lived to stage V and died at approximately 85 days after hatching. At 25 °C, 305 larvae were placed in trays, one, two, or four per compartment. Nearly all survived the first moult, which took place 7-8 days after hatching. Glaucothoes were obtained at this temperature in as few as 33 and as many as 52 days after hatching. Only 18 glaucothoes were obtained. One specimen spent 23 days as a glaucothoe, then died in the moult to first crab stage, 56 days after hatching. CAUSES OF MORTALITY At 10°C the mortality of fed animals paral- leled quite closely that for starved animals, in- dicating either that, despite presence of food, the animals were unable to feed or that the temperature alone was sufficiently low to kill the animals directly. Even at the higher tem- peratures starved larvae did not swim during the last few days. Hence we may suppose that in nature larvae unable to feed within the first few days after hatching seldom survive as long as they did in these experiments, but nothing is known of the capacity of larvae to resume feeding and normal growth after vary- ing periods of starvation. The shorter survival time at lower temperatures indicates that, at least at temperatures below 25 °C, the exhaus- tion of yolk reserves was not the factor causing death among starved larvae, but that tempera- ture had a direct negative effect on survival. Because of the large number of larvae hatched and the limited time available to tend to them, some were placed together in compartments. It is unlikely that crowding was a primary cause of mortality since each compartment con- tained 40-60 ml of water and, in a few compartments in which as many as 10 larvae were together, survival was better than in many others with fewer animals. There was no ap- parent negative effect of crowding on survival. It is obvious that the temperatures used were mostly below the satisfactory range for this species. At 10°C the larvae could not swim and died very quickly even though they had been gradually reduced to that temperature from the hatching temperature of only about 18°C. At 15°C the larvae were below the temperature at which normal development must take place, since none of them were able to moult. The fact that some lived as long as 35 days indicates that a few must have been able to feed at least occasionally even at that tem- perature, for starved larvae at 15°C were all dead by the tenth day after hatching. Even at 20 °C larvae were apparently under very mar- ginal conditions, since only three out of 329 lived to stage V. At 25 °C, although the percentage of sur- vival to metamorphosis was low (18 glau- cothoes were obtained from 305 original lar- vae), and although none of the glaucothoes actually survived to crab stage, the temperature was probably satisfactory, if still less than optimal. Contributing to the high mortality under laboratory conditions at 20° and 25°C 460 PACIFIC SCIENCE, Vol. XXI, October 1967 was an infection by a filamentous fungus-like organism which has occasionally struck experi- ments in the laboratory but which has not been identified. It is unfortunate that higher experi- mental temperatures were not available at the time. The 7-8 days required to reach the first moult at 25 °C is approximately the same amount of time as is required by some other tropical species of hermit crabs in the labora- tory, but is longer than for others. This period would almost certainly be shortened by sev- eral days at still higher temperatures. DESCRIPTION OF THE LARVAL STAGES There may be four or five zoeal instars in the development of T. magnifcus prior to the glaucothoe stage. General Features of the Zoeal Stages The rostrum is long, exceeding the cephalic appendages, rather broad and deep, with the tip slightly curved ventrad. Each of the anterio- ventral corners bears a small blunt spine pro- jecting anteriolaterally. The carapace bears no large spines posteriolaterally on the margins, but has numerous spinules which give the carapace a roughened appearance. These spi- nules extend onto the more posterior portions of the body as well, being especially noticeable on the dorsal surface of the abdominal somites and on the telson. As development proceeds, the spinules become relatively smaller until they are hardly noticeable in the last zoeal stage (Figs. 1 and 2). The telson is much broader than long in the first stage and in sub- sequent stages becomes progressively more elongate (Fig. 7). The appendages are generally symmetrical throughout larval development, with occasional differences of one or two setae between one side and the other, but a notable exception is the pair of mandibles which are quite asym- metrical throughout the zoeal stages. Because zoeal mandibles have seldom been described or illustrated in detail, the functional and pos- sible systematic significance of mandible arma- ture is not well understood. Therefore both mandibles of each stage of this species have been illustrated from two aspects. The zoeal stages have a yellow-orange overall color. Some of the parts of the exoskeleton, notably the tip of the rostrum and the ends of the antennules, are yellowish but not from chromatophores. The carapace has a very dif- fuse yellow-orange color, also apparently not due to chromatophores. There are orange-red chromatophores laterally under the anterior half of the carapace, and others deep in the body at the bases of the maxillipeds, and there is a very large orange chromatophore on each side of the fifth abdominal somite near the base of the lateral spines. There are two pairs of similar large orange chromatophores anteriorly on the telson. Red chromatophores are found at the base of the antennae, on the labrum, and perhaps on the bases of the mandibles. A pair of red ones occurs on the first abdominal somite. First Zoea CARAPACE length: 1.4-1. 6 mm TOTAL LENGTH: 2. 7-2. 9 mm (3 specimens) The first larval instar, as is typical of hermit crab larvae generally, has the eyes fused to the carapace. The sixth abdominal somite is fused to the telson, which bears the normal comple- ment of 7 -f- 7 marginal telson processes, the outermost of which is a heavy spine, the sec- ond a delicate hair, while the others are articu- lated plumose setae. The appendages of the first zoea (refer to figures) differ in no important respect from those of other species of hermit crabs at that stage (except that the well formed anterio- lateral spine on the antennal scale is not always present in other diogenid hermit larvae and the medio-proximal comer of the basipodite of the first maxilliped has only setae, not a hooked process as in some other species of Diogenidae) . Second Zoea CARAPACE LENGTH: 1.8-1. 9 mm total length: 3.5— 3.8 mm (4 specimens) The second larval instar differs from the first in many respects. The eyes are now free of the carapace and are stalked. The telson, while still fused to the sixth abdominal somite, has added a median pair of telson processes. All of the appendages have changed as shown in the figures. The antennule has added some terminal aes- thetascs, for a normal total of 6 or 7 terminal Zoeal Stages and Glaucothoe of Trizopagurus magni ficus — Provenzano 461 Fig. 1. Trizopagurus magni ficus. Dorsal views of the four zoeal stages and the glaucothoe. 462 PACIFIC SCIENCE, Voi. XXI, October 1967 Fig. 2. Trizopagurus magni ficus. Lateral views of the four zoeal stages and the glaucothoe. Zoeal Stages and Glaucothoe of Trizopagurus magni ficus — Provenzano 463 I II III IV G Fig. 3. Trizopagurus magnificus. Top , the antennule of zoeae I-IV and glaucothoe; bottom, the antenna of zoeae I-IV and glaucothoe. processes and 3 subterminal setae at approxi- mately the location of the future articulation of the rami. The antenna has changed little, adding only a seta on the scale and a small tooth at the base of the scale. The mandibles are not changed significantly. The maxillule now has 4 strong teeth on the distal endite instead of 2. There may be a very short seta on the proximal segment of the endopodite, but usually it is not dis- cernible. The maxilla has added 1 or 2 setae to the scaphognathite and 1 or 2 setae to some of the basal and coxal endites. The first maxilliped has added 2 natatory setae to the exopodite. The endopodite has lost the row of extremely fine setules on the lateral margins of the segments and has added to the three most proximal segments a single 464 PACIFIC SCIENCE, Vol. XXI, October 1967 Fig. 4. Trizopagurus magnificus. The paired mandibles of zoeae I-IV. Left column, posterior surface; right column, anterior surface. Third Zoea carapace length: 2.15-2.40 mm (8 speci- mens) TOTAL length: 4.25-4.55 mm (5 speci- mens) The most obvious gross change is that the telson is now articulated with the sixth ab- dominal somite, and a pair of uropods has ap- peared. The posterior telson margin bears 8 — (— 1 — (— 8 telson processes, the median process being articulated and of the same type | as the adjacent ones. However, the fourth pro- cess from each side is much enlarged, non- plumose, and is fused to the telson. There is long plumose seta at the distal lateral margin. The appendage is otherwise basically un- changed. The second maxilliped, like the first, has lost the fine row of setules on the endopodite, replacing them with long plumose setae on the two middle segments, and the exopodite has added 2 natatory setae. The third maxilliped, a mere bud in the first stage but now functional, consists of a basipodite bearing an exopodite with 5 or 6 natatory setae. On the basipodite a lobe which will be the endopodite originates proximally and may bear a terminal seta. Zoeal Stages and Glaucothoe of Trizopagurus magni ficus — Provenzano 465 now a pair of fine plumose setae submarginally on the dorsal surface of the telson. The uropods consist of unarmed and nonarticulated endo- uropodites and setose exo-uropodites, each of which bears 8-10 marginal setae and 2 sub- marginal ones ventrally. There is a small ventral spine on the posterior margin of the sixth abdominal somite. The antennule consists of a long peduncle with an articulated segment terminally which will be the dorsal flagellum. Proximal to the articulation there are 2 long plumose setae in place of 1 in the previous stages and 3-5 short simple setae. Usually there is evident a simple lobe which will become the ventral flagellum and which bears a plumose seta. The antenna has 11-13 plumose setae on the scale and the endopodite (which has elongated considerably) has lost its 2 long and 1 short plumose setae and replaced them with a single terminal process which appears to be a single flexible seta. The mandibles have added teeth. The maxillule has 7 or 8 setae on the proxi- mal endite and now the tiny seta occasionally present on the proximal segment of the endo- podite in earlier stages is missing. The maxilla bears 9-11 plumose setae on the scaphognathite, the endopodite carries a total of 6 or 7 setae. The basal endites each carry 4 or 5 setae. The distal coxal endite may have 3—5, the proximal coxal endite usually has 8 or 9 setae. The first maxilliped is basically unchanged but a third seta has been added to the medial margin of the proximal segment of the endo- podite. The second maxilliped is essentially un- changed. The third maxilliped is little changed except for a slightly greater development of the endo- podal lobe. The terminal seta of this lobe is sometimes missing. Fourth Zoea There is considerable variation in setation and relative degree of development of appen- dages in the fourth zoeal instar. Some individ- I II Fig. 5. Trizopagurus magnificus. Top , the maxillule of zoeae I-IV; bottom, the maxilla of zoeae I-IV. 466 PACIFIC SCIENCE, Vol. XXI, October 1967 II III IV Fig. 6. Trizopagurus magnificus. Top , first maxilliped; center, second maxilliped; bottom, third maxilli- ped; of ( left to right) zoeae I-IV. Zoeal Stages and Glaucothoe of Trizopagurus magnifcus — Provenzano 467 uals, better developed than some of their sib- lings, were able to moult directly to the glaucothoe stage following this zoeal instar, but others, less developed, moulted into a fifth zoeal instar before the glaucothoe. In all fourth stages, however, the uropods are articulated with the sixth abdominal somite via a protopo- dite. Each exo-uropod now has a large fused spine at the posteriolateral margin and in addi- tion may have 11-13 plumose setae marginally with 1—4 submarginally on the ventral surface. The endo-uropodites usually carry 5-7 marginal plumose setae and may have 1-3 submarginal setae ventrally. The telson itself is basically Fig. 7. Trizopagurus magnificus. Details of the telson of the zoeal stages. 468 PACIFIC SCIENCE, Vol. XXI, October 1967 unchanged from the previous stage except that it is more elongate and now may have either 1 or 2 pairs of submarginal setae. The medial telson process may be replaced by a pair of articulated processes. In series where the fourth zoeal instar was followed by another zoeal stage, the appendages were less well developed and the resulting fifth stage zoea did not differ significantly from the advanced fourth stage here described. The following remarks are based on specimens which moulted directly to glaucothoe from this stage. Terminal Zoea CARAPACE length: 2. 4-2. 9 mm (9 speci- mens) total length: 5. 5-6. 2 mm (8 specimens) The antennule shows subterminal groups of aesthetascs on the dorsal flagellum, and the lobe which will become the ventral flagellum is well marked and may have a terminal seta. There are 3 or 4 large plumose setae proximal to the distal articulation of the peduncle. The antennal scale may have 13-15 plumose setae on the medial margin. The endopodite may now reach as far as the base of the terminal spine of the scale, is still terminated with a single process, but consists of at least 2 or 3 segments with one or more distinct articulations. The mandibles are still more complex and show buds of the palps. The maxillule has added 2 strong teeth on the basal endite, and usually 1 or 2 setae on the coxal endite. A naked proximal lobe is present on the scaphognathite of the maxilla and as many as 22 plumose setae may be on the margin of the scaphognathite. The proximal lobe of the coxal endite of the maxilla has also increased in setation. The first maxilliped usually carries 7, some- times 6 or 8, natatory setae on the exopodite. The proximal medial corner of the basipodite may be rather prominently produced, with the usual pair of setae often reduced to a single seta. The second maxilliped may have 7 or 8 natatory setae on the exopodite but is other- wise unchanged. The third maxilliped has 7 or 8 (rarely 6) natatory setae on the exopodite. The endopodite is very well developed, segmented, and bears a total of 1-5 setae on the terminal segments. The pereiopods are well developed buds. The pleopods are represented by unarmed buds on abdominal somites 2-5. Glaucothoe SHIELD LENGTH: 0.9 mm (3 specimens) CARAPACE length: 1.3-1 .4 mm (2 speci- mens) TOTAL LENGTH: 3. 8-4.0 mm (3 specimens) The post-zoeal stage in hermit crabs, as in all reptant decapods, is radically changed from the last zoeal stage: the long rostrum has dis- appeared, the carapace of the glaucothoe being almost the form of the juvenile, the pereiopods are free and functional, the pleopods are setose, the telson and all the cephalothoracic appen- dages have undergone radical change. The il- lustrations show how the gross external mor- phology of the glaucothoe of T. magni ficus differs from the zoeal stages which preceded it. As in all other described glaucothoes of the family Diogenidae, except that of Diogenes pugilator, there are no ocular scales at the bases of the eyestalks. In three specimens checked, the setation of the pleopods varied from 8-10 per pleopod, with no consistency in pairs or by somite. Other morphological fea- tures of particular significance are shown in the illustrations and will be discussed below. The abdomen of the glaucothoe bears a few prominent chromatophores. In lateral view there is one red chromatophore anterior to the pleopods of the fourth abdominal somite. On the fifth abdominal somite there are two lateral and three ventral red chromatophores. Each of the protopods of the uropods, attached to the sixth abdominal somite, bears one red chro- matophore. In dorsal view the fifth abdominal somite shows a pair of chromatophores on the anterior border and a pair on the posterior dorsal margin. The telson bears a pair dorsally and two pairs ventrally. Other chromatophores may be present, but only those mentioned above were noted in a brief examination of a living specimen. Diffuse orange color was seen under the eyestalks and in the region of the mouth, but the precise location of the origin of the pigment was not determined. Zoeal Stages and Glaucothoe of Trizopagurus magni ficus — Prqvenzano 469 Fig. 8. Trizopagurus magnificus . Appendages of the glaucothoe. a, Mandible; b, maxillule; c, maxilla; d, first maxilliped; e, second maxilliped; f, third maxilliped; h—k, pleopods of abdominal somites 2-5; g, the tail fan. The posterior spine on the protopodite does not show in this view of the uropods. 470 PACIFIC SCIENCE, Vol. XXI, October 1967 Fig. 9. Trizopagurus magnificus. Pereiopods of the glaucothoe, right side, a, Cheliped, medial view; b, same, lateral view; c, same, dorsal view, slightly enlarged; d, second pereiopod; e, third pereiopod; f, fourth pereiopod; g, fifth pereiopod. DISCUSSION Variability in the number of larval instars in anomuran development is now established as a widely occurring phenomenon. Species of Coenobitidae (Provenzano, 1962^), Diogenidae (Provenzano, 1962^ and unpublished data), Galatheidae (Boyd and Johnson, 1963) and Hippidae (Rees, 1959) have been shown to have a variable number of instars in larval development of single species. Nor is this flexibility in development restricted to the anomurans. Caridean shrimps (Provenzano and Dobkin, 1962; Broad, 1957), scyllaridean lob- sters (Robertson and Provenzano, unpublished), some brachyurans (Costlow, 1965; Yang and Provenzano, unpublished data), and at least one dromiid crab (Rice and Provenzano, 1966) have shown this pattern when reared in the laboratory. This phenomenon apparently re- sults from the independence of the moulting and growth processes and probably is of posi- tive adaptive significance. Although apparently there is not the uni- formity in general appearance among larvae of the family Paguridae, as was thought only a few years ago, all described larvae of that family differ in certain features from larvae of the Coenobitidae and Diogenidae. The pagurid larvae which approximate most closely the diogenid larvae are those of the genu Para- pagurus, some of which have been described by Dechance (1964). The larvae of one species (Dechance’s sp. 1, which may be P. pilosimanus Smith), like those of Trizopagurus magni ficus, have minute denticulations over at least parts of the body surface. The only diogenid larvae which have been described as having any sculpturing on the cuticle are those of Dardanus as reported by Dechance (1962), in which the cuticle was reported to have extremely small overlapping Zoeal Stages and Glaucothoe of T rizopagurus magni ficus — Provenzano 471 scales with minute spinules. Larvae of a West Indian species, D. venosus (H. Milne-Ed- wards), have scutellations only on the rostrum in advanced stages and these scales are seen only with great difficulty, even under high magnification (Provenzano, unpublished data), but in that species there are minute tubercles distributed over the carapace and abdomen. These tubercles, much resembling those of Trizopagurus larvae, are more readily apparent in the early stages and, as development pro- gresses, they become less apparent, as in T. magni ficus. With this exception, T. magni ficus is the only diogenid thus far known in which the larvae are so obviously ornamented that at least in the early stages the sculpturing is suffi- ciently obvious that it aids in identification. Perhaps this sculpturing will prove to be a generic character, but otherwise it is not possi- ble yet to designate any single feature of these larvae of T . magnifcus as being generically distinctive. Since this is the first species in the genus Trizopagurus for which the larvae have been studied, it is not possible to compare these presently described stages with congeneric lar- vae from other parts of the world nor to point out which features may be reliable as specific versus generic characters. Moreover, since T. magnifcus is the first species of Diogenidae of the tropical eastern Pacific for which a descrip- tion of the larval stages is now available, it is not possible to offer a list of characters by which larvae of this species can be separated with certainty from other diogenid larvae with which they might occur in plankton. The enlargement and fusion to telson of the fourth telson process in the third and fourth zoeal stages of T. magnifcus is found in the two species of Calcinus which have been stud- ied, in Dardanus arrosor , and in the land her- mit crab Coenobita clypeatus. None of the three species of Paguristes so far studied show any change in this process in their zoeal stages, but since there are at least 25 species of Paguristes in the West Indian faunal region alone, and probably well over a hundred world-wide, this character may show some variation. In Cliba- narius and Diogenes there is fusion of this process, but instead of enlargement there is reduction, even approaching apparent absence. Some characters, such as the spine of the antennal scale, may vary in size within a par- ticular genus (see Dechance, 1962: Fig. 3), and hence may be of little value as an indicator of genus but may be reliable as a specific char- acter. The mediodorsal spine of the fifth abdominal somite in Trizopagurus is not known to occur in larvae of Clibanarius or Dardanus, but may be characteristic of Coenobita and of Calcinus, Diogenes, and at least some species of Paguristes, while the posteriolateral spines on that somite usually are found in these latter genera and in Dardanus as well. In combination, the characters which dis- tinguish the larvae of T. magnifcus from all other described diogenid larvae are : the peculiar surface sculpturing, the trio of large spines on the fifth abdominal somite (shared with sev- eral genera, but not with Dardanus or Cliba- narius), and the absence of the posteriolateral carapace spines which apparently characterize Calcinus . In addition to Coenobita, Calcinus, Cliba- narius, and Dardanus, for which larvae have been described from other faunal regions, there are within the range of T. magnifcus other re- lated genera ( Cancellus , Ante ulus, Alio dar- danus, Isocheles, and Petrochirus ) for which no larvae have been described from any part of the world. The glaucothoe of T. magnifcus is typical of the family Diogenidae in general features. As opposed to glaucothoes of the Paguridae, those of the Diogenidae (and of the Coeno- bitidae) are generally symmetrical (the aber- rant genus Diogenes is an exception), the chelae being of subequal size, and the tail fan especially being similar in both sides. The uro- pods in Diogenidae and Coenobitidae have well developed, functional endopodites, whereas in Paguridae the endopodites are very much re- duced. This glaucothoe differs in many respects from those known from other faunal regions and it is reasonable to expect that these fea- tures will be of value in separating planktonic T. magnifcus glaucothoes from those of other hermit crabs in the eastern Pacific when the latter have been studied. With respect to described glaucothoes of non-pagurid hermit crabs, that of T. magnifcus 472 PACIFIC SCIENCE, Vol. XXI, October 1967 differs notably from that of the West Indian Coenobita clypeatus in having a well developed exopodite on the third maxilliped and in not having an extremely long terminal seta on the antennal flagellum (both the reduced exopodite and the long terminal seta probably are generic or familial characters of the land hermit crabs (Provenzano, 1962^). The glaucothoe of T. magnifcus differs from those of the Pacific Dardanus scutellatus and the West Indian D. insignis and D. venosus (Provenzano, 1963^ 1963^) in size (all of which are much larger), in eye shape (in Dardanus the cornea is wider than the eyestalk, not narrower) , in not having their peculiar armature of the ambulatory dac- tyls, in having a shorter telson, and in having a reflexed palp on the endopodite of the maxillule, which those species lack. The glau- cothoe attributed to the West Indian Petro- chirus dio genes (Provenzano, 1963^) was erroneously identified (Provenzano, in prepara- tion), but the true glaucothoe of Petrocbirus (a genus represented in the eastern Pacific by P. californiensis) probably differs from that of Trizopagurus in those same features as does Dardanus. The glaucothoe of Clibanarius erythropus from the Mediterranean (Dechance, 1958) dif- fers from that of T. magnifcus in being smaller, in having a suboval telson, in lacking a promi- nent spine on the protopodite of the uropod, in having a smaller number of segments on the antennal flagellum, and in details of setation. The four species of Paguristes for which glau- cothoes have been described, P. turgidus Stimp- son, from the northeastern Pacific (Hart, 1937); P. oculatus (Fabricius), from the Medi- terranean (Issel, 1910 and Pike and William- son, I960); P. abbreviate Dechance, from the western Indo-Pacific (Dechance, 1963); P. sericeus A. Milne-Edwards, in the West Indies (Rice and Provenzano, 1965), all differ from that of T. magnifcus in having longer dactyls on the second and third pereiopods, in armature of the chelipeds, in having a very small number of segments in the antennal flagellum (8 segments or less), and in having only 2 segments in the ventral ramus of the antennule. Glaucothoes of Calcinus — specifi- cally, C. ornatus (Roux), in the Mediterranean (Pike and Williamson, I960); and C. tibicen (Herbst), in the West Indies (Provenzano, 1962^) — apparently bear closest resemblance to that of T. magnifcus , but when the eastern Pacific glaucothoes of Calcinus have been stud- ied, probably there will be size differences and details, such as armature of the protopodite of the uropod, by which these forms may be dis- tinguished. An apparently unique feature of the glau- cothoe of T. magnifcus is the patch of granula tions on each chela. The only other diogenid genera occurring in the range of T. magnifcus, and for which no glaucothoes have been described from other regions, are Allodardanus (A. bredini Haig and Provenzano, 1965), Isocheles (several species), An i cuius (A. elegans Stimpson) and Can cell us (C. tanneri Faxon). In Allodardanus and Iso- cheles the dactyls of the second and third pereiopods are rather long, and it is likely, though not certain, that the glaucothoe will show the same condition. Both Aniculus elegans and Cancellus tanneri have very short dactyls, but neither species has a reflexed palp on the endopodite of the maxillule, and so their glaucothoes should be distinguishable from that of Trizopagurus. Particularly important characters for the fu- ture discrimination of diogenid glaucothoes should be the overall body size, the shape and armature of the telson, the armature of the protopodite of the uropods, the shape of the eyes, the relative lengths of the dactyls and propod i of pereiopods two and three and the armature of these dactyls, the length of the setae on the antenna relative to lengths of antennal segments, presence or absence of a reflexed palp on the endopodite of the maxil- lule, and the armature of the chelipeds. REFERENCES Boone, L. 1932. The littoral crustacean fauna of the Galapagos Islands. Part 2. Anomura. Zoologica 14:1-62, 19 figs. Bouvier, E. L. 1898. Sur quelques Crustaces anomoures et brachyures recueillis par M. Diguet en Basse-Californie. Bull. Mus. Hist. Nat. Paris 4:371-384. Boyd, C. M., and M. W. Johnson. 1963. Variations in the larval stages of a decapod Zoeal Stages and Glaucothoe of Trizopagurus magnifcus — Provenzano 473 crustacean, Pleuroncodes planipes Stimpson (Galatheidae) . Biol. Bull. Woods Hole 124(3) :141— 152. Broad, A. C. 1957. Larval development of Palaemonetes pugio Holthuis. Biol. Bull. Woods Hole 112(2) :l44-l6l. Coffin, H. G. I960. The ovulation, embryol- ogy and developmental stages of the hermit crab Pagurus samuelis (Stimpson). Walla Walla College Publ. Biol. Sci. 25:1-30. Costlow, J. D. 1965. Variability in larval stages of the blue crab, Callinectes sapidus. Biol. Bull. Woods Hole 128(1) :58-66. Dechance, M. 1958. Les glaucothoes de Cata- paguroides timidus (Roux) et de Clibanarius erythropus (Latreille) . Remarques sur le stade post-larvaire des Pagurides. Bull. Soc. Zool. France 83(2-3) :274-293. 1962. Remarques sur les premiers stades larvaires de plusiers especes indopacifiques du genre Dardanus. Bull. Mus. Natl. Hist. Nat, 2e ser., 34(1) :82-94. 1963. Developpement direct chez un paguride, Paguristes abbreviatus Dechance, et remarques sur le developpement des Paguristes. Bull. Mus. Natl. Hist. Nat., 2e ser., 35(5):488-495. 1964. Developpement et position sys- tematique du genre Parapagurus Smith (Crus- tacea Decapoda Paguridea), I. Description des stades larvaires. Bull. Inst. Oceanog. Monaco 64(1321) :1— 26. Forest, J. 1952. Contributions a la revision des Crustaces paguridae I. Le genre Trizo- pagurus. Mem. Mus. Natl. Hist. Nat., ser. A, Zoologie 5(1): 1-40. Haig, J., and A. J. Provenzano. 1965. A new genus and two new species of diogenid her- mit crabs (Decapoda, Anomura). Crusta- ceana 9(2) T99-207, pis. IV, V. Hart, J. F. L. 1937. Larval and adult stages of British Columbia Anomura. Can. J. Res., D, 15:179-220, pi. I. Issel, R. 1910. Ricerche intorna alia biologia ed alia morfologia dei crostacei decapodi. Parte I. Studi sui Paguridi. Arch. Zool. Ital. Napoli 4:335-397, pis. 9-1 1. Pike, R. B., and D. I. Williamson. I960. Larvae of decapod Crustacea of the families Diogenidae and Paguridae from the Bay of Naples. Pubbl. Staz. Zool. Napoli 31(3): 493-552. Provenzano, A. J. 1962a. The larval develop- ment of the tropical land hermit Coenobita clypeatus (Herbst) in the laboratory. Crus- taceana 4(3) : 207-228. 1962&. The larval development of Calcinus tibicen (Herbst) (Crustacea, Ano- mura) in the laboratory. Biol. Bull. Woods Hole 123(1) T79-202. 1963^. The glaucothoe of Dardanus venosus (H. Milne-Edwards) (Decapoda: Anomura) . Bull. Mar. Sci. Gulf and Carib. 13(1) :11— 22. 1963A The glaucothoes of Petrochirus dio genes (L.) and two species of Dardanus (Decapoda: Diogenidae). Bull. Mar. Sci. Gulf and Carib. 13(2) :242-26l. ■ and S. Dobkin. 1962. Variation among larvae of decapod Crustacea reared in the laboratory. Am. Zool. 2(3):439. and A. L. Rice. 1964. The larval stages of Pagurus marshi Benedict (Decapoda, Anomura) reared in the laboratory. Crus- taceana 7(3) :217-235. Rees, G. H. 1959. Larval development of the sand crab Emerita talpoida (Say) in the laboratory. Biol. Bull. Woods Hole 117(2): 356-370. Rice, A. L., and A. J. Provenzano. 1965. The zoeal stages and the glaucothoe of Paguristes sericeus A. Milne-Edwards (Anomura, Di- ogenidae). Crustaceana 8(3) :239-254. and A. J. Provenzano. 1966. The larval development of the West Indian sponge crab Dromidia antillensis (Decapoda, Dromiidae) based on laboratory rearings. J. Zool. 149(3) :297-319. Bathymetric Distribution of Chaetognatha, Siphonophorae, Medusae, and Ctenophorae off San Diego, California1 * Angeles Alvarino The samples studied were obtained at various depths with the bongo or bmoc open-closing paired zooplankton net (McGowan and Brown, 1966), at 30°30'N, 120°00'W, from 27 Au- gust to 8 September 1965. These collections were taken during night hours (27 August to 8 September 1965), and daylight hours (30 August to 5 September 1965) . Quantitative data for each of the spe- cies were determined for each of the nets (left and right) for collections at the various strata surveyed; and variations in the number of specimens and in the species were observed at times in the same haul for the samples from the left and right nets. Diagrams were made showing the qualitative and quantitative dis- tribution of the species in the two nets for each of the stratified hauls. However, in the final diagrams the medians for both left and right samples were plotted for each of the stratified hauls. The data from the collections made at differ- ent times and depths during the same half day (dark or daylight) were plotted together. Since the hauls for the same half of the day were made on different dates (there was a lapse of time from one haul to the other), it is ob- vious that changes might take place at times, either in the depth at which the various spe- cies were found or in the quantity of individuals obtained. In the case of the siphonophores, the necto- phores of a physonectes appeared in either the right or left net and in the other net the 1 These studies were conducted under the Marine Life Research Program, the Scripps Institution’s com- ponent of the California Cooperative Oceanic Fish- eries Investigations, and with support from the Na- tional Science Foundation (NSF GB-2861), and AEC Contract AT(ll-l)-34, Project 127. Contribution from Scripps Institution of Oceanography, University of California, San Diego, California. Manuscript re- ceived July 21, 1966. pneumatophores with the nectosoma and siphosoma attached. Similarly, in other cases, one of the paired nets contained the bract or the superior nectophore and the other net the gonophores or the inferior nectophore of the eudoxid or paragastric forms of the Diphyidae. The quantitative distribution of Chaetognatha and Medusae are noted in the diagrams; but in the case of the Siphonophorae, although data on the number of superior and inferior nectophores, gonophores, bracts, pneumato- phores, etc. were obtained, because of the peculiar anatomical structure of these orga- nisms it was not convenient to attempt a quantitative representation. In the case of the Monophyidae and Diphyidae, for instance, it will be easy to establish the number of indi- viduals present from the part obtained in the sample, but this is not so for the Physophorae and Hippopodiidae. Therefore, only the quali- tative distribution of the siphonophores was considered in preparing the final diagrams. During each tow about 10,000 m3 of water were filtered through the net; thus the sam- ples were directly and quantitatively compar- able. Collections were made at the following depths (in meters) : NIGHT DAYLIGHT 100-10 110-0 100-20 350-250 300-235 460-0 460-410 525-401 500-420 1720-1340 620-530 2300-1880 775-685 2630-2210 840-690 3040-2620 1030-860 1040-890 1242-1090 1710-1450 2020-1800 2170-1950 2320-2100 2460-2100 474 Bathymetric Distribution — Alvarino 475 Several gaps in the bathymetric distribution are obvious, since collections were not obtained at some strata. The most important sampling gap was in the daylight series, of about 800 m (1340-525 m), which interrupts the data on the bathymetric distribution. This lack of data, and the one haul from 460 m to the surface during the daylight series (considered only for the siphonophores ) , do not permit the loca- tion of the upper or lower limits of the dis- tribution of several species. Quantitative data were obtained by the method explained by Alvarino (1965 c and 1966^) . The bathymetric zones considered are: epi- planktonic (upper 200 m), mesoplanktonic (200-1000 m), bathyplanktonic (below 1000 m). The vertical division into zones cannot be static, however, because the stratification of the organisms is controlled by bio-physico- chemical factors. CHAETOGNATHA Figures 1 and 2 show the quantity of each of the species found and their distribution for the day and night series, respectively. The epiplanktonic species observed here were: Krohnitta subtilis, Pterosagitta draco, Sagitta bierii, S. blpunctata, S. enflata, S. euneritica, S. hexaptera, S. minima, S. pad pc a, S. pseudosenatodentata, and S. scrlppsae. A typical mesoplanktonic species, S. decipiens, also was present in the upper 100 m, but ex- tended to 620 m depth. Other species characteristic of the meso- planktonic levels which extended their distribu- tion into the bathypelagic domain were: S. maxima, S. macrocephala , and S. zetesios. The bathyplanktonic species reaching various levels of the mesoplanktonic zone were: Eukrohnia bathypelaglca, E. fowled, and E. hamata. During the nighttime collections, K. sub- tills and S. scrlppsae did not appear in the upper 100 m, but did appear at 500-235 rn and 460-235 m, respectively. S. pad flea was absent from any level. The species present in the upper 100 m layers for both night and daylight series, S. bierii, S. declplens, S. euneritica, and S. pseudo- senatodentata, were more abundant at night than during daylight, whereas S. minima was more abundant during the day, and the others appeared within the same range of abundance for both periods. The species spreading from the surface to 300 m depth during daylight were S. euneritica and S. padflca, and at night, S. enflata and S . hexaptera , the latter appearing down to 525 m in the daytime. Sagitta bierii populated the upper 100 m down to 500 m during both day and night, showing the greatest concentration in the up- per 100 m and the lowest between 400 and 500 m at night, whereas during daylight the distribution was homogeneous along the layers it populated; but S. hexaptera presented the highest concentration from 300 to 235 m at night, and was homogeneous at daylight down to 525 m depth. K. subtilis extended during daylight from the surface to 525 m, with higher concentrations in the upper 100 m, whereas at night it was present only between 235 and 500 m. The presence of S. declplens in the upper 100 m both at night and by day appears to indicate that upwelling phenomena took place at this location. Two specimens of S. declplens were observed in the left net for the tow from 2630 to 2210 m, and two specimens in the right net from 3040 to 2620 m during the day series. These showed evidence of con- tamination, however, and were omitted from the figures. S. scrlppsae extended during daylight from the surface to about 500 m, with maximum concentration at 350 to 250 m. At night it extended only from 235 to 460 m, with maxi- mum concentration between 300 and 235 m. The influence of light in the bathymetric distribution could be understood when observ- ing the vertical distribution of S. bierii, S. declplens, S. euneritica, and S. pseudoserrato- dentata . However, if the factor of light is re- sponsible for the vertical migration of these organisms, it fails to explain the distribution of K. subtilis, S. enflata, S. maxima, and S. scrlppsae . Therefore, the factors influencing the verti- cal distribution and displacements of the spe- cies of chaetognaths may be of various kinds, and interacting: light, temperature, oxygen, 476 PACIFIC SCIENCE, Vol. XXI, October 1967 1000 — - — 1000 1500 — — 1500 2000 — — 2000 ESTIMATED NUMBER PER 10,000m3 WATER I 1-49 II 50 - 499 III 500 - 4999 llll >5000 — 2500 3000 — — 3000 Fig. 1. Bathymetric distribution of Chaetognatha during the daylight series, Bathymetric Distribution — Alvarino 0 m — - II III - 1,1000 — 1500 — m~ o o II o CD o CO co O co CD CD fvj O CT> O CO 477 0 m I — 500 a A V V A w —1000 — 1500 2000 — ESTI MATED NUMBER PER 10,000m3 WATER I 1-49 I! 50 - 499 III 500 - 4999 1111 >5000 — 2000 Fig. 2. Bathymetric distribution of Chaetognatha during the night series. 478 PACIFIC SCIENCE, Vol. XXI, October 1967 food, hydrostatic pressure, and bio-physico- chemical interrelations with other organisms of the pelagic realm. Also to be considered is the structure of the population, since young individuals appeared to be at levels closer to the surface than were adults (Alvarino 1964a, 1965*). SIPHONOPHORAE A greater number of species was observed in the upper 100 m in the daylight series than in the night (Figs. 3 and 4). The species ob- served in the upper 100 m for both series were Muggiaea atlantica, Chelophyes appendiculata, and Eudoxoides spiralis. These were joined in daylight by Eudoxia russelli, Lens id hotspur , L. subtiloides, Amphicaryon acaule, and Steph- anomia bijuga, and only by A. ernes ti, L. multi- cristata (extending also to the mesoplanktonic levels), and Nectodroma reticulata at night. The last two inhabited the mesoplanktonic domain in daylight. The daylight haul from 460 to 0 m was in- cluded in the diagrams for the siphonophores only, to show the presence of several species at those levels. However, the bathymetric limits cannot be determined for some species; thus, the upper limit of the daylight distribu- tion for Chuniphyes moserae, Ch. problematica, and Heteropyramis maculata , and both upper and lower limits for L. challengeri and Necto- droma dubia, are not yet established. The mesoplanktonic species appeared in higher numbers at night than in daylight. Typical mesoplanktonic species such as L. ajax, L. conoidea, L. grimaldii, Bargmannia elon- gata, Stephanomia rubra, Physophora hydro- statica, and N. dubia were not observed at night; and Nectopyramis diomedeae, N. thetis, and N. natans were not observed in daylight. Species appearing in the mesoplanktonic levels at night which extend to deeper layers at day- light were L. achilles, Vogtia kuruae, and Rosacea p lie at a. Dimophyes arctica was ob- tained in the mesoplanktonic levels at night and only at the bathyplanktonic zone during daylight. Species occupying both meso- and bathy- pelagic regions were Ch. multidentata, Ch. moserae, Ch. problematica, Clausophyes galeata, Crystallophyes amigdalina, Heteropyramis maculata, L. achilles, L. hostile, L. lelouveteau, L. reticulata, and Nectodroma reticulata (in daylight) . Species observed at the bathyplanktonic levels only were Clausophyes ovata, Ceratocymba dentata, and L. havock. One specimen of Velella (longest axis 70 mm, and sail oriented NW-SE) was obtained with a dip net, 8 September 1965. One complete colony of Physophora hydro- statica was obtained in daylight at depths of 350-250 m. MEDUSAE Higher numbers of species were observed during the night series than in the daytime (Figs. 5 and 6). Liriope tetraphylla was the only species found for both series in the upper 100 m, presenting a higher number during daylight. Sibogita geometric a and Cunina pere- grina were observed only at night and in the upper 100 m. Phialidium discoidum and Cros- sota alba were observed during daylight in these upper strata, and extended from 685 to 1030 m at night. The mesoplanktonic species were Sarsia coc- cometra, Zanclea costata, Pandea violacea, Heterotiara anonima, Colobonema sericeum, Cross ota alba, C. brunnea (extending deeper at night), and C. pedunculata. Species populating both the meso- and bathypelagic levels were Halicreas papillosum, Atolla wyvillei, and Periphylla hyacinthina (ob- served only at night). Medusae observed exclusively at the deepest levels were Homeonema alba, Aegina citrea, and Nausithoe rubra. The medusae showed some degree of stratifi- cation related to the size of the specimens; thus individuals of A. wyvillei 12-30 mm in diameter appeared at 300-235 m, whereas specimens 75 mm in diameter were found at 1710-1450 m. However, specimens up to 100 mm in diameter were obtained in the 620-530 m level. The size-stratifications for this species appeared to be more clear-cut during the day- light series, when individuals 7-20 mm in diameter appeared in the 460-410 m level, and those 30-60 mm in diameter at 1720-1340 m. Bathymetric Distribution — Alvarino 479 1000- — — 1000 1500 — — 1500 — 2000 — 2500 — 3000 2000 — 2500- 3000 — Fig. 3 Bathymetric distribution of Siphonophorae during the daylight series. 480 PACIFIC SCIENCE, Vol. XXL October 196: — 0 m — 500 — 1000 — 1500 2000 2000 Fig. 4. Bathymetric distribution of Siphonophorae during the night series. Bathymetric Distribution — Alvarino 481 0 m — o a> E o o o o o CO m — 0 m 500 — — 500 1000 — — 1000 1500 — 1500 2000- — — 2000 ESTIMATED NUMBER PER 10,000m3 WATER I 1-49 II 50 - 499 III 500 - 4999 Mil >5000 — 2500 3000 — - — 3000 Fig. 5. Bathymetric distribution of Medusae during the daylight series, 482 PACIFIC SCIENCE, Vol XXI, October 1967 CD — CD O 0 m- O O - — CO 0 m 500 — 1000 — — 500 — 1000! 1500 — — 1500 ESTIMATED NUMBER PER 10,000m3 WATER I 1-49 II 50 - 499 III 500 - 4999 mi >5000 — 2000 Fig. 6. Bathymetric distribution of Medusae during the night series, Bathymetric Distribution — AlvarinO 483 Halicreas papillosum 30 mm in diameter were present at 775-685 m, and those 50 mm in diameter at 1242-1090 m. Specimens of Peri- pbylla hyacinthina up to 25 mm high appeared at 300-235 m, whereas at 840-690 m they were 35 mm high. Colobonema sericeum up to 40 mm in height were found at 460-410 m, and below this level those 50 mm in height. CTENOPHORAE Beroe spp. extended in the night series from 10 m (uppermost sampling) to 500 m, and during daylight from 0 to 525 m. CONCLUSIONS ON BATHYMETRIC DISTRIBUTION Several striking features were observed in the bathymetric distribution : 1. The number of species of Chaetognatha and Siphonophorae in the upper 100 m was higher during daylight than at night. 2. The number of specimens for the species of Chaetognatha present both in daylight and at night in the upper 100 m was either of the same numerical magnitude or, in most cases, higher at night. 3. In general, the difference in the number of specimens observed in the right and left net for the upper 100 m was greater during day- light than at night. Points 1 and 3 suggest either that during daylight the patches of specimens are denser, or that at night the individuals are scattered throughout a bigger region, thus providing fewer individuals per cubic unit of water. This conclusion appears to be in contradiction with established statements, which maintain that planktonic organisms congregate more at night than in daylight. Another possibility is that they can avoid the net better in daylight than at night. The sky condition at night, when the hauls were taken, was one of darkness, with- out moonlight; but there are no data on bio- luminiscence. Therefore, these preliminary studies appear to indicate that individuals are not evenly dis- tributed, but that there is a small pattern of patchiness included in the total region popu- lated by certain species. By using the paired net it will be possible to detect either this patchiness within the distributional region of the species, or the flocking of individuals when disturbed (a general behavior response observed in nature). DISTRIBUTION OF THE ORGANISMS AND THE POSITION OF THE SCATTERING LAYER Unfortunately, samplings were not made at depths at which the scattering records appeared. However, they could be determined easily by correlating records and samples taken at the same time and date. For example, at 1238- TABLE i Species in the Upper no m Correlated with the Shallower Scattering (Daylight) CONCENTRATION OF INDIVIDUALS PER 10,000 M3 OF WATER FILTERED group GREATER THAN 5,000 4999-500 499-50 LESS THAN 50 Chaetognatha S. bier it S. minima S. pseudoserratodentaia K. subtilis S. bipunctata S. dec ip 'tens S. enflata S. euneritica P. draco S. hexaptera S. pacifica S. scrippsae Siphonophorae M. atlantica E. russelli Ch. appendiculata N. reticulata E. spiralis L. subtiloides A. acaule St. bijuga Medusae L. tetraphylla P. discoidum C. alba Ctenophorae Beroe spp. 484 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE 2 Species in the 350-250 m Layer, Partially Coincident with the Deepest Scattering Record (Daylight) CONCENTRATION OF INDIVIDUALS PER 10,000 M3 OF WATER FILTERED GROUP GREATER THAN 5,000 4999-500 499-50 LESS THAN 50 Chaetognatha S. bierii S. decipiens S. scrippsae S. maxima K. subtilis S. bipunctata S. euneritica S. hexaptera S. minima S. pacific a S. zetesios Siphonophorae Ch. appendiculata L. multi crist at a L. conoidea N. dubia St. rubra Ph. hydrostatica B. elongata Medusae S. coccometra L. tetraphylla C. alba C. brunnea 1550 hours on 5 September 1965, bright and depth of 218.40 m it was 36.40 m thick. The at 91 m, 200 m, and 345 m. The species ob- served coincidently are detailed in Tables 1 and 2. The nighttime scattering layer at 2038-2345 hours on 3 September 1965 appeared to be 54.60 m thick at the upper levels, and at a species observed coincidently are detailed in Tables 3 and 4. The siphonophores considered to be most probably responsible for the production of scattering are the Physonectae, those with floats containing gas (CO). Species of that group ( Stephanomia bijuga, St. rubra, and Physo- TABLE 3 Species in the 100-20 m Layer, Partially Coincident with the Shallower Scattering Record (Night) CONCENTRATION OF INDIVIDUALS PER 10,000 M3 OF WATER FILTERED GREATER GROUP THAN 5,000 4999-500 499-50 LESS THAN 50 S. bierii S. euneritica P. draco S. hexaptera S. enfaia Chaetognatha S. minima S. pseudoserra- todentata M. atlantica A. ernesti Siphonophorae E. spiralis Ch. appendiculata N. reticulata Medusae L. tetraphylla S. geometrica C. peregrin a Ctenophorae Beroe spp. Bathymetric Distribution — Alvarino 485 TABLE 4 Species in the 300-235 M Layer, Coincident with the Deepest Scattering Record (Night) CONCENTRATION OF INDIVIDUALS PER 10,000 M3 OF WATER filtered GROUP GREATER THAN 5,000 4999-500 499-50 less than 50 Chaetognatha S. decipiens S. bierii S. hexaptera S. scrippsae K. subtilis S. enflata S. maxima 5“. zetesios Siphonophorae M. atlantica L. conoidea Ch. appendiculata L. achilles L. havoc k L. multicristata St. bijuga B. elongata A. ernes ti R. plicata Medusae A. wyvillei P. hyacinthina Ctenophorae Beroe spp. phora hydrostatica) were observed at the levels of scattering. It is also probable that the head armature (hooks, teeth, and chitinous plates) of the chaetognaths will contribute to scattering. REFERENCES Alvarino, A. 1964^. Bathymetric distribution of chaetognaths. Pacific Sci. 18(l):64-82. 1964 A Zoogeografia de los Queto- gnatos, especialmente de la region de Cali- fornia. Ciencia 23(2):51-74. 1965^. Chaetognaths. Oceanography and Marine Biology: Annual Review 3:115— 194. 1965 A El Mar de Cortes: Quetognatos, Sifonoforos y Medusas. Proc. Second Mex- ican Congress of Oceanography, 1965. 1965c. Distributional Atlas of Chaeto- gnatha in the California Current Region. CalcOFi Atlas 3, pp. 1-291. 1966^. Zoogeografia de California: Quetognatos. Bob Soc. Mex. Hist. Nat. (In press.) 1966A A new Siphonophora, Vogtia kuruae, n. sp. Pacific Sci. 21(2): 236-240. McGowan, J. A., and D. M. Brown. 1966. A new opening-closing paired zooplankton net. Univ. Calif. Scripps Inst. Ocean. 66-23, pp. 1-50. Moore, H. B., and E. G. Corbin. 1956. The effects of temperature, illumination, and pressure on the vertical distribution of zoo- plankton. Bull. Mar. Sci. Gulf and Caribb. 6(4) :273-287. Reid, J. L. 1965. Intermediate waters of the Pacific Ocean. Johns Hopkins Oceanog. Studies (2): 7-8 5. Ecological Significance of a Drifting Object to Pelagic Fishes Reginald M. Gooding and John J. Magnuson1 Pelagic fishes frequently gather around drift- ing material in the open sea. Commercial and sport fishermen regard the immediate vicinity of drifting material as a potentially good area for trolling. Commercial seine and pole-and- line fishermen in Japan, Indonesia, and Malta anchor floating material to attract fish. Fish have been reported gathered around floating algae, coconuts, and pumice (Besednov, I960; Senta, 1965); floating logs (Inoue, Amano, and Iwa- saki, 1963; Kimura, 1954; Yabe and Mori, 1950) ; coconut fronds and slabs of cork (Har- denberg, 1949; Soemarto, I960; Galea, 1961); and rafts (Kojima, I960; Heyerdahl, 1950; Evans, 1955). In addition to clustering near these inanimate objects, the young of many pelagic fishes gather beneath jellyfish (Man- sueti, 1963); fish- jellyfish associations have much in common with the associations studied in the present paper. Hypotheses suggested to explain the accumu- lation of fish around inanimate floating objects include: (1) fish seek shelter from predators (Soemarto, I960; Suyehiro, 1952); (2) larger fish prey on the concentration of smaller fish (Kojima, 1956) ; (3) fish feed on algae or decaying coconut fronds (Reuter, 1938; Soe- marto, I960); (4) fish seek the shade under the object (Suyehiro, 1952); (5) fish use float- ing objects as a substrate on which to lay their eggs (Besednov, I960); (6) the shadow of the object makes zooplankton more visible to the fish (Damant, 1921). At the beginning of the present study we suggested still another hypoth- esis: floating objects are cleaning stations, where pelagic fishes have their parasites removed by other fish. Such symbiotic cleaning associations are well documented for fishes in inshore waters (Eibl-Eibesfeldt, 1955; Limbaugh, 1955, 1961; Randall, 1958). To test these hypotheses, studies were made 1 Bureau of Commercial Fisheries Biological Lab- oratory, Honolulu, Hawaii. Manuscript received August 19, 1966. from a raft with an observation chamber (Fig. 1) built at the Bureau of Commercial Fisheries Biological Laboratory, Honolulu, and set adrift in the central Pacific (Gooding, 1965). The present paper describes and interprets the ob- servations in light of the above hypotheses. areas and methods of observation Observations were made in two areas, one off the leeward coast of the island of Hawaii and the other near the Equator in the central Pacific (Fig. 2). Observations were made in Hawaii between September 28 and October 11, 1962, and be- tween August 1 and August 26, 1965. This area offers two advantages: first, it is sheltered from the northeast trade winds and the sea is relatively calm; second, essentially pelagic con- ditions (water deeper than 800 m) occur within 1 mile of shore. During 345 hours of drift, 173 hours of daylight observations and 9 hours of night observations were recorded. Eleven drifts were made, the longest of which was '52 hours. Two drifts were made between February 14 and March 20, 1964 in the storm-free belt at the convergence of the northeast and south- east trade winds near the Equator. On the first drift the raft was launched 9 nautical miles north of the Equator in an area of upwelling. During 194 hours of drift, 91 hours of daylight observations were made. The second equatorial drift began 153 nautical miles south of the Equator. During 215 hours of drift, 100 hours of daylight observations were made. The raft drifted 585 nautical miles west dur- ing the first equatorial drift and 395 nautical miles west during the second. Most of the drift was due to surface currents. To reduce wind- induced drift a 28-foot parachute was used as a sea anchor during part of the first drift and all of the second. (It was also used during several of the Hawaiian drifts.) While the raft was adrift, wave heights ranged from 0 to 1 m at Hawaii and from 1 to 486 Drifting Object and Pelagic Fish — Gooding and Magnuson 487 Fig. 1. The observation raft used in study. 2 m at the Equator. Average wind speeds ranged from 10 to 15 knots. Cloud cover seldom exceeded 30%. The observation chamber beneath the raft (Fig. 3) accommodated a single observer, who could view the area beneath and around the raft. Two observers manned the drifting raft from dawn to dusk. Watch positions in the chamber were rotated each hour. Nights were spent on the ship, which remained 1-3 miles from the raft. A skiff provided transportation between ship and raft. The observers noted the number of each kind of fish at the raft, their position under or near the raft, and their reaction to the raft and to other fish or invertebrates. Night observations were made under bright moonlight, but a flash- light was used at intervals to determine more accurately the positions of the fish. The ac- cumulation was quantified by making population counts of the species present at intervals during the day. An estimate of population changes during the night was obtained by comparing the last count in the evening with the first count on the following morning. 158° 157° 156° 155°W n \ \ i P j: o I's \ j r •FANNING I. j -1 i ^ CHRISTMAS I. j / „ DRIFT NO.l (8 DAYS) . L ! Z. ; j * 9 (QDAvfSj — „ 1 DRUM j 155° 150° W Fig. 2. Areas in which drifts were made with the observation raft, off Hawaii ( upper panel) and near the Equator ( lower panel). Fig. 3. The observation chamber of the raft. Dark specks to right of chamber are small fish. The white object behind the chamber is the parachute drogue. 488 In addition to direct observations, 6,200 ft of 1 6-mm color movies, and numerous still pic- tures were taken. Fish were captured at the raft with dip nets, baited hooks on hand lines, casting and trolling lures, and a small purse seine net attached to the sides of the raft. To avoid interference with the accumulation of animals, collections were made only at the end of the drifts. Stomach PACIFIC SCIENCE, Vol. XXI, October 1967 contents and external parasites of fish captured at the raft were preserved. FISHES AT THE RAFT Animals seen from the observation chamber (some are shown, as photographed from the chamber, in Figures 4a-f) were broadly Fig. Ad. Juvenile dolphin. Fig. Ab. Adult dolphin. Fig. Ae. Whitetip shark accompanied by pilotfish and remora. Fig. Ac. Amberjack. Fig. Af. Whale shark accompanied by remora. Drifting Object and Pelagic Fish — Gooding and Magnuson 489 grouped as transients, visitors, or residents (Table 1) on the basis of their reaction to the raft and the length of time they remained near it. Transients (many of which were flyingfish, Exocoetidae) did not appear to react to the raft, but were usually visible only momentarily as they swam by. Visitors did not aggregate at the raft, but appeared to react to it; they usually remained near it for several minutes to an hour. Residents aggregated at the raft; some stayed in view more or less permanently, and others swam out of view for several hours but usually returned. Different individuals of certain species did not always react in the same way to the raft; these species were consequently placed in more than one category. Residents were of two types: smaller fishes which stayed in the immediate vicinity of the raft and were usually in view of the observer; and iarge carnivores that were frequently out of view for several hours. When reappearing after a prolonged absence, the individual or group could often be identified by distinguishing characteristics such as abrasions, parasites, scars, the number in the group, and body size. The relation of all resident species to the raft was facultative, since each also occurs independently of any association with drifting objects. Small resident fishes were: freckled drift- fish, Psenes cyanophrys (Cuvier) ; juvenile pilot- fish, Naucrates duct or (Linnaeus); rough trig- gerfish, Canthidermis maculatus (Bloch) ; scrawled filefish, Alutera script a (Osbeck) (but only individuals exceeding about 20 cm, smaller ones behaving as visitors); amber jack, Seriola rivoliana Cuvier and Valenciennes; juvenile greater amber jack, Seriola dumerili (Risso) ; juvenile jack, Caranx sp.; adult and juvenile mackerel scad, Decapterus pinnulatus (Eydoux and Souleyet) ; juvenile skipjack tuna, Katsu- wonus pelamis (Linnaeus); juvenile yellowfin tuna, Thunnus albacares (Bonnaterre) ; juvenile dolphin, Coryphaena sp.; and juvenile stages of four reef fishes — damselfish, Abudefduf abdominalis (Quoy and Gaimard) ; sea chub, Kyphosus cinerascens (Forskal); goatfish, Mul- loidichthys samoensis Gunther; and squirrel- fish, Holocentridae. The large predatory residents were: dolphin, Coryphaena hippurus Linnaeus; wahoo, Acan- thocybium solandri (Cuvier) ; rainbow runner, Elagatis bipinnulatus (Quoy and Gaimard) and whitetip shark, Carcharhinus longimanus, usu- ally accompanied by adult pilotfish and remoras, Remora remora (Linnaeus). The freckled driftfish was by far the most common resident in both drift areas. On all drifts it was the first to appear, had the highest rate of accumulation (Table 2), and attained the largest population. At the end of the second equatorial drift, 729 were caught in the purse seine and several hundred escaped. Many were also caught at the end of other drifts. Freckled driftfish usually came to the raft singly or in small groups. Once a green turtle, Chelonia mydas, came to the raft accompanied by nine driftfish and one remora. The turtle left with the remora after a few minutes, but the drift- fish remained with the raft. Residents accumulated more rapidly by day than by night. Statistics on the average rate of accumulation of some of the more common residents appear in Table 2. Less common resi- dents, not listed in Table 2, also accumulated more rapidly by day than by night. Species composition differed between the Hawaiian and equatorial areas. Only 38% of the 27 fish identified to species in Table 1 were seen in both areas. Three of the more common species off Hawaii, the rough triggerfish, dol- phin, and damselfish, were either absent or rare in the equatorial waters. Of species that were residents at some stage in their life history, 62% were common to both areas, whereas none listed only as a visitor was common to both areas. Some of the apparent differences between the areas could have resulted from differences in the time of year or could even be attributable to the sample sizes. For example, the occurrence of rainbow runners, pompano dolphin ( Cory- phaena equiselis Linnaeus), and green turtles in the equatorial but not the Hawaiian area may well be irrelevant, for all are common in Hawaiian waters. ADAPTIVE SIGNIFICANCE Our observations provided relevant informa- tion on the hypotheses that floating material (1) provides protection from predators, (2) concentrates the food supply, and (3) acts as a cleaning station. These hypotheses, of course, 490 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE l Animals Seen from the Observation Chamber of a Drifting Raft* SPECIES, GENUS, OR FAMILY (Common Name in Parentheses) DRIFT LOCATION BEHAVIOR CATEGORY FORK LENGTH (cm) MAXIMUM NUMBER SEEN AT ONE TIME Abudefduf abdominalis (damselfish) H R 0.7-1 .0' 24 Acanthocybium solandri (wahoo) H03 R 45-90 3 Alutera scripta (scrawled filefish ) H RV 10-35 2 Canthidermis maculatus (rough triggerfish) H R 2 5 — 3 5f -33 Caranx kalla (golden jack) H V 30 1 Caranx sp. (jack) H R 2.9-5. 3f 3 Carcharhinus longimanus (whitetip shark) HO 3 RV 125-175 2 Chelonia mydas (green turtle) 0 V 60 1 Coryphaena equiselis (pompano dolphin) 03 V 30 100+ Coryphaena hippurus (dolphin) H03 R 60-100* 70+ Coryphaena sp. H03 R 10-15 80 Decapterus pinnulatus adult (mackerel scad) H03 RT 20-25 1,000+ juvenile 3 R 13.1* 1 Diodontidae (spiny puffer) 0 V 12 1 Echeneidae (free-swimming) (remora) 3 R 8 1 Elagaiis bipinnulatus (rainbow runner) 3 R 75 1 Exocoetidae ( flyingfish ) H03 T 10-15 10+ Fistularia petimba (cornetfish) H V 20-40 2 Globicephala scammoni (pilot whale) HO V 375 2 Holocentridae (squirrelfish) H R 2 1 Istiophoridae (marlin) H T 125 1 Katsuwonus pelamis adult (skipjack tuna) H3 T 45 1,000+ juvenile 3 RV 10-15 50 Kyphosus cinerascens (sea chub) H R 2.5* 13 Manta aljredi (manta ray) H V 100-125* 1 Manta sp. 0 V 1 Drifting Object and Pelagic Fish — Gooding and Magnuson 491 TABLE 1 ( continued ) SPECIES, GENUS, OR FAMILY (Common Name in Parentheses) DRIFT LOCATION BEHAVIOR CATEGORY FORK LENGTH (cm) MAXIMUM NUMBER SEEN AT ONE TIME Mulloidichthys samoensis (goatfish) H RV 10-12 1,000+ Naucrates ductor adult (pilotfish) HO 3 RV 15-30 7 juvenile H03 R 2.6-6.7f 7 Nomeus gronowi (man-of-war fish) 0 V 2 1 Prionace glauca (great blue shark) 0 V 150 1 Psenes cyanophrys (freckled driftfish) H03 R 1.5-1 2. 4f 1,000+ Remora remora (attached) (remora) H03 RV 15-30 Rh/ncodon typus (whale shark) 3 V 300 1 Seriola rivoliana% (amber jack) H R 20* 1 Seriola dumerili (greater amberjack) H R 3.7 1 Sphyraena barracuda (great barracuda) H V 50 1 Thunnus albacares (yellowfin tuna) H3 RV 25-40 37 Tur slops sp. HO V 150-200 20+ (bottlenose dolphin) * Drift Location: H -- Hawaii; 0 r= 0° Latitude; 3 = 3° S. Behavior Category: R = Resident; V = Visitor; T rr Transient, t Measured length; all other lengths are estimated, i Breadth. § The first record for Hawaiian waters, identified by Dr. Frank J. Mather, Woods Hole Oceanographic Institution, from a specimen preserved after capture at the raft. TABLE 2 Average Net Increase or Decrease in Number of Residents* at the Raft per 12-Hour Day and 12-Hour Night1 in Three Drift Areas (Number of 12-Hour Periods in Parentheses) FISH HAWAII OCTOBER 1962 HAWAII AUGUST 1965 0° LATITUDE FEBRUARY 1964 3° MARCH s 1964 Day (9) Night (7) Day (8.5) Night (4) Day (7-5) Night (9) Day (8.5) Night (9.0) Psenes cyanophrys * 24 1 107 1 18 1 100 0 Coryphaena sp. (juvenile) - - - - 11 — 2 1 —1 Canthidermis maculatus 7 2 3 1 - - - - Coryphaena hippurus (adult) 4 1 10 0 - - - - Abudejduf abdominalis (juvenile) 4 — 1 - - - - - - Decapterus pinnulatus (adult) ■M - 10 — 5 2 0 3 -3 Katsuwonus pelamis (juvenile) - - - - - 3 -3 Naucrates ducior (juvenile) - - - - 1 — 1 - - * Only the residents with an average accumulation equal to or greater than one fish per 12 hours of daylight are in- eluded. t Population changes during the night were estimated by comparing the last count in the evening with the first count the following morning. t Increases are based on the rate for the first 100 to gather because larger numbers could not be counted accurately. 492 PACIFIC SCIENCE, Vol. XXI, October 1967 are not mutually exclusive. The observations provided less information about the other hypotheses mentioned earlier. All the above hypotheses consider the adaptive significance of floating material in the ecology of pelagic fishes. The stimuli that release the approach of fishes to the raft are not discussed. Protection from Predation At least nine species of fish, both large and small, reacted to the raft in a way that made them less vulnerable to predation. Typically, when a predator approached the raft, the prey formed a compact group very close to the under- structure. When the predator left or ceased harassments, the prey again dispersed about the raft. Often the predator chased the prey to the raft. The value of the raft to the prey was demonstrated by the fact that only one species, the amberjack, frequently caught fishes that had taken shelter under the raft. Observa- tions on individual prey species are described below. The most common resident, the freckled driftfish, usually took a position far below and downwind from the raft and was sometimes out of view. Driftfish were able to match their background. They had a silvery countershaded coloration when not under the raft, but took on a mottled brown coloration when close un- der it, and those collected from under an orange drogue buoy had an orange color. Most of their predator-avoidance activity was in response to dolphins, although some was in response to pompano dolphins, wahoos, bottlenose dolphins ( Tursiops sp.), or to pilotfish which approached the raft swimming with a whitetip shark. The hundreds of such responses followed an un- varying sequence: when one of the predators came into the vicinity, the freckled driftfish suddenly formed a compact school and swam rapidly back to the raft or the parachute drogue. (They also fled to the raft when an observer entered the water.) When an amberjack was preying upon them, they remained within about 20 cm of the viewing chamber. They attempted to stay on the opposite side of the chamber from the amberjack or dodged into the gaps between the frames of the viewing windows. When the amberjack was not actively feeding, the driftfish ranged out again. Small damselfish, pilotfish, greater amber jacks, and jacks behaved similarly to driftfish in response to predation, but did not change coloration. Rough triggerfish ranged far from the raft, sometimes out of sight. Their rapid return to it usually heralded the appearance of a predator (billfish, a great barracuda, bottlenose dolphin, whitetip shark) or apparent predators (schools of mackerel scad or a powerboat). They re- sumed ranging before the potential predator de- parted, except when the predator was a bottle- nose dolphin. None of the above species exhibited a predatory response towards rough triggerfish. The triggerfish did not return to the raft when manta rays appeared and they usually swam out and met approaching dolphins. Rough triggerfish and dolphins may often be associated in the absence of drifting material; sometimes they arrived simultaneously at the raft. On several occasions, the most successful piscivore, the amberjack, itself became the po- tential prey of dolphins and took shelter beneath the raft. Although amber jacks fre- quently ranged 10 to 15 m from the raft un- molested, when the dolphin began pursuit the amberjack eluded the predator by swimming close to the chamber. It remained there for some time before ranging out again. The dolphin, one of the largest residents, took shelter close under the raft three times: once in response to a bottlenose dolphin, once to a billfish, and once to a swimmer. Each time the dolphin swam around the chamber just un- der the flotation drums and took on a coloration (Fig. 5) that occurred in no other situation Fig. 5. The lower dolphin assumed the dark coloration when one of the observers entered the water. Drifting Object and Pelagic Fish — Gooding and Magnuson 493 and had not previously been recorded (for other colorations of this species, see Murchison and Magnuson, 1966). The dorsal half of the body turned a dark brownish-black. A sharp separa- tion extended longitudinally along the side between the dark dorsal area and the silvery ventral half of the body. The above behavior and coloration, observed only when 1 or 2 dolphins were at the raft, were different from those seen on similar occasions when 13 or more dolphins were present. Then the group of dolphins swam immediately behind a bill- fish, a whitetip shark, bottlenose dolphin, and a swimmer near the raft. A position im- mediately behind a potential predator may be of advantage to the prey provided the animal has the speed and maneuverability to maintain such a position. Large schools of goatfish attempted to avoid dolphins and amber jacks by swimming to the other side of the raft, but only rarely did in- dividuals use the maximum shelter of the raft by swimming under it. As a consequence both predators were able to prey upon them success- fully. One of the most clearcut examples of preda- tor avoidance occurred when a golden jack was chased to the raft by the feeding attacks of five dolphins. The dolphins stopped their feeding passes after the jack swam under the raft. For several hours the jack swam within inches of the chamber. The observer on deck could reach into the water and touch the fish without driving it away. After several hours it began to swim under the flotation drums, but not away from the raft. About 8 hours after it arrived the jack joined a whitetip shark and six pilotfish which swam close by, and left the raft in their com- pany. The dolphins took on their feeding colora- tion, but did not attack the jack as it swam off with the shark. This incident provided evidence for the protective role that both floating objects and large animals such as sharks play for the fish that accompany them. Concentration of Food Supply It has often been said that floating material concentrates the food supply — smaller fish, zoo- plankton, or sessile biota. Most piscivores did net successfully prey on fish that sought shelter beneath the raft, but they did prey extensively on those that gathered at the raft but did not take shelter beneath it. Zooplankton was not concentrated at the raft, nor did large numbers of sessile organisms attach themselves to it. Kojima (1956) suggested that dolphins were found near floating objects because more food was available there, but was unable to demon- strate that they fed substantially on other fishes gathered at anchored bamboo rafts (Kojima, I960, 1961). Yabe and Mori (1950) argued that abundance of food was an inadequate ex- planation for the presence of yellowfin and skipjack tuna near floating logs because the fish took bait readily and did not have much food in their stomachs. The simultaneous pres- ence of piscivores and potential prey near the raft was well documented, yet, as mentioned above, only amber jack successfully preyed on the small fish that took shelter there. We saw them chase and eat freckled driftfish. The stomach of the only amber jack taken at the raft contained three driftfish. The only other species we saw catch smaller fish was the adult dolphin. Both it and the amber jack, as has been men- tioned, preyed on schools of goatfish that were near the raft, but not under it. The stomachs of 53 dolphins caught near the raft contained only 5 scrawled filefish; 1 sargassum trigger- fish, Xanthichthys r in gens (Linnaeus); and 1 puffer, Diodon holocanthus Linnaeus. All were juveniles. Once we saw an adult dolphin seize and eat a freckled driftfish which was attempt- ing to reach the raft. This incident suggests that dolphins sometimes intercepted driftfish seeking shelter. Possible supporting evidence for this supposition came from observations off Hawaii. While the raft was anchored for several days, numerous freckled driftfish, 19 dolphins, and 1 amber jack accumulated. The raft was then towed by the ship 30 miles down the coast and set adrift. During the tow the driftfish were outdistanced and all were lost; only the dolphins and amber jack remained. Thus, un- like other drifts, this drift began with a number of fish — 19 dolphins and 1 amberjack — at the raft. During 52 hours of drifting no freckled driftfish appeared. Yet in the same area, two weeks earlier, approximately 500 and 200 drift- fish gathered at the raft on two drifts of 50 and 32 hours, during which only 2 and 7 dolphins had accumulated. 494 PACIFIC SCIENCE, Vol. XXI, October 1967 Two other predators, wahoos and adult pilot- fish (with sharks), actively chased smaller fishes at the raft, but were not observed to catch any. Although zooplankton was not concentrated at the raft, a number of fishes that eat zooplankton gathered there. For example, stomachs of 10 rough triggerfish caught at the raft contained many pteropods and stomatopods, and lesser numbers of crab megalops and zoea, amphi- pods, and copepods. Stomachs of 81 freckled driftfish contained small pelagic tunicates ( Oikopleura sp.), copepods, fish eggs, chaeto- gnaths, and various coelenterates. These fish also bit at macroplankton such as ctenophores and tunicate colonies. Stomachs of 24 damselfish contained only Oikopleura sp. Stomachs of nine small pilotfish contained mostly copepods. All of these fishes, and also scrawled filefish and goatfish, frequently darted after and caught zooplankton around the raft. The wind slowly pushed the raft through the water at a speed faster than the swimming speed of the small zooplankters. Thus, there was no accumulation of zooplankton, but rather a continuous stream of macroplankton and microplankton slowly moving past the underwater windows. Finally, fishes at the raft did not feed on the small amounts of sessile or ambulating biota present. Only the rough triggerfish bit at the raft. Crab megalops occasionally settled on the underside of the raft or on the triggerfish, but those in the stomachs could have been taken as well from the plankton as from the raft. Perhaps a greater growth of biota on the raft would have altered the feeding behavior, es- pecially of the triggerfish, which has a dentition suited for grazing. Evans (1955) reported that triggerfishes (Balls tes sp., and Canthidermis sp.) cropped barnacles fringing the waterline of a drifting vessel in the Atlantic North Equa- torial Current. Removal of Ectoparasites At the beginning of this study we hypoth- esized that floating objects serve as cleaning stations where fishes may gather to have para- sites removed by other fish. Many fish observed at the raft carried ectoparasites, and several events suggested that these were eaten by other fish. Fish also chafed against the raft, another possible aspect of cleaning behavior. Small copepods were found on captured dolphins, freckled driftfish, and rough trigger- fish, and were seen on whitetip sharks and juvenile dolphins (Coryphaena sp.). Crab megalops and parasitic isopods were also seen on triggerfish. The megalops walked freely over the fish; the isopods were firmly attached. Biting behavior was common among rough triggerfish and was directed toward a trigger- fish that was headstanding (body oriented head down), apparently soliciting predation on para- sites. This behavior occurred only when more than one triggerfish was present; it was common 3 to 12 m from the raft. The headstanding fish did not flee the biting fish and once even appeared to rotate its body, keeping the side with the parasitic isopod toward the biting fish. The biting was always directed at the headstand- ing fish even though several other fish were very close by. Although we did not witness directly the removal of a parasite, we saw one rough triggerfish bite at a parasitic isopod on the caudal peduncle of another, and soon after- ward the isopod was missing. Biting did not appear to represent aggressive behavior; intra- specific aggression among triggerfish frequently occurred immediately under the raft, but did not include headstanding. In aggression one triggerfish repeatedly chased others from under the raft. Once a rough triggerfish swam to a dolphin and apparently nipped at it. The dolphin, some distance from the raft, had begun leaning to one side. It had also stopped swimming and was almost motionless in the water. It leaned four times within 2 minutes, for periods of about 9 seconds. Similar leaning behavior by dolphins in the presence of rough triggerfish was seen on several other occasions, but did not elicit nipping by the latter. This behavior was not unlike that of inshore fishes soliciting parasite-cleaning labrids (Randall, 1958). Ba- listids are not among the reported inshore parasite-pickers, but their dentition should make them efficient in this role. A juvenile dolphin, Coryphaena sp., with a small reddish copepod attached near the fork of the caudal fin repeatedly positioned itself so that its caudal fin was close to the head of Drifting Object and Pelagic Fish — Gooding and Magnuson 495 Fig. 6. Adult dolphin chafing against a 55-gallon drum beneath the raft. another juvenile dolphin, Coryphaena sp. Dur- ing the display the fish with the ectoparasite stopped caudal movements and treaded water with its pectorals. It did not lean to one side as did the adult dolphin mentioned above. On numerous occasions, the juvenile dolphin, Coryphaena sp., to which the display was di- rected made passes at the caudal fin of the parasitized fish. At the end of the day, how- ever, the copepod was still attached. Several species chafed their sides on the raft, skiff, or lines hanging in the water. Adult dolphin commonly chafed against the bottom of the raft and skiff (Fig. 6). Sanchez Roig and Gomez de la Maza (1952) and Heyerdahl (1950) have reported similar behavior. Some- times dolphin chafe against other fish (Breder, 1949). In one of our film sequences, a small abrasion can be seen on the side that the fish was rubbing against the skiff. Other species at the raft which were seen chafing were rough triggerfish on the bottom of the raft; juvenile dolphin on ropes and on the caudal and dorsal fins of whitetip shark; whale shark, whitetip shark, and scrawled fiiefish on the rope to the parachute drogue; and a spiny puffer, on a small floating can. This behavior, especially common in the coryphaenids, could remove parasites or relieve skin irritation. Some predation on ectoparasites occurred at the raft, but the question remains whether the removal of parasites is concentrated near the raft and other floating objects. It is obvious that removal of parasites by chafing on hard objects would be concentrated near floating material or larger fishes. In addition, the op- portunity to feed on ectoparasites or to solicit parasite cleaning would appear to be greater near the raft because the fishes usually arrived in small groups or alone and formed larger aggregations at the raft. Other Possible Explanations The hypothesis that fishes seek shade under floating objects has no substance. Yabe and Mori (1950) and Kojima (1956) also reached this conclusion. None of the smaller species tended to remain in the shade of the raft. Larger species such as rough triggerfish, wahoo, dolphin, and whitetip shark often ranged far from the raft and were seldom in its shadow. The hypothesis (Besednov, I960) that fish use floating material as a substance on which to lay their eggs could not be substantiated. Even though fish eggs are frequently found on drifting material, no fish deposited eggs on the raft nor were any. eggs seen on the undersurface. No data were obtained to test the hypothesis (Damant, 1921) that the shadow of an object makes the zooplankton more visible to fish. Four species fed upon zooplankton; the visi- bility of these zooplankters may have been increased by the raft’s shadow. CONCLUSION A floating object in the pelagic environment provides a relatively rare "superstrate” in an environment notable for its horizontal homo- geneity. This superstrate has some of the same ecological significance to certain pelagic fishes that a substrate has to inshore fishes. Obviously, no single biological association or adaptive ad- vantage can explain the occurrence of fish around floating objects at sea. Of the ecological hypotheses considered, shelter from predation is substantiated best and appears to be the most significant factor in the evolution of fish com- munities that gather beneath inanimate drifting material in the open ocean. ACKNOWLEDGMENTS We thank Randolph K. C. Chang, Robert T. B. Iversen, and Everet C. Jones, Bureau of Commercial Fisheries Biological Laboratory, Honolulu, who assisted in making the observa- tions; Frank J. Mather, Woods Hole Oceano- 496 PACIFIC SCIENCE, Vol. XXI, October 1967 graphic Institution, who verified the new record for Hawaii of Seriola rivoliana; John E. Randall, The Oceanic Institute and the Bishop Museum, Honolulu, Howard E. Winn, University of Rhode. Island, Philip Helfrich, University of Hawaii, Thomas A. Manar, and Donald W. Strasburg, Bureau of Commercial Fisheries Biological Laboratory, Honolulu, Charles M. Breder, Jr., American Museum of Natural History, and John H. Hunter, Bureau of Commercial Fisheries Tuna Resources Lab- oratory, La Jolla, California, who reviewed the manuscript during various phases of its develop- ment. REFERENCES Besednov, L. N. I960. Some data on the ich- thyofauna of the Pacific Ocean flotsam. Trudy Inst. Okeanol. 41:192-197. [Trans- lated from the Russian by W. Van Campen, Bureau of Commercial Fisheries, Honolulu, Hawaii.] Breder, C. M., Jr. 1949. On the relationship of social behavior to pigmentation in tropical shore fishes. Bull. Am. Mus. Nat. Hist. 94(2) :83-106. Damant, G. C. C. 1921. Illumination of plankton. Nature (London) 108:42-43. Eibl-Eibesfeldt, I. 1955. fiber Symbiosen, Parasitismus und andere besondere zwischen- artliche Beziehungen tropischer Meeresfische. Z. Tierpsychol. 12(2) :203-219. Evans, F. 1955. The "Petula” transatlantic ex- pedition, 1953-1954. J. Inst. Navig. 8(3): 205-210. Galea, J. A. 1961. The "kannizzati” fishery. Proc. Tech. Pap. Gen. Fish. Coun. Mediterr. 6:85-91. Gooding, R. M. 1965. A raft for direct subsur- face observation at sea. Spec. Sci. Rept. Fish. U. S. Fish Wildl. Serv. 517:1-5. Hardenberg, J. D. F. 1949. Development of pelagic fisheries. 1st Proc. Indo-Pacific Fish. Coun., Sec. 4:138-143. Heyerdahl, T. 1950. Kon-Tiki: Across the Pacific by Raft. Rand McNally and Co., Chicago. 304 pp. Inoue, M., R. Amano, and Y. Iwasaki. 1963. Studies on environments alluring skipjack and other tunas, 1. On the oceanographical condition of Japan adjacent waters and the drifting substances accompanied by skipjack and other tunas. [In Japanese with English summary.] Rept. Fish. Res. Lab. Tokai Univ. 1(1) :12— 23. Kimura, K. 1954. Analysis of skipjack ( Katsu - wonus pelamis ) shoals in the waters of "To- hoku Kaiku” by its association with other animals and objects based on records by fish- ing boats. [In Japanese with English sum- mary.] Bull. Tohoku Reg. Fish. Res. Lab. 3:1-87. Kojima, S. 1956. Studies of dolphin fishing conditions in the western Sea of Japan, II. "Tsuke” rafts and their attraction for the fish. (Fishing for dolphins in the western part of the Japan Sea. II. Why do the fish take shelter under floating materials?) [In Jap- anese with English summary.] Bull. Jap. Soc. Sci. Fish. 21(10) :1049-1052. [Translated from the Japanese by W. Van Campen, Bu- reau of Commercial Fisheries, Honolulu, Hawaii.] I960. Ibid. V. On the species of fish attracted to "Tsuke” rafts. (Fishing for dolphins in the western part of the Japan Sea. V. Species of fishes attracted to bamboo rafts.) [In Japanese with English summary.] Bull. Jap. Soc. Sci. Fish. 26(4) :379-382. [Translated from the Japanese by W. Van Campen, Bureau of Commercial Fisheries, Honolulu, Hawaii.] 1961. Ibid. III. On the stomach con- tents of dolphin. (Studies on fishing condi- tions of dolphins, Coryphaena hippurus L., in the western region of the Sea of Japan. III. On food contents of the dolphin.) [In Japanese with English summary.] Bull. Jap. Soc. Sci. Fish. 27(7) :625-629. [Trans- lated from the Japanese by W. Van Campen, Bureau of Commercial Fisheries, Honolulu, Hawaii.] Limbaugh, C. 1955. Fish life in the kelp beds and the effects of kelp harvesting on fish. Univ. Calif. Scripps Inst. Oceanog., Inst. Mar. Res. Rept. 55-9, 156 pp. 1961. Cleaning symbiosis. Sci. Am. 205 (2) : 42-49. Mansueti, R. 1963. Symbiotic behavior be- tween small fishes and jellyfishes, with new data on that between the stomateid, Peprilus Drifting Object and Pelagic Fish — Gooding and Magnuson 497 alepidotus, and the scyphomedusa, Chrysaora quinquecirrha. Copeia 1963 (1) :40— 80. Murchison, A. E., and J. J. Magnuson. 1966. Notes on the coloration and behavior of the dolphin, Coryphaena hippurus. Pacific Sci. 20(4) :515— 517. Randall, J. E. 1958. A review of the labrid fish genus Lahroides, with descriptions of two new species and notes on ecology. Pacific Sci. 12(4) :327— 347. Reuter, J. 1938. Voorlopig mededeling orn- trent het roempononderzoek. Mededeling no. 2B, Instituut voor Zeevisscheri j , Batavia. Sanchez Roig, M., and F. Gomez de la Maza. 1952. La Pesca en Cuba. Ministerio de Agricultura, Republica de Cuba, La Ha- bana. 272 pp. Senta, T. 1965. The importance of drifting seaweeds in the ecology of fishes. Japanese Fishery Resources Conservation Agency, Tokyo. [In Japanese.] Fish. Res. Bull. 13, 56 pp. Soemarto. I960. Fish behaviour with special reference to pelagic shoaling species: Lajang ( Decap terus spp.). 8th Proc. Indo-Pacific Fish. Coun., Sec. 3:89-93. Suyehiro, Y. 1952. Textbook of Ichthyology. [In Japanese.] Iwanami Shoten, Tokyo, 332 PP- Yabe, PL, and T. Mori. 1950. An observation on the habit of bonito, Katsuwonus vagans, and yellowfin, Neothunnus macropterus, schools under drifting timber on the surface of the ocean. [In Japanese with English summary.] Bull. Jap. Soc. Sci. Fish. 16(2): 35-39. [Translated from the Japanese by W. Van Campen, Bureau of Commercial Fisheries, Honolulu, Hawaii.] A Possible Relation between the Occurrence of a Dendritic Organ and the Distribution of the Plotosidae (Cypriniformes) W. J. R. Lanzing1 ABSTRACT: Three marine species of Plotosidae are found along the coastlines of the Indian and Pacific oceans, but the other 25 species occur exclusively in the Australian region. The majority of the Plotosidae are freshwater inhabitants, some of which are indigenous to both Australia and New Guinea. The marine members of the family and two freshwater members possess a dendritic organ. It is sug- gested that this organ has an osmoregulatory function. In their description of the catfish Plotosus anguillaris , Bloch (1794) and LaCepede (1803) mention the presence of a peculiar external structure situated posterior to the vent and between the pelvic fins. Cuvier and Valen- ciennes (1840) observed that this structure had no connection with the urogenital system, but was attached to the last abdominal verte- bra by means of a long tendon. According to Brock (1887) and Hirota (1895) this so-called dendritic organ consists of numerous well- vascularized epithelial folds. Weber and de Beaufort (1913), Taylor (1964), and Munro (1966) used the presence of this organ as a criterion in their keys to the family Plotosidae. However, apart from its use in taxonomy this structure has attracted very little attention. Recently, electron microscope studies by van Lennep and Lanzing (1966) have shown that the dendritic organ of Plotosus anguillaris (Bloch), Cnidoglanis macrocephalus (VaL), and Eur isthmus lepturus (Gunther) possesses two main cell types: principal cells containing parallel groups of cytoplasmic tubules and many mitochondria, and clear cells containing an unusual three-dimensional network of cyto- plasmic tubules. Because of a similarity be- tween these cells and those occurring in salt glands of sharks and marine birds and the chloride cells in fish gills, the authors suggested that the plotosid dendritic organ is involved in salt transport. This assumption has led to a study of a possible relation between the occur- 1 School of Biological Sciences, University of Syd- ney, Sydney, N.S.W., Australia. Manuscript received September 23, 1966. rence of a dendritic organ and the distribution of the Plotosidae in marine and freshwater en- vironments. DISTRIBUTION OF THE SILUROIDEI The siluroids of the Austral- Asian region are best represented in the area bounded by Thailand, Vietnam, and Indonesia. The follow- ing families occur in this region: Akysidae, Amblycipitidae, Bagridae, Chacidae, Clariidae, Heteropneustidae, Plotosidae, Schilbeidae, Si- luridae, Sisoridae (Bagariidae) , and Tachy- suridae (Ariidae). Except for the Tachysuridae and three species of Plotosidae none of these catfish are found east of the line of Wallace. In the Australian region, which includes Aus- tralia, New Guinea, and some adjacent islands (Darlington, 1957), only the Doiichthyidae, Plotosidae, and Tachysuridae occur. DISTRIBUTION OF THE PLOTOSIDAE All known species of Plotosidae are inhabi- tants of the Australian region, but three marine species have a much wider range (Fig. 1). Paraplotosus albilabris (C. et V.) occurs in Indonesia, Vietnam, and the Philippines (Suvatti, 1950; Herre, 1953; Kuronema, 1961). Plotosus canius Ham. Buch. is reported from East Africa as well as from Fiji (Fowler, 1959), but does not seem to occur in China or Japan. On account of its wide range it is surprising that it has not yet been reported from Australia, although it is present in New Guinea (Munro, 1958). Plotosus anguillaris (Bloch) is distributed over a vast area. The western 498 Distribution of the Plotosidae — Lanzing 499 Fig. 1. Distribution of the Plotosidae. , Principal range of the Plotosidae; . . . , range of Plotosus anguillaris ; / / /, range of Paraploiosus albilabris. For letter symbols see footnote for Table 1. limit of its range is formed by the Southwest Asiatic Barrier (Ekman, 1953), as it has been found in the Red Sea and the Suez canal (Fowler, 1956). Although present along the East African coast, it does not reach Cape Town (Smith, 1949). It occurs in Korea (Mori, 1952) and Japan (Okada, 1955) as well as in the Society Islands; the Eastern Pacific Barrier (Ekman, 1953) apparently forms the eastern limit of its range. Plotosus anguillaris is not recorded from the Marshall and Marianas Islands (Schulz, 1953), which ichthyologically are more related to the Hawai- ian region. The distribution of the other plotosids, how- ever, is strictly limited to the Australian region. Of the marine plotosids only Cnidoglanis macrocephalus is found all along the coast of mainland Australia, Tasmania, and New Guinea; the other marine species occur only in the northern half of Australia and in New Guinea. Most of the freshwater plotosids are indigenous to the Leichhardtian fluvifaunula (see map by Whitley, 1959), although Tan- danus tandanus Mitchell, for instance, is con- fined to the Mitchellian fluvifaunula. Table 1 shows that several freshwater plotosids are common to both Australia and New Guinea (e.g., Porochilus obbesi Weber). Other spe- cies are exclusively New Guinean, e.g., Neo- silurus gjellerupi (Weber), or Australian, e.g., Tandanus tandanus. Munro (1964) recently drew attention to the existence of differences between the ichthyological fauna of northern New Guinea (Gaimardian fluvifaunula) and that of southern New Guinea (Riechian fluvi- faunula). Apparently, some plotosids like Neosilurus gjellerupi and N. ater sepikensis (Whitley) live in the Gaimardian area, whereas N. brevidorsalis (Gunther) and N. ater ater (Perugia) are confined to the Riechian area. THE PRESENCE OF A DENDRITIC ORGAN AMONG THE PLOTOSIDAE Table 1 lists two groups of Plotosidae: the species in group A possess a dendritic organ, whereas this organ is lacking in the plotosids of group B. The species of group A occupy a marine or estuarine habitat, but thus far Olo- plotosus mariae Weber and Plotosus papuensis Weber have been found only in fresh water 500 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE 1 Distribution and Habitats of the Plqtosidae* SPECIES AUSTRALIA NEW GUINEA HABITAT Group A (dendritic organ present) 1. Cnidoglanis macrocephalus (Yal.) + + marine 2. C. micro ceps (Rich.) + — marine 3. C. muelleri (Klunzer) + — marine 4. Euristhmus lepturus (Gunther) + + marine 5. E. nudiceps (Gunther) + + marine 6. Oloplotosus mariae Weber — + R fresh water 7. Paraplotosus albilabris (Val.) + + marine 8. Plotosus anguillaris (Bloch) + + marine 9- P. canius Ham. Buch. — + marine 10. P. papuensis Weber — + R fresh water Group B (no dendritic organ) 1. Anodontiglanis dahli Rendahl + L — fresh water 2. Neosilurus argenteus (Zietz) + S — fresh water 3. N. ater ater (Perugia) + E + R fresh water N. ater sepikensis (Whitley) — + G fresh water 4. N. hart on i Regan — + R fresh water 5. N. hrevidorsalis (Gunther) + L,J + R fresh water 6. N. equinus (Weber) — + R fresh water 7. N. gjellerupi (Weber) — + G fresh water 8. N. glencoensis (Rendahl) + L — fresh water 9. N. hyrtlii Steindachner + L,J — fresh water 10. N. idenburgi (Nichols) — + G fresh water 11. N. meraukensis (Weber) — + R fresh water 12. N. mortoni Whitley + L — fresh water 13. N. novaeguineae niger (Nichols) — + G fresh water N. novaeguineae novaeguineae (Weber) — + R fresh water 14. N. perugiae (Ogilby) — + R fresh water 15. N. rendahli (Whitley) + E — fresh water 16. Porochilus obbesi Weber + E + R fresh water 17. Eandanus bostocki Whitley + Gr — fresh water 18. T. tandanus Mitchell + M — fresh water * Symbols used: +, present; — , absent; G, Gaimardian fluvifaunula; Gr, Greyian; J, Jardinian; L, Leichhardtian; M, Mitchellian; R, Riechian; S, Sturtian. (Weber and de Beaufort, 1913). All the spe- cies of group B are freshwater inhabitants. Examinations carried out on adult and juve- nile individuals of Plotosus anguillaris, Cnido- glanis macrocephalus, and Eur isthmus leg turns showed that the dendritic organ is present and equally developed in both sexes, and also that it is already conspicuous in juvenile catfish ranging in size between 45 and 56 mm. Re- cently, van Lennep (unpublished) found that, in comparison with adults, the dendritic organ of juvenile catfish contains only a relatively small number of fully developed glandular cells. The mean length of nine adult Plotosus anguillaris was 322 mm (286-361 mm) total length. Since some of the gonads were either in a fully mature or in a spent condition, these measurements must represent the size of adult catfish of this species. Comparable figures have been reported by Delsman and Hardenberg (1934), 300 mm; Okada (1955), 250 mm; and Fowler (1959), 460 mm. A much higher figure for maximum size (30 inches) is given by Smith (1949), and is quoted by Munro (1954) and Fowler (1956), but probably is erroneous. DISCUSSION More than half of the 28 species of Ploto- sidae are freshwater inhabitants. This makes Distribution of the Plotosidae — Lanzing the Plotosidae a predominantly freshwater family rather than a chiefly marine family, as is often implied in the literature (Berg, 1957; Darlington, 1957; Nikolsky, 1961; Sterba, 1963). It appears that 25 species are found in the Australian region. Three marine species occupy a much larger area covering most of the Indian and West Pacific oceans. Sterba’s map (1963) of the distribution of the Plotosidae therefore actually shows the range of one species, namely Plotosus anguillaris. It is of interest that among the non-plotosid siluroids only T achy sums th al- ias inus (Rueppel) has a range as wide as that of Plotosus canius , except that the former is also reported from Japan (Matsubara, 1955). As yet no physiological work has been car- ried out with regard to the function of the dendritic organ. The available evidence indi- cates that: (a) its structure resembles that of salt-secreting glands in other vertebrates, (b) it is present in both juvenile and adult ploto- sids, (c) there exist no sexual differences, and (d) it is not present in freshwater plotosids other than Oloplotosus mariae and Plotosus pap Mens is. Hardenberg and Delsman (1934) suggested that the dendritic organ is involved in reproduction, but produced no evidence in support of this claim. A possible respiratory function merits consideration since other cat- fish (Clariidae, Heteropneustidae) possess ac- cessory respiratory organs. These, however, are in connection with the branchial chambers and are structurally different from the dendritic organ. Furthermore, there seems to be no rea- son why only marine plotosids should require an accessory respiratory organ. Both marine and freshwater plotosids enter muddy environ- ments that could contain oxygen-deficient water. A salt-excretory function seems to be the most likely, although, admittedly, the presence of a dendritic organ in Oloplotosus mariae and Plotosus papuensis does not fit in with this theory. Too little is known about these two species to venture any explanation. According to Darlington (1957:46) marine plotosids may have invaded the Australian region and, after entering a freshwater habitat, reached the end of a complicated line of teleost evolution. Although evolutionary aspects of zoogeog- 501 raphy must remain speculative, it is suggested that the dendritic organ was developed while plotosid ancestors in the Southeast Asian region invaded the sea. This invasion would be differ- ent from that of the Tachysuridae, which were able to cope with osmotic stresses by means of mechanisms similar to those used by other old teleost families, such as the Salmonidae. Per- haps because of tachysurid competition, the plotosid ancestors became firmly settled only in the Australian region. From them would have evolved, on the one hand, the freshwater spe- cies which lost the dendritic organ in the process. Some of the marine species, on the other hand, managed to disperse radially along the edges of the Indian and West Pacific ocean basins. I am grateful to Mr. I. S. R. Munro (C.S.I. R.O. Marine Laboratory, Cronulla) for his ad- vice and for putting at my disposal the manu- script for his forthcoming book on the fishes of New Guinea. REFERENCES Berg, L. S. 1957. Classification of Fishes Both Recent and Fossil. Edward Bros. Ann Arbor, Michigan. Bloch, M. E. 1794. Naturgeschichte der Aus- landischen Fische, vol. 8. Schlesinger, Berlin. Brock, J. 1887. tJber Anhangsgebilde des Urogenitalapparates von Knochenfische. Z. wissen. Zool. 45:532. Cuvier, G., and A. Valenciennes. 1840. His- toire Naturelle des Poissons. XV. Levrault, Paris. Darlington, P. J. 1957. Zoogeography. The Geographical Distribution of Animals. John Wiley & Sons, New York. Delsman, H. C., and J. D. F. Hardenberg. 1934. De Indische Zeevissen en Zeevisscheri j . Visser and Co., Batavia. Ekman, S. 1953. Zoogeography of the Sea. Sidgwick and Jackson, London. Fowler, H. W. 1934. The fishes of Oceania, Suppl. 2. Mem. B. P. Bishop Mus. 10(6). Honolulu. 1938^. The Fishes of the George Van- derbilt South Pacific Expedition. Acad. Nat. Sci. Philadelphia, Monogr. 2. 502 PACIFIC SCIENCE, Vol. XXI, October 1967 1938A A List of Fishes Known from Malaya. Fish. Bull. 1. Govt. Printing Office, Singapore. 1956. The Fishes of the Red Sea and Southern Arabia. Vol. 1. Weizmann Sci. Press, Jerusalem. — 1959. Fishes of Fiji. Govt. Fiji Publ., Suva. Herre, A. W. 1953. Checklist of Philippine Fishes. U.S. Fish Wildl. Serv. Res. Rept. 20. Hirota, S. 1895. On the dendritic appendage of the urogenital papilla of Plotosus anguil- laris. J. Coll. Sci. Imp. Univ. Japan 8:367. Kuronema, K. 1961. A Checklist of Fishes of Vietnam. Div. Agric. Nat. Resources, U. S. Operations Miss. Vietnam. LaCepede, B. G. E. 1803. Histoire Naturelle des Poissons. V. Plassan, Paris. Lennep, E. W. van, and W. J. R. Lanzing. 1967. The ultrastructure of glandular cells in the external dendritic organ of some marine catfishes. J. Ultrastr. Res. (In press.) Matsubara, K. 1955. Fish Morphology and Hierarchy. Ishizaki-Shoten, Tokyo. Mori, T. 1952. Check list of the fishes of Korea. Mem. Hyogo Univ. Agric. 1:228. Munro, I. S. R. 1957. Handbook of Australian fishes, No. 10. Fish. News Letter 16(4) :4l. 1958. The fishes of the New Guinea region. Papua and New Guinea Agric. J. 10(4) :97. - 1964. Additions to the fish fauna of New Guinea. Papua and New Guinea Agric. J. 16(4) :141. 1966. Fishes of New Guinea. (In press.) Nikolsky, G. V. 1961. Special Ichthyology. Israel Program. Sci. Transl., Jerusalem. Norman, J. R. 1951. A History of Fishes. Ernest Benn, London. Okada, Y. 1955. Fishes of Japan. Maruzen, Tokyo. Schulz, L. P., et al. i960. Fishes of the Marshall and Marianas Islands. Vols. I and II. Smithsonian Inst., U. S. Natl. Mus. Bull. 202. Smith, J. L. B. 1949. The Sea Fishes of South- ern Africa. Central News Agency, Cape Town, South Africa. Sterba, G. 1963. Freshwater Fishes of the World. Vista Books, London. Suvatti, C. 1950. Fauna of Thailand. Dept. Fish., Bangkok. Taylor, W. R. 1964. Records of the American- Australian Expedition to Arnhemland. Vol. 4, Zoology. Melbourne Univ. Press. Weber, M., and L. F. De Beaufort. 1913. The Fishes of the Indo- Australian Archi- pelago. II. Brill, Leiden. Whitley, G. P. 1959. The freshwater fishes of Australia. In: Ed. Keast, et ah, Bio- geography and Ecology in Australia. Junk, The Hague. The Family Olividae John Q. Burch and Rose L. Burch1 Shells of the family Olividae are certainly among the most beautiful in form, color, and markings that we possess. Nevertheless, they attract comparatively little attention from the general collector. This is remarkable because the shells are extensively distributed over the globe, are easily collected, and are easily procured at a moderate rate. We believe that this lack of attention is due to the uncertainty with regard to the number of species, and to the lack of reference material. It is difficult to define the limits of many of the species on account of the great variation in color and the extraordi- nary manner in which the markings gradually change in character while species of other groups are easily distinguished. Linnaeus (1758) divided the shells we know as the Olividae into 3 species, V oluta oliva, V . porphyria, and V. ispidula. Gmelin (1791) added a few, and Lamarck (1811), extended the number to 62. Dillwyn (1817) reduced it to 18. Duclos (1835) figured 84 species; he considered that 22 of the species which Lamarck described as distinct were only varieties of other species. Reeve (1850) published figures of 100 species. In 1858 J. E. Gray published "An at- tempt to distribute the species of Oliva into natural groups,” but his work did not meet with general acceptance by other authors. In 1870- 1871 F. P. Marrat published his "Monograph of the Genus Oliva ” in which he figured 220 species. In 1883 George W. Tryon published volume V of the Manual of Conchology, which contains his monograph of the Olividae. This was a monumental work, but Tryon was dis- posed to group related species to a greater extent than most authors have approved. Tryon re- duced the number of species to 55. The more recent major publications on this family include those of Johnson (1910-1911, 1915, 1928); Dautzenberg (1927); Dodge (1950); and Olsson and Dance (1966). 1 1400 Mayfield Road, Apt. 61 L, Seal Beach, California 90740. Manuscript received May 25, 1966. Marrat (1870-1871) made the following statement regarding the Olividae: "Specific dif- ferences confined within limited areas constitute the exceptions not the rule. In almost every case where the shells can be obtained in numbers they approach the so called species above and below them so as to render it a matter of un- certainty whether they constitute a variety of one or the other.” In an effort to trace the rela- tionships between the species Marrat introduced many new names. Sowerby (1870-1871) com- mented in Thesaurus Conchyliorum : "In his study of the affinities he has been led to register and nominally to admit as species many forms which will appear to the readers as they do to the editor quite indistinguishable.” Ford (1953) said of Marrat: "Judged from the number of forms to which he gave names Marrat might be considered a splitter. Later, however, he stated that the 220 species of Oliva might, if carefully examined, be reduced to twenty, and the greater part of his own species reduced to varieties.” The Marrat collection of shells of the Olivi- dae was purchased by the Liverpool Museum in 1875 and remains intact and available. The work of the late J. R. le B. Tomlin listing the species and designating the various type speci- mens was published by Ford (1953). Tomlin’s comments are of great value to the worker on this group. Lamarck’s types of the Annales du Museum are in Caen. The Duclos collection is in the Geological Society of London. The Lovell Reeve collection of olives was sold at public auction at the Steven’s Auction Rooms. A large part of the specimens were purchased by Marrat and incorporated in his collection. The Weinkauff collection is in the museum at Frankfurt, Germany. All of the great collections of Olividae remain intact in large public in- stitutions. The type specimens representing the work of leading scholars from Lamarck to those of the present day remain available to the seri- ous student. These specimens were named and studied in good faith by recognized workers 503 504 PACIFIC SCIENCE, Vol. XXI, October 1967 who did original investigation, and they have been studied by a host of serious authors who followed them. It is tragic that we are com- pelled to abandon such solid material and ac- cept references to a series of poorly drawn old wood cuts. Typical of these are some of Roding in the notorious Museum Boltenianum in which, for some, Roding lists as many as four refer- ences all to entirely different species, some of them unrecognizable, and the actual shell has been sold as a curio and lost. What the species may have been is known only to God. We propose to recognize four major divisions in the Olividae with others as subgenera: Oliva, Olivella, Agaronia, and Ancilla. On shell char- acters alone the generic distinctions between Oliva and Olivella are difficult to define. In gen- eral, the species of Oliva are larger, although there are many exceptions. Oliva has no oper- culum and no epidermis. In general, Olivella is distinguished from Oliva by the small size of the shell, its more produced spire, and the presence of a thin horny operculum. Neverthe- less, some species of Olivella, e.g., Olivella nivea (Gmelin, 1791), lack an operculum. Anatomi- cal characters must be considered. The animal of Olivella is like that of Oliva, but the tentacles and eyes are wanting, the foot is shorter, rounded behind, and does not extend beyond the tip of the spire. Species of Olivella are best distinguished from Oliva by the radulae. In Oliva the radular ribbon generally shows but little variation, the differences between species being small. The ribbon of Oliva is generally long and narrow, with many rows of teeth (about 100), while the rhachidian teeth are tricuspidate, the basal margin of the ribbon is wide and often yoke-shaped. In Olivella the ribbon is short and wide, with fewer rows of teeth (generally less than 50), the rhachidian teeth are multicuspidate, the cusps being small and numerous. While the rhachidian teeth of both Agaronia and Olivancillaria are tricuspi- date, there are small denticles on the sides that are not present in Oliva. The rhachidian teeth of both are very similar in this character. The radula of Ancilla is somewhat different. The rhachidian teeth are tricuspidate, but there are two strong lateral teeth. Some species of Oliva are a source of con- fusion, with various authors accepting different names for the same species. These notes are an attempt to indicate our diagnosis of the tax- onomy. We have freely adopted the work of many others. It would seem that the only ad- vantage to be gained from the recognition of color forms is to give the references to the writ- ing of authors in which they frequently add substantially to an understanding of the species involved. With few exceptions there are inter- grades in and out of all of them. It is our opin- ion that naming them is somewhat analogous to describing all of the kittens in a litter. We will discuss the better known color forms that have been given names. In this paper we will discuss the taxonomy of certain species from the Indo-Pacific of the genus Oliva Bruguiere 1789. Oliva oliva (Linnaeus 1758), Systema Naturae, ed. 10, p. 729, no. 350; ed. 12, p. 1188, no. 399. This species has been confused by many au- thors with O. ispidula (Linnaeus 1758). The recent publication by Olsson and Dance (1966) seems to have established the fact that the true Voluta ispidula Linnaeus is a fossil species of Agaronia, and that the O. ispidula of authors is the Voluta oliva of Linnaeus. Generations of authors have described and figured this species as O. ispidula. The synonymy is extensive. The figures and discussion given by Reeve (1850) are excellent. Try on (1883) gives a clear de- scription: "White, ash, yellow, brown, chestnut or chocolate colored, without markings, or with nebulous spots, zigzag lines or reticulations, often with a band near the top of the body whorl; columella white; interior chocolate colored. Length 1-1.5 inches.” It is impossible to enumerate the shades and patterns of coloring of this species. The chocolate-colored interior is the most characteristic feature. We will men- tion a few of the described color forms. algida Vanatta 1915. Nautilus 29:67-72. It is our opinion that this variety is not suf- ficiently distinct. It was based, with doubt, on a figure of Reeve (1850). Nevertheless, other workers have accepted this as a variety, and Reeve’s figure is clear. The shell is a bluish- white with light-brown longitudinal streaks, a yellow-brown lip, and a shorter spire than usual. The Family Olividae — Burch and Burch A brown and white callus is seen in the pos- terior comer of the aperture. Candida Lamarck 1811. Ann. Mus. Hist. Nat. 16:322. It is doubtful if this is a form of O. oliva. The figures represent a much shorter and more obese olive than any example of O. oliva, and also the interior of the aperture is white. Never- theless, others have accepted this form. Dodge (1955) lists it as a recognized variety stating: "An albino form with the aperture more orange than brown.” flaveola Duclos 1845. In Chenu 1845. Duclos has represented under this name four shells of different colors. According to the de- scription, this variety is light or dark, yellow or orange, the others being of unusual pattern. In this form the interior of the aperture is fre- quently whitish or rose, but we possess a number of specimens in which it is dark brown. The dark coloration of the aperture nearly always permits identification of this species. This form can be regarded as albinistic, with an aperture of light brown to almost white. Dodge (1955) states: "A yellow form with a white aperture.” gratiosa Yanatta 1915. Nautilus 29:71. This is a doubtful form and we have failed to recognize it. Vanatta’s description is brief and without references: "Shell slender, dark brown, spire elevated, columellar callus cream-white.” Dodge (1955) states: "A dark brown, slender form with an elevated spire. It is possibly the form which Lamarck called O. oriola (1811) although the spire is considerably higher. The columella is cream-white.” jayana Ducros de St. Germain 1857. Revue Critique, p. 68, pi. Ill, fig. 44a, 44b. O. jayana, described as a distinct species by Ducros, is only a form of O. oliva, very close to tigridella Duclos 1835, having a slender body and short spire. Its pattern is formed of very fine lines composing a condensed network of small triangular meshes. Ducros admitted that he considered his species doubtful. Dodge (1955) described jayana : "White or flesh colored with fine longitudinal brown lines and two or more less prominent bands of irregular markings.” 505 lacteana Dautzenberg 1927. J. Conchyl. 71: 49. Dautzenberg stated that the external color is entirely white, without pattern, the aperture brown. He also stated that this form has been confused by many authors with O. Candida Lamarck 1811. Large series in the Burch collec- tion contain specimens that fit the description in every way, but the gradation into other forms is obvious. longispira Bridgman 1906. Proc. Malacol. Soc. London 7:195. This form differs from typical O. oliva by hav- ing a more elevated spire. Dodge (1955) com- ments: "A name given to a high spired form which Johnson believed to be identical with the latter’s samar ensis 1915.” martini Dautzenberg 1927. J. Conchyl. 71:53. Dautzenberg described this form as whitish or flesh colored, ornamented at the top of the last whorl with a transverse orange band. Numerous sets in the Burch collection could be assigned to this form, but all are associated with other forms. oriola Lamarck 1811. Ann. Mus. Hist. Nat. 16:321. O. oriola of Lamarck is a black form of O . oliva in which the interior is usually lighter and sometimes nearly white. The external black coloration is not always uniform. It is often blended with white spots more or less wide- spread, which in some specimens occupy half the surface. Sets in the Burch collection from numerous localities could be assigned to this form. Duclos (1835) has represented under the name O. oriola several dark examples of O. reticularia Lamarck 1811. samarensis Johnson 1915. Nautilus 29:71. Johnson stated that his shells were from Samar, Philippines. All were uniform in color repre- senting the dark reticulated form (Thes. Con- chyl., fig. 248). The types are in the Academy of Natural Sciences at Philadelphia (no. 111759). We consider this name questionable on all counts. stellata Duclos 1835. Monogr. Oliva, -pi. 8, figs. 11, 12. This form is ornamented with little lines or dots more or less disposed in zigzags. It is devoid 506 PACIFIC SCIENCE, Vol. XXI, October 1967 of a transversal zone. Dodge (1955) describes it as "A white form with broad coarse mark- ings, and a shorter spire than in most forms of this species.” An interesting series in the Burch collection contains specimens of this color form, and would fit the figure of Reeve (1850, pi. 17, fig. 34 d), being ivory-white marked spar- ingly with dashes of violet-brown. However, other color forms are taken from the same locality. Specimens are easily selected from large sets from the Philippines and other localities. taeniata Link 1807. Besch. Nat. Samml. 2:98. This is a form stated by Link to be distinguished by the dark unicolored band at the upper end of the last whorl. This description is far too inadequate, and in our opinion the name should be ignored. tigri della Duclos 1835. Monogr. Oliva , pi. 8, figs. 13, 14. This form differs from stellata by the pattern of numerous punctations, sometimes isolated, sometimes close together, and aligned longi- tudinally in zigzags. The background is yellow- ish gray, rather dark. Dodge (1955) lists this form: "Fawn colored with dark spots.” We do not accept the findings of Bridgman (1905), who made a case for the recognition of O. tigri della as a distinct species. Reeve (1850), Weinkauff (1878), Tryon (1883), and others have considered it to be a form of the species under discussion. The name has been used by collectors, and specimens of this color form may be selected from large sets from many localities. Oliva miniacea (Roding 1798). Mus. Bol- tenianum, p. 33, sp. 391. 6. Das Mergen- roth Gmel. V, Porphyria , sp. 16 b; Martini 2, t. 45, f. 476, 477, 9 St. These references are unmistakably to the species long known as O. erythrostoma (Meu- schen 1787), Mus. Geversianum, p. 376. The work of Meuschen has been officially declared invalid by the International Commission of Zoological Nomenclature in opinions 260 and 26L It is our opinion that O. sericea (Roding 1798) and O. tremulina Lamarck 1811 are recognizable distinct species. Johnson (1910) and others wished to incorporate these species as forms of O. miniacea . Typical O. miniacea (Roding) is composed of shells ornamented with wavy longitudinal lines, and two trans- verse bands more or less interrupted, encircling one at the top and one at the center of the last whorl. All, however, are yellow-white, streaked and banded with blue, green, and purple. The aperture is always a deep orange. O. miniacea is well illustrated in many pub- lications and often as O. erythrostoma (Meu- schen 1787). Among the more recent publica- tions are those of Kira (1962, pi. 32, figs. 4, 5), and Habe (1966, vol. II, pi. 27, fig. 16). The species name porphyretica Martini 1773, used by some authors, can not be defended because it was not established according to the International Rules. The name was used by Marrat (1871) and was based on a small specimen otherwise close to the typical Melvill and Standen (1897) cited an O. mes- saris Duclos 1835 which is perhaps a form of O. miniacea, but it is impossible to know ex- actly the species they intended to designate since Duclos described and figured two very different shells. One (pi 12, fig. e), which agrees with the description, is a large example of O. tremulina , while the other (pi. 22, figs. 7, 8), half as large, is, according to Ducros de St. Germain (1857), a worn and discolored O. miniacea. O. azemula Duclos 1835 should be nullified, as it is based on O. ponder osa Duclos 1835, pi 15, figs. 1, 2, and on an O. miniacea with- out bands (pi 15, figs. 10, 11). It is impossible to mention the many color forms of O. miniacea , but some of the better known follow. efasciata Dautzenberg 1927. J. Conchy!. 71: 39. Dautzenberg named this form on the theory that since Duclos (1835) had first figured an O. ponder osa under the name azemula , the second figure could not be designated as O. azemula. In any event, this is the form without bands which is not uncommon from many localities. joknsoni Higgins 1919. Nautilus 33:58. This form is based on Figure 110- of Marrat The Family Olividae — Burch and Burch (1871). It is a form of dark brown to black color with large white markings. The form is common from many localities, and in colonies merging into other color forms. man at i Johnson 1910. Nautilus 24:51. This form is entirely dark brown to black. It is common in many localities. saturata Dautzenberg 1927. J. Conchyl. 71 : 39- This form is described as having darker longi- tudinal lines, more numerous and the bands more colored, in such a manner that the entire shell has a more sombre aspect. Shells of this description can be selected from almost any long series of specimens from the Philippines and other localities. sylvia Duclos 1835. Monogr. Oliva, pi. 14, figs. 10, 11. It would seem that the shells generally as- signed to this form are orange-yellow with ir- regular lines, and having two bands of brown usually smaller than in other forms. This color form is quite common in specimens from Zamboanga, Philippines. Oliva tremulina Lamarck 1811. Ann. Mus. Hist. Nat. 16:310. The only reference given by Lamarck for O. tremulina is the figure of Lister (1685, pi. 727, fig. 14). It is a large shell of which the back- ground is yellowish-white, and is ornamented with heavy longitudinal lines and purplish- brown dots. The last whorl is crossed by two transverse bands of wide blackish dots. The interior of the aperture is fleshy white. O. nobilis Reeve 1850, pi. 2, sp. 3 a.b.c, of which the dimensions are exactly those of Lister’s figure, and of which the pattern is quite simi- lar, falls into synonymy with typical O. tremu- lina. Johnson (1928:8-9) considered O. tre- mulina to be a form of O. miniacea (Roding 1798). A number of authors accepted this conclusion, but it seems to us that the speci- mens before us labelled O. tremulina not only have a fleshy-white aperture, but seem to be less swollen at the posterior or shoulder of the shells. They are otherwise close, but we think that they are easily separable. 507 Some of the color forms of O. tremulina follow. concinna Marrat 1871. Thes. Conchyl., p. 13, pi 7, figs. 100, 101. This form is of a uniform blackish brown, or well sprinkled with a few unusually white spottings more or less triangular. Occasionally, one also sees examples irregularly marked with brown and gray. O. tenebrosa Marrat 1871 differs only in the smaller size, which is insuf- ficient to make another form inasmuch as many examples of intermediate sizes are encountered. We think that Johnson (1928) was incorrect in placing O. concinna Marrat 1871 with O. pica Lamarck 1811 as these shells are very different. Weinkauff (1878) placed O. concinna in the synonymy of O. zeilanica Lamarck 1811. A few of the sets before us at this time follow. The shells from Ceylon are larger, lack- ing the brown edge of the interior lip, marbled with brown and gray. These seem to fit this form. Shells from Bougainville, Solomon Is- lands, are of uniform blackish brown, some with a few white spottings usually triangular. Large sets from Zamboanga, Philippines, con- tain shells with all patterns of this form. chrysoides Dautzenberg 1927. J. Conchyl. 71:139. This form is a golden-yellow or orange, solid- colored or with a very faint pattern. The slender form and high spire with open suture is very close to the form zeilanica (Lamarck) Philippi 1845, from which it differs only in coloration. It is difficult to see why Reeve (1850) and Marrat (1871) united it with O. iris an s (Lamarck) Duclos 1835, in which the flattened spire is entirely covered by a callosity. Duclos (1844) cited with doubt as forms of O. tremulina, O. obtusaria and O. hepatic a, but it is impossible to identify these names of Lamarck (1811), the descriptions of which are insufficient and which are not accom- panied by a reference. The Burch collection contains sets that fit this form from Zamboanga, Philippines, and also from the Great Barrier Reef, Australia, but the last are larger shells and more slender than those from the Philippines. 508 PACIFIC SCIENCE, Vol. XXI, October 1967 Oliva sericea (Roding 1798). Mus. Bolteni- anum, p. 33, sp. 390. Roding gave two references. One was to Martini (1773, pi. Ll, figs. 559, 56l), which obviously describes the species known to many as O. textilina Lamarck 1811. The other refer- ence is to Gmelin (1791, sp. 17, var. 88), based on Figure 489 of Martini (1773), and is, according to Pfeiffer (1840), an O. reticularis Lamarck, but it is probably O. tricolor Lamarck because of the black points which adorn the upper whorls. It is not O. sericea in any event. O. textilina Lamarck 1811 is in the synonymy of O. sericea (Roding 1798). In comparing O. sericea with O. tremulina Lamarck 1811 one notices that its columellar edge is adorned up to the top of the aperture by a thick callosity which spreads wide on the base, and that the columellar plaits are stronger and less numer- ous. Its pattern is composed of a multitude of very fine intercrossed lines which form a network. The bands also are composed of very fine and closely knit lines. Finally, the inside of the aperture, which is white in the back- ground and barely flesh-colored at the base of the columella and along the lip in O. tremulina, is entirely light yellow to slightly salmon in O. sericea. The species is well illustrated as O. textilina Lamarck 1811 by Reeve (1850, pi. 6, fig. 9, a.b.c.) . Tryon (1883, pi. 27, figs. 59, 60) figured the shell well, but in our opinion con- fused the species with his concept of O. irisans Lamarck 1811. Hirase (1938, 1951, ph 113, fig. 1) and Kira (1955, pi. 31, fig. 15) figured the species well as O. sericea. Specimens matching our concept of this species may be seen, bearing a monumental assortment of names, in most of the major col- lections. It is understandable that many are confused with O. tremulina Lamarck 1811. We noted a number of sets assigned to O. sabulosa Marrat 1868. We place this name in the synonymy. The comments of Ford (1953) on the Liverpool types may be of interest here: "Two possible syntypes. No locality. 52 mm. X 22 mm., 40 mm. X 18 mm- I recently found these shells in a tray with the cut out descrip- tion of the species. They had not been seen by Tomlin. G. L. Wilkins has identified them as young specimens of O. sericea (Roding 1798).” A few of the described color forms of O. sericea follow. granitella Lamarck 1811. Ann. Mus. Hist. Nat. 16:314. Oliva granitella has been regarded by some au- thors as a distinct species, by others as a synonym of O. sericea, and by Ducros (1857) and Weinkauff (1878) as a variety of that species. In comparing the descriptions of Lamarck, one sees that there are no transverse bands in O. granitella, while O. sericea offers two transverse bands more or less marked, composed of little brown lines closed in zigzags and resembling the characters of script. One can thus suppose that O. granitella is a variety of O. sericea without the bands, but in the absence of all figuration this interpretation remains doubtful. albino. Melvill and Standen 1897. J. Con- chyl. 8:404. This form, designated as being ivory-white without other shell characters, may have been created for an albino specimen of O. sericea, as cases of albinism are known in many species of olives, but, to be sure of its determination, we must know in what sense O. sericea has been considered by Melvill and Standen. O. sericea, O. tremulina, O. ponderosa, etc. have often been considered forms of the same species. As for O. sericea var. albescens Johnson 1915, this can only be an albino form of O. lignaria Marrat 1868. Johnson stated that the spire is callous. Oliva lignaria Marrat 1868. Ann. Mag. Nat. Hist., 4th series, 2:212. The holotype is in the Liverpool Museum and was mentioned by Tomlin (1953). A set of Marrat’s types is in the collection of the ' Academy of Natural Sciences at Philadelphia, no. 150597. These are presumably paratypes and are labelled as coming from Broome, Western Australia. It seems that the first au- thor to recognize the distinctive characters of this species was Ford (1891), in his description of O. cryptospira Ford 1891. It is to be re- gretted that this name must fall into the synonymy. The synonymy (as well as color forms) is extensive, but the two names given The Family Olividae — Burch and Burch most often by authors are both homonyms : O. ornata Marrat 1867 (not Roding 1798) and O. cylindrica Marrat 1867 (not B orson 1830, nor Sowerby 1850). Reeve (1850), Tryon (1883), and others assigned the species to O. irisans Lamarck 1811. No reference was given to a figure by Lamarck in his description. The fact that the figure of Reeve (1850) does not agree with Lamarck’s description in any essen- tial features is quite apparent. Deshayes (1844) refers to Dillwyn (1817). Dillwyn gives one reference to Martini (1773, fig. 561), which is obviously O. sericea (Roding 1798). It is dif- ficult to explain how Reeve (1850) and Marrat (1871) could have united O. tremulina La- marck 1811 and O. irisans (Lamarck) Duclos 1835, in which the flattened spire is entirely covered with callosity. Tryon (1883) followed Reeve (1850). Kira (1955, 1962) illustrated the species under the name O. ornata Marrat 1867. The typical form from western and northern Australia is slender and white with a fine zigzag pattern in ash to purple-brown. The columella is white, faintly tinged with lilac, the aperture light to deep violet. The color of the interior is not a constant character: a cer- tain percentage of specimens will range from white through light to deep violet. The apex of the Australian form is not as flat and cal- loused as are those from elsewhere throughout the Indo-Pacific. A brief comment on a few of the described color forms follows. albescens Johnson 1915. Nautilus 28:99. This pure-white albinistic form is not rare. cryptospira Ford 1891. Nautilus 4:135-136. The types are in the Academy of Natural Sci- ences of Philadelphia. This form is predomi- nantly orange in color. The spire is short with the sutures entirely covered by a heavy callus. fordi Johnson 1910. Nautilus 24:51. This is the dark brown form. It is common with others from the Philippines, Ceylon, and many other localities throughout the Indo-Pacific. Oliva vidua (Roding 1798). Mus. Boltenianum, pp. 34, 412, 20, Porphyria vidua. Die ungarische Wittwe. Gmel. V, sp. 17, Mar- tini 2, t. 45, f. 472, 473 St. The figures of Martini are clearly the solid 509 black form of the species known to many as O. maura Lamarck 1811. There has been an interval in which authors have placed this species in the synonymy of O. oliva (Linnaeus 1758). Olsson and Dance (1966) show that the true O. oliva is, in fact, the species known to many as O. ispidula (Linnaeus 1758). The name O. vidua must be restored. The shape of O. vidua seldom varies. It is relatively a little elongated. The spire is very depressed, often completely flat. A projecting callosity restricts, at the top of the columellar ridge, a scanty canal. This species is remark- able for the richness and variety of the patterns and colors. The typical coloration is a brilliant black, which has attracted the attention of au- thors. Reeve (1850, pi. 7, sp. lOa-lOg) figured this plate for O. maura Lamarck 1811, and in his listings of synonyms cited three of Lamarck’s 1811 species, O. fulminans, O. septuralis, and O. funebralis, and also O. macleaya and O. leucostoma of Duclos (1844). These are all recognized color forms of O. vidua with the exception of O. funebralis and O. leucostoma. However, Reeve did recognize in all figures the details of the spire and shoulder of O. vidua. Tryon (1883) figured the species well as O. maura Lamarck, as did Weinkauff (1878) with the color forms. T. Habe (1966) figured the species well as O. oliva. Some of the color forms of O. vidua follow. albofasciata Dautzenberg 1927. J. Conchyl. 71:70. Dautzenberg based this name on a figure of Duclos (1844, pi. 25, fig. 4). The gray back- ground covered with a compact and faint pattern is crossed by two decurrent bands linked by little irregular black swatches. aurata Roding 1798. Mus. Boltenianum 33. This form is composed of shells with a uniform golden-yellow to orange color. We may add here that, as with other color forms, we have long series showing the gradual merging of one into others. c in eta Dautzenberg 1927. J. Conchyl. 71:63. This form is characterized by dark transverse lines on a background of light yellow, gray, or brown. The author referred to a figure of Reeve (1850, pi. 7, fig. lOe). 510 PACIFIC SCIENCE, Vol. XXI, October 1967 cinnamonea Menke 1828. Menke Synopsis 76. This form is of cinnamon brown color with irregular longitudinal zones of darker brown. fenestrata Roding 1798. Mus. Boltenianum 34. This form, based by Roding on Figure 502 of Martini, is distinguished from the form cincta by the addition of horizontal lines which cross the vertical lines to create a trellis of quad- rangular meshes. Vanatta (1915), instead of using the reference to Martini for the form fenestrata, substituted that of Tryon (1883, pi. 23, fig. 23), which represents an individual of the form cincta. macleaya Duclos 1835. Monogr. Oliva, pi. 21, figs. 13-16. This form is gray or yellowish gray covered with lines and inconspicuous compact dots sometimes broken by two unbroken transverse bands. rump hi Dautzenberg 1927. J. Conchyl. 71: 66. This form is based on the figure of Rumpf (1705, p. 119, pi. 39, fig. 4). It is ornamented with lines and black spots arranged in the axial plane of the shell. sepulchuralis Lamarck 1811. Ann. Mus. Hist. Nat. 16:312. The comments of Dautzenberg (1927), trans- lated from the French, regarding this form may be of interest. "The name sepulturalis has been borrowed by Lamarck from the old litera- ture. Rumpf (1705) explains that it means sepulchurae or prinsegraaffnis (funeral of a prince). These olive-like shells are ornamented with spots and black lines arranged in a man- ner to represent a theory of persons dressed in grand fashion and following the funeral. How- ever, Lamarck has cited as being his O. sepul- turalis fig. 1, of pi. 365 of the Encyclopedia, on which the pattern is arranged in transverse bands and not longitudinal, as in the figure of Rumpf, while it is the variety b which agrees with the figure. The name sepulchuralis should therefore be reserved for the form with the transverse bands, and we propose for the form with longitudinal swatches the name of rump hi Oliva angustata Marrat 1868. Ann. Mag. Nat. Hist., 4th series, 2:213; Thes. Conchyl., p. 16, pi. 13, %s. 182, 183. Tomlin (1953) stated: "There are two syn- types in the Liverpool Museum labelled from China. Original of fig. 182, 25 mm. X 12 mm.; shell of fig. 183, 26 mm. X mm. These are very young shells of Oliva vidua (Roding).’’ Oliva cana Marrat 1871. Thes. Conchyl., p. 15, pi. 11, fig. 152. Tomlin (1953) stated that the holotype is in the Liverpool Museum from New Guinea, 37 mm. X 15 mm. It is a poor example of O. vidua. Oliva reticulata (Roding 1798). Mus. Bol- tenianum, p. 33, sp. 396. Generations of authors have assigned this species to O. sanguinolenta Lamarck 1811 (Ann. Mus. Hist. Nat. 16:316: Anim. sans Vert. 7:426). In a discussion of references this name must be used. The references of Roding are: 396.10. Porphyria reticulata, Die Netz-dattel. Gmel. V, Oliva sp. 17 ; Martini 2. t. 48, f. 512, 533. 9 St. V oluta oliva var. x was established by Gme- lin (1791) first on Plate 739, Figure 28 of Lister (1685-92), which is probably an O. reticulata of light coloration; second, on Figure 3 of Plate 39 of Rumpf (1705), which is cer- tainly an O. sericea (Roding); and third, on Figures 512 and 513 of Martini (1773), which represent typical O. reticidata. Johnson (1928: 11) assumed the role of first reviser and selected the name O. reticulata (Roding) for this species on the basis that both Roding and Lamarck used the same figures of Martini. Johnson here abandoned the use of the name variegata (Roding). Earlier Johnson (1910:67) had sug- gested that the name Porphyria variegata Rod- ing be accepted for this species. Roding’s refer- ences for this species are: Mus. Boltenianum, p. 33, sp. 393, 8. P. variegata, Die schackigte Dattel. Gmel. V oluta sp. 17, 3; Martini 2, t. 45, f. 478, 479. 24 St. This species was based by Roding (1798) on the variety of V oluta oliva of Gmelin (1791) which includes, first, an O. vidua (Roding) of the form sepulturalis, li The Family Olividae — Burch and Burch Knorr (1768, III, pi. 17, fig. 3); second, an 0. vidua Regenfuss (1758, pi. 1, fig. 2); and third, an O. vidua of the form hi fas data Mar- tini (1773, fig. 474), and on Figures 478 and 479 of Martini, which are O. reticulata of light coloration. Variety a of this Porphyria is based by Roding on Figures 480 and 481 of Martini, which represent O. episcopalis La- marck 1811 of the form lugubris, and upon the figure of Knorr (1768, v, pi. 19, fig. 1), which is an O. vidua of the form sepulturalis orna- mented with a yellow transverse thread upon the middle of the last whorl. The opinion of the majority of students was that such an assemblage did not justify restoration of the name. The fact remains that it is in use by some authors. The design of typical O. reticulata is com- posed of lines so close and condensed that it nearly hides the background of the shell. The last whorl is crossed by two blackish bands, and the columella is a beautiful blood-red. This type has been well figured by Martini (1773, figs. 512, 513), Duclos (1835, pi. 22, fig. 16), Weinkauff (1878, pi. 10, fig. 4), Reeve (1850, pi. 13, fig. 25b), Tryon (1883, pi. 23, fig. 28), and Habe (1966, vol. 2, pi. 27). Habe uses the name O. variegata (Roding) . The name viridescens, borrowed from Mar- tini by Morch (1863), H. and A. Adams (1858), Marrat (1871), and others, cannot be used since this is only a part of a descriptive phrase in the work of Martini. O. pintamella Duclos 1835, which has been regarded by Weinkauff (1878) and Tryon (1883) as a variety of O. sanguinolenta, was figured by Duclos in 1835 (Monogr. Oliva, pi. 33, figs 7 and 8). It is a small, short shell; the edge of the columella is very callous and strongly folded in throughout the length. The aperture is narrow. Duclos (1844) added for the same species two figures in Chenu (Illus. Conchy 1., pi. 35, figs. 9 and 10) which do not agree with the previous ones. Evidently it was these figures that Ducros (1857) considered to be yellow and discolored O. sanguinolenta . The true O. pintamella (figs. 7 and 8) seems to us to be a good species. Marrat (1771) figured it in Thes. Conchyl., pi. 15, figs. 212 and 213. A few of the described color forms of O. reticulata follow. 511 azona Dautzenberg 1927. J. Conchyl. 71:109. This form differs from the typical only by the absence of transverse bands. evania Duclos 1835. Monogr. Oliva, pi. 20, figs. 3 and 4. Reeve (1950, pi. 13, figs. 25a and 25b) ; Marrat (1871, fig. 163) ; Tryon (1883, vol. 5, p. 79, pk 23, fig. 29). Ducros (1857) says with reason that O. evania is but a form of O. sanguinolenta with pale background and strongly banded. It may be added that the pattern is much less closed than that of the form pallida, and that the bands are composed of large isolated spots, sometimes nearly black. pallida Dautzenberg 1927. J. Conchyl. 71:110. This form differs from the typical in that the pattern covers less of the background, giving a lighter aspect to the entire coloration. zigzag Perry 1811. Perry Conchyl., pi. 41, fig. 4. In this form the pattern consists of longitudinal lines disposed in zigzags and isolated from each other without decurrent bands. Oliva rubrolabiata H. Fischer 1902. J. Conchyl. 50:409-410, pi. 8, figs. 12 and 13. Type locality, New Hebrides. A comparison of this species with O. reticu- lata (Roding 1798) would seem logical from form alone, but the folds on the columella are much more numerous, and run the length of the columella to the suture. In addition both the columella and the outer lip are a bright crimson. The body whorl is dark brown, banded with light, close, concentric bands. This species must be comparatively rare. We have seen only a few specimens. There is a set of two in the collection of the California Academy of Sciences (no. 37876). These specimens match the description and figures in all details. Dautzenberg (1927) reported the species from New Caledonia, but we have seen none from this locality. The Burch collection at this time contains two specimens, both from the New Hebrides, which is the type locality. One is from Tasariki, Espiritu Santo, New Hebrides, from black volcanic sand in about 7 m of water (J. R. Bollard, October 1966), the other from 512 PACIFIC SCIENCE, Vol. XXI, October 1967 the north coast of Tanna, New Hebrides, from black volcanic sand (Mrs. H. Dale, August 30, 1966). Oliva tricolor Lamarck 1811. Ann. Mus. Hist. Nat. 16:316. The coloring of this species consists of a profuse mottling of clouded blue and saffron- yellow spots with, in most specimens, a large proportion of green, showing two bands, one around the middle of the shell, and one beneath the sutures. But the most characteristic feature of the species is that the spire is obliquely tes- sellated with black and a slight mixture of red. The aperture is white. This species has the outline of O. reticulata (Roding) and not of O. elegans Lamarck. It has the salmon-colored fasciole, but the color of shell is very different from either. The dark specimens are bluish green with bands of slightly darker shade. The entire shell is spotted with yellow; spire and lip are coarsely marked with brown. Light-colored specimens often have bright yellow and blue spots with the bands obsolate. Such specimens often resemble O. caerulea (Roding) so closely as to be separ- ated only by the violet-colored aperture of the latter. O. philantha Duclos 1835 is a light- colored form approaching O. caerulea in ex- ternal appearance. This species is common from many Indo- Pacific localities. Oliva caerulea (Roding 1798). Mus. Bolten- ianum. Porphyria caerulea. Die himmel- blaue Dattel. Gmel. V, Oliva fp. 17x. Martini 2, t. 48, f. 518. Rumpf t. 39, f. 5. 13 St. Of the three references cited by Roding the one of Martini (1773, pi. 48, fig. 518) is the only one which agrees with the species. That of Gmelin (1791), Voluta oliva var. x, is based on three figures: Lister (1685, pi. 739, fig. 28), which is O. reticulata (Roding 1798); Rumpf (1705, pi. 39, fig. 3), which is O. sericea (Roding 1798) ; and Martini (1773, figs. 512, 513), which is O. reticulata (Roding). The third reference by Roding to Rumpf (1705, pi. 39, fig. 5) represents O. tricolor Lamarck 1811 and O. elegans Lamarck 1811, but cer- tainly not O. caerulea. This species is known in almost all of the literature as O. episcopalis Lamarck 1811. It is unfortunate that we are compelled to accept the O. caeriilea of Roding, but the name has priority and has been extensively used. It was proposed by H. and A. Adams (1858), and Morch (1863). Needless to add, the references to the literature, with few exceptions, are to O. epis- copalis Lamarck. The typical pattern is a shell covered with scattered punctations mingled with a few little black specks, but in certain examples the spots are grouped in such a manner as to form two interrupted bands situated one at the top, the other about the middle of the last whorl. In others the pattern is transformed into two longi- tudinal undulations. The aperture is a deep violet. This is a common species, distributed throughout the Indo-Pacific. We will mention two of the named color forms. luguhris Lamarck 1811. Ann. Mus. Hist. Nat. 16:313. This form differs from the typical by the coloration being darker throughout. The pattern is more marked, and runs into zigzags and large blackish-brown spots. In certain individuals the brown color overruns nearly all the surface, not allowing sight of the white background, which takes the shape of little isolated spaces. emelliodina Duclos 1844. In: Chenu, Illus. Conchyl., pi. 21, figs. 19, 20. According to Ducros (1857) this is a peculiar form of the species. The figure of Duclos repre- sents a shell of small size, short, with a spire very little elongated when compared with most specimens. Oliva atalina Duclos 1835. Monogr. Oliva, pi. 10, figs. 9, 10. Tryon (1883) stated that O. atalina Duclos and O. quersolina Duclos 1835 are discolored specimens of O. caerulea (Roding 1798). Tryon used the name O. episcopalis Lamarck. Ducros (1857), while agreeing that O. quer- solina is a discolored state of O. caerulea, was The Family Olividae — Burch and Burch 513 of the opinion that O. atalina is a distinct species. We share his opinion because, despite the resemblance of the pattern of O. atalina and O . caerulea, the background of the aperture is always white in the first and purple in the second. Oliva ti grina Lamarck 1811. Ann. Mus. Hist. Nat. 16:322. This species is based on Figure 475 of Martini (1773), which represents a shell swollen about the top, with a short spire on which the pattern is composed of numerous greenish-gray punctations, and with a few groups of short brown lines. The band is ornamented with black spots, but most speci- mens are destitute of lines and black spots on the band, and the background is tawny gray instead of white. Meuschen (1787:370) created a Cylindrus tigrinum. This work has been declared invalid but, in any event, Meuschen supported his species by three figures representing olives, of which none is determinable. The name has no meaning. Marrat (1871) replaced O. ti grina Lamarck with O. holos erica Martini, a substitution which cannot be accepted since the nomenclature used in the first volume of the Conchylien Cabinet is only occasionally binomial. Furthermore, in the present case, the words Cylinder holosericus are used as part of a descriptive phrase. Finally, Martini’s species is based on four figures of which only one (Fig. 475) concerns the species in question. According to Ducros (1857), who examined the type of Duclos, O. othonia Duclos 1844 is a young specimen of O. ti grina Lamarck. It is impossible to identify in a satisfactory manner O. glandiformis Lamarck 1811. For some authors, this is a variety of O. ti grina. The description is quite insufficient, and is accompanied by no reference. The variety associated by Lamarck, girol Adanson 1757, is an entirely different species and already has been named O. flammulata by Lamarck himself. We mention one of the named color forms frequently used by authors. fallax Johnson 1910. Nautilus 2 4:64. In this form the bands are suffused and cover the entire shell. We noted Johnson’s type in the Academy of Natural Sciences of Philadelphia. It is the common all-black color form. It may be mentioned that this form is often confused with the black form of O. vidua (Roding), but the shell is much less cylindrical in outline, and the sutural callus is less elevated. Oliva elegans Lamarck 1811. Ann. Mus. Hist. Nat. 16:312. The shell is olive or brownish yellow closely covered with zigzag lines or punctations or both, varying from chocolate to nearly black. The fasciole is salmon-colored. This species seems to be confusing to many, with other species being assigned to it in error. It is separable from O. ti grina Lamarck 1811 and O. tricolor Lamarck 1811 by the shorter and more tumid growth, and the erect callous pro- duction of the last whorl upon the spire, which is proportionally depressed. Although this species is smaller, it has the more cylindrical form and elevated sutural callus of O. vidua (Roding 1798). Light-colored specimens with the bright salmon-colored fasciole resemble in a general way O. reticulata (Roding 1798). It also has a range in color similar to that of the latter species, and lacks the dark fulvous and melanic forms of O. vidua (Roding). Small dark specimens are often very close to specimens referable to O. funebralis Lamarck 1811. This species is common, with a wide distribution throughout the Indo-Pacific. Oliva lecoquiana Ducros 1857. Revue Critique, p. 43, pk 2, figs. a-c. The shell is banded with chocolate-colored triangular markings, as in O. elegans Lamarck 1811. The fasciole is stained with saffron. The form is somewhat more bulbous, and the interior of the aperture is violaceous. O. similis Marrat 1867 is a minor form of O. lecoquiana. O. calosoma Marrat 1871 (not Duclos 1835) is a small form. We have specimens that fit the description of O. lecoquiana from Madagascar, Fiji Islands, and other localities. 514 Oliva calosoma Duclos 1835. Monogr. Oliva , pi. 26, figs. 1, 2. Tryon (1883) describes the species as "Pure white, or with slight indications of three bands composed of occasional triangular brown mark- ings. Length 27 mm. China.” Weinkauff (1878) accepted the species. Oliva bulbiformis Duclos 1835. Monogr. Oliva, pi. 27, fig. 10. The shell is short and very bulbous, colored as O. elegans Lamarck 1811, but the interior of the aperture is chocolate brown. Reeve (1850) figured and described O. bulbiformis, but mentioned O. dactyliola Duclos 1835 and O. caroliniana Duclos in the synonymy. We do not accept this (see below). This is a common species from many localities throughout the Indo-Pacific. Oliva dactyliola Duclos 1835. Monogr. Oliva, pi. 27, figs. 5-8. We have eliminated from the synonymy Figure 9 of Duclos. It presents no well-defined character, and Ducros (1857), who studied the shell represented by this figure, stated that it is O. bulbiformis Duclos and not O. dactyliola. Reeve (1850) supposed that O. dactyliola could be a synonym of O. bulbiformis Duclos, and Tryon (1883) made it a variety of O. funebralis Lamarck, to which he compared, moreover, O. picta Reeve and, with doubt, O. blanda Marrat. Finally, Johnson (1910), although maintaining that O. dactyliola is a distinct species, said that it appears to be intermediate between O. funebralis Lamarck and O. bulbi- formis Duclos, having the spire of the first and the form of the second. Sowerby (1900) accepted the species with the statement that he had specimens from Pondoland and also from Cebu, Philippines. We have specimens from the Philippines, New Guinea, Indonesia, New Caledonia, and elsewhere that agree well with the Figures 5-8 of O. dactyliola Duclos. Oliva funebralis Lamarck 1811. Ann. Mus. Hist. Nat. 16:332. The shell is more cylindrical than O . bulbi- formis Duclos. It differs from O. lecoquiana PACIFIC SCIENCE, Vol. XXI, October 1967 Ducros in that the lower band of the fasciole is deeply strigated with chocolate. The aperture is slightly bluish or chocolate. Johnson (1910) stated that this species seems to occupy an intermediate position between O. tigrina Lamarck and O. elegans Lamarck. It is beautifully illustrated by Marrat (1871, pi. 11, figs. 143-148) under the name leucostoma Duclos 1835 and labradorensis (Roding 1798) . The figure attributed by Roding to Lister (1685, tab. 731, fig. 20) is unrecognizable, and so labradorensis can be dropped. The narrower form suggests a relationship to O. mustellina Lamarck, while the broader form shows a ten- dency toward the more inflated O. dactyliola Duclos. Johnson (1915) discussed the species again, stating that it is extremely variable, with limits which are difficult to define. Reeve (1850, pi. 7, sp. 10) figured a specimen of this species which he considered a form of O. vidua (Roding). Tomlin (1953) considered O . leu- costoma Duclos to be separable from O. fune- bralis, but we are disposed to place it in the synonymy. Some authors have thought the following species of Marrat (1871) to be valid, but we think that they belong in the synonymy of O. funebralis : O. clara, O. pro- pin qua, O. lutea, and O. inornata. It is probable that further study will indicate that this is one entity including a number of related forms. Oliva similis Marrat 1867. Ann. Mag. Nat. Hist., 3rd series, 20:215. Thes. Conchyl. pi. 14, figs. 205, 207. Tomlin’s comments on the shells in the Liverpool Museum follow: "Four (not types). Eastern seas. Max. 35 mm X 16 mm, Min. 31 mm X 14 mm. Av. 32.2 mm X 15 mm. I do not think the shells numbered 206 and 207 are the originals of these figures. None of the four seem to fit any of the three figures that Marrat gives of O. similis though they are that species right enough.” Weinkauff (1878, pp. 27, 7, 7, 11, sp. 10) accepted the species. At first we thought to place this species in the synonymy of O. bulbiformis Duclos 1835 but, after seeing more specimen material, we are now disposed to admit the species. A few sets noted are: Acad. Nat. Sci. Philadelphia, no. 128327 and 15853 The Family Olividae — Burch and Burch from Ceylon, and 104736 from New Guinea; Am. Mus. Natl. Hist. no. 48428 from Singa- pore. The specimens are white or cream in color, the pattern is somewhat as in O. scripta Lamarck, the size about 32 mm. All seem to be more cylindrical than O. bulbijormis. All have a violaceous aperture. Oliva laevis Marrat 1871. Thes. Conchyl. p. 4, pi. 20, figs. 330-331. Tomlin (1953) states that the holotype only is in the Liverpool Museum. It is from the Sey- chelles, 18 mm X 7 mm. A very young shell of O. similis Marrat (spelled laevis on page 26, laeve on caption to plate). Weinkauff (1878) thought it to be a juvenile O. elegans or O. tigrina. The confusion here is apparent. Oliva caroliniana Duclos 1835. Monogr. Oliva, pi. 19, figs. 5, 8. This species is close to O. mustellina Lamarck 1811. The shells are more bulbous and the spire more exserted. Weinkauff (1878) and Marrat (1871) accepted the species. Sets so labelled in the major collections are of interest. Acad. Nat. Sci. Philadelphia, no. 15855 from Singapore, and set no. 104782 from Mauritius are the same. Oliva mustellina Lamarck 1811. Ann. Mus. Hist. Nat. 16:316. The shell is cylindrical, the aperture long and narrow, the spire short. The color is a pale yellow covered with light chestnut figurations. The interior of the aperture is a deep violet. Many specimens in the large museum collections seem to be assigned to this species in error. It is surprising to note so many assigned to the quite different O. elegans Lamarck. Variation in one biological entity or grouping of several related species are possibilities to consider in this as well as other species in this family. Specimens from Japan, China, Singapore, India, and the Philippines seem to be typical. We place the following species tentatively in the synonymy. Oliva pacifica Marrat 1871. Thes. Conchyl., p.15, pi. 11, fig. 151. Some authors have accepted O. pacifica as valid. Weinkauff (1878) accepted the species and placed O. arctata 515 Marrat in the synonymy. Shell labelled from China. Oliva arctata Marrat 1871. Thes. Conchyl., sp. 99, p. 20, figs. 229, 230. The holotype only is in the Liverpool Museum, labelled China Sea, 25 mm X 10 mm- Tomlin (1953) stated that this is obviously the young of O. pacifica Marrat. Oliva ponder os a Duclos 1835. Monogr. Oliva, pi. 15, figs. 8, 9. The shell is cream-white, slightly colored with obscure bluish or violet short interrupted streaks, and a few brown reticulations. The columella and interior of the aperture is car- nelian-white, sometimes tinged with flesh-pink. The shell is thick and stout, with the spire but little exserted. The last whorl is more or less produced toward the apex. Long sets from Mauritius, Seychelles, Maidive Islands, and other localities of the Indian Ocean produce shells with both white and light-salmon aper- tures. The high callosity above the suture becomes less pronounced in the more juvenile specimens. The pronounced growth of the last whorl in all adults of this form, and the consis- tently more obese shape leads us to admit the species. Other species are considered distinct on grounds much less apparent. The species was well illustrated by Reeve (1850), Weinkauff (1878), and Marrat (1871). Oliva rufula Duclos 1835. Monogr. Oliva, pi. 19, figs. 9, 10. The shell is fawn colored, crossed diagonally or transversely by dark-chestnut bands formed by the coalescence of trigonal markings. The aperture is white. This species seems to maintain both form and pattern consistently, showing no noticeable variations. It is distinctive and easily recognized. This is a fairly scarce shell. Most specimens come from the Philippines. Oliva scripta Lamarck 1811. Ann. Mus. Hist. Nat. 16:315. The shell is yellowish brown with pale chestnut zigzag markings, and two bands of 516 PACIFIC SCIENCE, Vol. XXI, October 1967 brown waved characters. The shell is cylindri- cally ovate with a short spire. All authors agree about the similarities of this species with some forms from the West Indies. At times it is difficult to separate O. scrip ta from specimens of the West Indies called O. jamaic- ensis by Marrat (1871) and O. caribbensis Dali and Simpson 1901. The subcylindrical form a little bulging at the top of the last whorl, the short spire, the pattern and color are identical. The pattern of O. say ana Ravenel 1834 from the western Atlantic resembles that of O. scripta, but the form is more elongate, the spire higher, and the last whorl is not swollen above. It is our opinion that O. scripta may be easily separable from the above forms by the fact that the produced posterior of the last whorl gives the shell more the shape of O. mustellina. The species was well illustrated by Reeve (1850) and by Tryon (1883). Kuroda and Habe (1952) listed the species from Japan. We have this species from China, Ceylon, the Moluccas, and from Thailand. We also have sets from the Cook Islands that seem to be identical. Oliva an mil at a (Gmelin 1791). Syst. Nat., ed. 13, p. 3441. This species is listed widely in the literature as O. emicator (Meuschen 1787). Meuschen’s names have been declared invalid by the In- ternational Commission. The name O. guttata Lamarck 1811 has also been used by many authors. It is unfortunate that the first available name for this species seems to be O. annulata. This name was given to an abnormal ringed shell. The color is entirely white and the shell is characterized by a protruding ring encircling the middle of the last whorl. The ring is not uncommon in this species. We have many specimens of all color forms with this character. We know the shell from the figure of Lister (1685), and that of Martini (1773), the last having been copied by Wood (1828) in the Index Testalogicus. Vanatta (1915) cited as representing typical O. annulata Figure b of Plate 16 of Duclos, and Figure 60 of Plate 5 of Marrat. Both figures represent a yellow-orange shell, without a ring in the middle of the last whorl. This is com- pletely wrong, since it is especially the ring which characterizes the O. annulata of Gmelin. j O. leucophaea Lamarck 1811 is an absolute synonym. It is odd that Lamarck substituted the name for annulata, even though he cited it among the references to his species. Some of the named color forms follow. amethystina (Roding 1798). Mus. Bolteni- anum, p. 35. This should perhaps be recognized, as this repre- sents the common color which is ornamented with round, purple spots, fairly regularly spaced on a clear flesh-colored background. alba Sowerby 1825. Cat. Tankerville, p. 86. The shell is entirely white inside and out. carnicolor Dautzenberg 1927. J. Conchyl. 71:22. The background is yellowish white without spots. The dorsal region of the last whorl is reddish pink. intricata Dautzenberg 1927. J. Conchyl. 71:23. This form differs in the pattern of a compact confusion of little brownish lines sprinkled throughout with black dots. The white back- ground appears between the meshes of the net- work in the shape of little gaps more or less triangular. mantichora Duclos 1835. Monogr. Oliva , pi. 15, figs. 7, 8. The form mantichora presents at the top of the last whorl an angle more or less pronounced. The pattern and coloration resemble those of the form intricata, but there are also examples of which the background is more open. Marrat (1871, pi. 5, fig. 29) shows an individual in which the keel, situated lower down, tends to approach that of the typical annulata, but this keel is blunt and does not have the aspect of a ring. nebulosa Dautzenberg 1927. J. Conchyl. 71:22. This form was figured by Reeve (1850, pi. 14, species 30g) as O. leucophaea Lamarck. The The Family Olividae — Burch and Burch spots have the character of scattered triangular blotches, and the ground is of a ruddy tinge. O. annulata (Gmelin) is a common species distributed throughout the Indo-Pacific. We have sets of mixed and intergrading forms from Tahiti, Solomon Islands, Guam, New Guinea, Marquesas, Mauritius, Admiralty Islands, and many localities in the Philippines. Oliva carneola (Gmelin 1791). Syst. Nat., ed. 13, p. 3443. The figure of Martini (1773, fig. 495) upon which Gmelin based this species is very poor, but one can still recognize that it concerns a shell with reddish background crossed on the last whorl by bluish-gray bands running down- ward. This coloration may be regarded as typical. O. carneola is either short or elongate, either swollen or cylindrical; the spire is constantly covered with a thick callosity which completely hides the superior whorls, that of the last whorl alone being free. The pattern and coloration give this species a great number of modifications, and change so often during growth that it is impossible to place the shells into determined color forms. We will mention a few of the most characteristic. We have avoided discussion of generic names in Oliva. The differences in radula or anatomy are slight, and we see no systematic advantage in the recognition of such names in this group. For example, some authors place this species in the genus Galeola Gray 1858. This is a common species throughout the Indo-Pacific. Inasmuch as many or all of the described color forms appear in most of a large series from all localities, it seems futile to attempt a division of them. A few of the named color forms follow. adspersa Dautzenberg 1927. J. Conchyl. 71:9. The shell is irregularly sprinkled with numerous small, white, rather conspicuous triangular flecks. bizonalis Dautzenberg 1927. J. Conchyl. 71:8. This shell has a red background crossed in the middle of the last whorl by two white bands rather large and close. (See Duclos 1844, pi. 28, fig. 13.) 517 candidula Dautzenberg 1927. J. Conchyl. 71:8. The bands are barely visible at the beginning of the last whorl and disappear completely there- after. (See Duclos 1844, pi. 28, figs. 12, 16.) coccinata Dautzenberg 1927. J. Conchyl. 71:8. The shell is a nearly uniform red, with bands visible. (See Duclos 1844, pi. 28, fig. 8.) trichroma Dautzenberg 1927. J. Conchyl. 71:9. The shell differs from unizonalis by the presence at the summit of the last whorl of a deep-purple band at the suture which imparts to the shell a three-colored aspect. tmizonalis Dautzenberg 1927. J. Conchyl. 71:9. The shell has a red background crossed by one large white band in the middle of the last whorl. (See Duclos 1844, pi. 28, figs. 6, 14.) Oliva athenia Duclos 1835. Monogr. Oliva , pi. 26, figs. 17, 18 (excl. 19, 20). O. athenia is a well defined species, charac- terized by the squat form, the spire nearly flat and mucronated in the center. Nevertheless, it has often been misunderstood. Duclos himself, after correctly representing it on Plate 28, Figures 17, 18, has added under the same name Figures 19 and 20, which are of O. mucronata Marrat and O. fab a Marrat. Tryon (1883) considered it a synonym of O. sidelia Duclos 1835. A typical O. athenia is ornamented with longitudinal blotches standing out clearly in zigzags on a tawny background. These blotches are either separated or approximate. In certain examples they stand out on a dotted or ob- scurely reticulated background. We have speci- mens from New Caledonia, Andaman Islands, Fiji Islands, and Australia. This species has been the source of much disagreement. Reeve (1850) placed it in the synonymy of O. carneola (Gmelin 1791). Johnson (1910) said under his discussion of O. mustellina Duclos, "O. athenia Duclos re- sembles this species in miniature.” Johnson’s comment might be in order if one failed to 518 PACIFIC SCIENCE, VoL XXI, October 1967 note the spire. We think that it is distinct. Most of our specimens were received labelled O. faba Marrat 1867, a name which we now place in the synonymy. Oliva sidelia Duclos 1835. Monogr. Oliva , pi. 19, figs. 1, 2. Duclos (1844, p. 23, pi. 21, figs. 1, 2). Ducros (1857, p. 69). Marrat (1871, pi. 15, figs. 231, 232) (a copy of the figures of Duclos). Tryon (1883, pi. 33, figs. 27, 44) (a copy of the figures of Duclos) . Tryon (1883) united, under the name O. sidelia, O. volvaroides Duclos 1835, O. lepida Duclos 1835, and O. tod osina Duclos 1835. We share his opinion for sidelia, volvaroides, lepida, and todosina, but it is our opinion that athenia, mucronata, and faba constitute a distinct species, more squat, with a more depressed spire, and closely mucronated in the middle. The pattern is much plainer and darker. Finally, sidelia, vol- varoides, lepida, and todosina have the last whorl separated by a suture greatly canaliculated, while in athenia, mucronata, and faba the suture of the last whorl is deep but very tight. We find ourselves presented with two distinct species, sidelia and athenia. It is to be regretted that the name sidelia, which was created for a young shell, should have been selected for this species. O. volvaroides, O. lepida, and O. todosina ap- ply to adult specimens. The type of O. sidelia is a shell 10 mm in length, white, with faint violet undulations, and having on the dorsal re- gion a large brown spot which occupies nearly the whole length of the last whorl, except for a white zone close below the suture. This coloration must be infrequent, since Marrat (1871) and Tryon (1883) copied the figures of Duclos. Weinkauff (1878) gave three figures of O. sidelia, but this interpretation is difficult, the author admitting that they are not elongated enough, and that the relation between the height and width is not exact. Reeve (1850, pi. 22, sp. 59) figured and described O. volvaroides. Specimens from China, Mauritius, Seychelles, Solomon Islands, and other localities indicate a distribution throughout the Indo-Pacific. A few forms of this species follow. lepida Duclos 1835. Monogr. Oliva, pi. 25, figs. 15 to 20. The shell has a pattern of triangular spots. We have specimens that may fit this form from Mauritius, Seychelles, China, and the Philip- pines. todosina Duclos 1835. Monogr. Oliva, pi. 25, figs. 9, 10. This form was based by Duclos on a shell of which the pattern and especially the middle band of the last whorl are darker. The habitat of California indicated by Duclos is certainly erroneous. volvaroides Duclos 1835. Monogr. Oliva, pi. 25, figs. 11 to 14. The type of O. volvaroides represented by Duclos (pi. 25, figs. 11 and 12) is coffee-and- cream colored with some transverse lines hardly visible. His Figure 13 is of a variety entirely white, and Figure 14 is of a uniform blackish brown which is so similar to Figure 20 (in- scribed as a variety of lepida ) that one might be tempted to believe that these two figures were made from the same shell. While main- taining O. volvaroides as a species, Ducros (1844) said that a study of a great number of individuals from diverse localities might au- thorize a reunion with O. lepida. Oliva tessellata Lamarck 1811. Ann. Mus. Hist. Nat. 16: 320, b. 28. The shell is yellow, spotted with purple. The aperture and columella are deep violet. Dillwyn (1817) and Marrat (1871) took the name O. tigrina (Meuschen 1787) for this species. Meuschen’s names have been declared invalid by the International Commission, but in any event it is an obvious error because no reference cited in the Museum Geversianum re- lates to O. tessellata. O. tessellata varies only in the number of punctations with which the surface is orna- mented. In adult specimens they deviate rarely on the end of the last whorl. The dark purple of the interior is constant. Some authors place this in the genus Neo- cylindrus Fischer 1883, with O. tessellata La- marck as the type. This is a common species distributed through- out the Indo-Pacific. The Family Olividae — Burch and Burch Oliva bulbosa (Roding 1798). Mus. Bolteni- anum, p. 34, Porphyria bulbosa. It is to be regretted that the vagaries of the system permit no escape from abandoning the name O. inflata Lamarck 1811 under which this well known species has been recognized for generations. It is needless to add that al- most all of the following references to the literature are the work of authors who knew the species as O. inflata Lamarck. Both Roding (1798) and Lamarck (1811) refer to the same figures of Martini (Conchyl. Cab. II, tab. 47, figs. 507, 508). These figures represent specimens having undulating longi- tudinal stripes of brown. Hence we must ac- cept this form as typical. The form of this species is very character- istic: swollen gibbous growth, fasciole with a heavy callous ridge which is independent of the columellar plaits. This is the only species of Oliva with this character. Inasmuch as most of the large sets in collec- tions from various localities seem to contain specimens of almost all of the described forms, it seems futile to consider them in great detail. However, a few will be mentioned to give the references. bicingulata Lamarck 1811. Ann. Mus. Hist. Nat. 16:94. bicincta Lamarck 1822. Two revolving dark lines are present. fabagina Lamarck 1811. Ann. Mus. Hist. Nat. 16:94. The bands fuse and cover irregularly the greater portion of the shell. This form was figured by Marrat (1871) as O. crassa Martini. immaculata Vanatta 1915. Nautilus 29:68. This is the white albino form. In some collec- tions this form is labelled O. alba Dillwyn 1817. inflata Lamarck 1811. Ann. Mus. Hist. Nat. 16:310. Specimens with only the small uniform bluish- gray spots. lacertina Quoy and Gaimard 1825. Voy. Uranie, p. 432, pi. 72, figs. 4, 5. This is a peculiarly banded color form of this species. 519 This species is generally distributed through- out the Indo-Pacific. We have large sets from Mozambique, Zanzibar, the Red Sea, Mauritius, Suez, Tanzania, Singapore, Indonesia, etc. Oliva paxillus Reeve 1840. Conchol. Icon., pi. 21, sp. 56, a, b. The shell is cone shaped, yellowish white in color, smooth, with triangular brown markings farming interrupted bands, and spots beneath the sutures and on the fasciole. The interior of the aperture is sometimes two- or three- banded, but this is not a constant character. We are accepting the opinion of the late J. R. le B. Tomlin (1934), a very careful observer, who studied the types of all of the species involved. We also studied the types. All of them are in the British Museum (Natural His- tory). Omogymna was described by von Martens (1897) as a subgenus of Oliva, with O. paxillus Reeve as the type. Tryon (1883) pointed out that O. nitidula Duclos 1835 is preoccupied by nitidula Dillwyn 1817 (not Gmelin 1791)- Oliva ozodona Duclos 1835 is certainly not O. paxillus. On the other hand, Oliva sand- ivichiensis Pease I860 and Oliva thomasi Crosse 1861 are both the same as O . paxillus Reeve. While not common, the species seems to be rather generally distributed. It seems to be the only species of Oliva from the Hawaiian Is- lands. We have specimens from various localities in the Indian Ocean, and from Okinawa, Guam, Marshall Islands, etc. Oliva duclosi Reeve 1850. Conchol. Icon., pi. 19, sp. 44. The shell is orange-yellow, thickly reticulated with olive-brown, the spaces of the network being rather distant and sharply triangular. The columella and interior of the aperture are orange-yellow. It is difficult to know how so many authors have confused this species with O. paxillus Reeve 1850. O. duclosi has been reported from numerous Pacific localities, but it is more com- mon from Tahiti and neighboring islands. According to Ducros (1857), O. natalia 520 PACIFIC SCIENCE, Vol. XXI, October 1967 Duclos 1844 is typical O. duclosi, but blighted and discolored, and his figure is enlarged, too red, and embellished to suit. Ducros studied the type. The same author considers O. stainforthi Reeve 1850 to be a squat specimen of the form lentiginosa Reeve 1850. We wish to mention the two following names as color forms of O. duclosi because both are occasionally seen in the literature. Both are placed in the synonymy. lentiginosa Reeve 1850. Conchol. Icon., pi. 19, sp. 45, a.b. This seems to be no more than a lighter-colored specimen. esiodina Duclos 1844. In Chenu, Illus. Con- chyl., p. 18, pi. 16, figs. 19, 20. This form or species is questionable, but it is obvious that if it is possible to identify O. esiodina, and that it is conspecific with O. du- closi, the name of Duclos has priority. We have failed to locate even one locality record. The shell is very thick with the spire excep- tionally elevated. We know this only from the figures of Duclos, of which one has been re- produced by Tryon (1883). Oliva panniculata Duclos 1835. Monogr. Oliva, pi. 15, figs. 15, 16. In Chenu, Illus. Con- chyl., p. 12, pi. 6, figs. 15, 16. The pattern of O. panniculata represented by the Figures 15 and 16 of Duclos is composed of very fine longitudinal lines of a light fawn color. The last whorl is crossed, a little below the middle, by a narrow band of gray spots, and by another of the same color under the suture. Figure 17 of Duclos represents a form slightly more slender, and Figures 19 and 20 an ex- ample in which the last whorl is crossed by two decurrent gray bands. Some references to this species are: Reeve (1850, pi. 26, sp. 77); Tryon (1883, pi. 32, figs. 24, 25) ; Ducros de St. Germain (1857, p. 64); Marrat (1871, p. 10, figs. 83, 84); Wein- kauff (1878, p. 84, pi. 22, figs. 10, 11, 12). Marrat (1871) figured under the name O. panniculata an exceptionally large shell orna- mented with a highly-colored pattern. Schep- man (1911) listed the species with comments on variation. Melvill and Sykes (1896) re- ported the species from the Andaman Islands. Melvill and Standen (1897) described a variety williamsi as one having the pattern more sharply marked and farther apart as well as by the absence of a band on the last whorl. While not a common species, this has a wide distribution. We have it from Andaman Is- lands, Thailand, Madagascar, Mauritius, and Ceylon. We place the following species in the syn- onymy of O. panniculata: O. ozodona Duclos 1835, Monogr. Oliva, pi. 5, figs. 19, 20. In studying the work of Duclos, we note that he figured the species under discussion on Plate 5, Figures 15 to 18 (which are O. panniculata ), and that Figures 19 and 20 on the same plate are O. ozodona. They are very similar in all characters observable in Duclos’ fine color plate. Oliva concavospira G. B. Sowerby 1914. New mollusca of the genera Pleurotoma ( Sur - cula) , Oliva, and Limopsis from Japan. Ann. Mag. Nat. Hist., 1914, ser. 8, 13: 445, pi. 18, fig. 2. The sunken spire of this species seems to be a constant character. The spire is sunk in a con- cavity below the shoulder of the last whorl. Excellent color figures of the species showing both views were given by Kira (1955, p. 63, pi. 31, fig. 6), and Kira (1962, pi. 32, fig. 6). REFERENCES Adams, Henry, and Arthur Adams. 1858. The Genera of Recent Mollusca; Arranged According to Their Organization. Van Voorst, London. Ad anson, Michel. 1757. Histoire naturelle du Senegal. Paris. Bridgman, F. G. 1905. Note on Oliva tigri- della Duclos. Proc. Malacol. Soc. London 6:187. 1906. Notes on a new variety (var. longispira) of Oliva ispidula. Proc. Malacol. Soc. London 7:195. Bruguiere, J. G. 1789- In: Encyclopedic Methodique. Histoire Naturelle des Vers. Panckoucke, Paris. Crosse, J. C. H. 1861. J. Conchyl. 9:173, pi. 6, figs. 3, 4. Dale, W. H., and C. T. Simpson. 1901. Mol- The Family Olividae — Burch and Burch lusca of Porto Rico. Bull. U.S. Fish Com- mission (1900), p. 391, pi. 56, fig. 9. Dautzenberg, Ph. 1927. Olivides de la Nou- velle Caledonie et de ses dependences. J. Conchy 1. 71:1-72, 103-147. Deshayes, G. P. 1835-45. Histoire naturelle des animaux sans vertebres. Deuxieme edi- tion. Paris. Dillwyn, L. W. 1817. A Descriptive Catalogue of Recent Shells, Arranged According to the Linnaean Method, with Particular Attention to the Synonymy. Arch, London. Dodge, Henry. 1955. A historical revue of the mollusks of Linnaeus. The genera Bulla and Voluta of the class Gastropoda. Bull. Am. Mus. Nat. Hist. 107, art. 1. Duclos, P. L. 1835, 1840. Histoire naturelle . . . de tous les genres de coquilles univalves marins a l’etat vivant et fossile, publiee par monographies. Genre Olive. Paris. 1844-48. Oliva. In: J. C. Chenu, Il- lustrations conchyliologiques. Paris. Ducros de St. Germain. 1857. Revue critique du genre Oliva de Bruguiere. Fischer, H. 1902. Description d’un Oliva nou- veau provenant des Nouvelles-Hebrides. J. de Conchyl. 50:409-410. Ford, John. 1891. Description of new species of Anctus and Oliva. Nautilus 4:134-136. Ford, W. K. 1953. Notes by J. R. le B. Tomlin on Marrat’s species of Oliva. Northwestern Naturalist 1953:442-449. Gmelin, J. F. 1791. Caroli a Linne Systema naturae per regna tria naturae. Editio decima tertia. Vol. 1, part 6, Vermes, pp. 3021- 3910. Leipzig. Gray, J. E. 1858. An attempt to distribute the species of Oliva into natural groups, and to define some of the species. Proc. Zool. Soc. London, part 26, 1858, pp. 38-58. Gualtiere, Nicolo. 1742. Index Testarum Conchyliorum quae adservantur in Museo N. Gualtieri . . . et methodice etc. Florence. Habe, Tadashige. 1961. Coloured Illustrations of the Shells of Japan. Vol. II. — 1965. Ibid. Rev. ed. Vol. II. 1964. Shells of the Western Pacific in Color, Vol. II. Hoikusha, Osaka. 1966. Shells of the World in Color. Vol. II, The Tropical Pacific. 521 Higgins, H. C. 1919. A new variety of Oliva sericea miniacea. Nautilus 33:58. Hirase, Shintaro. 1938. A Collection of Jap- anese Shells with Illustrations in Natural Colors. Sixth ed. Maruzen, Tokyo. 1951. An Illustrated Handbook of Shells from the Japanese Islands and Ad- jacent Territory. Revised and enlarged by Isao Taki. Johnson, Charles W. 1910-1911. Some notes on the Olividae. Nautilus 24:49-51, 64-68, 121-124. 1915. Further notes on the Olividae. Nautilus 28:97-103, 114-116. 1928. A review of certain species of the Olividae. Nautilus 42:6-13. Kira, Tetsuaki. 1955. Coloured Illustrations of the Shells of Japan. Vol. I. Hoikusha, Osaka. 1962. Shells of the Western Pacific in Color. Vol. I. Knorr, G. W. 1760-73. Les delices des yeux et de l’esprit, ou collection generate des dif- ferentes especes de Coquillages que la Mer referme, etc. Nuremberg. Kuroda, T., and T. Habe. 1952. Checklist and Bibliography of the Recent Mollusca of Japan. L. W. Stach, Tokyo. Lamarck, J. B. P. A. de M. de. 1811. [Suite de la] Determination des especes de mollus- ques testaces: continuation du genre Ovule, Tarriere, Ancillaire et Olive. Ann. Mus. Hist. Nat. (Paris) 16 (for 1810): pp. 89- 114, 300-328 (Jan.-Mar., 1811). 1815-22. Histoire naturelle des ani- maux sans vertebres. Paris. 18 16. Tableau encyclopedique et methodique des trois regnes de la nature. Vol. 3, pis. 361-368. Paris. Link, H. F. 1807. Beschreibung der Naturalien- Sammlung der Universitat zu Rostock, etc. Rostock. Linnaeus, C. 1758. Systema naturae per regna tria naturae. Editio decima, reformata. Vol. 1, Regnum Animale, 824 pp. Holmiae. 1766-67. Ibid. Editio duodecima, re- formata. Vol. 1, Regnum Animale, part 1, pp. 1-532 (1766); part 2, pp. 553-1327 (1767). Holmiae. Lister, M. 1685-92. Historiae sive Synopsis methodicae Conchyliorum etc. London. 522 PACIFIC SCIENCE, Vol. XXI, October 1967 Marrat, F. P. 1867. On some new species of Oliva and a new species of Trivia. Ann. Mag. Nat. Hist., 3rd series, 2:214. 1868. Observations on some proposed new species of Olivia, 1. Ibid., 4th series, 2:212. — - 1870-71. Oliva Bruguiere. In: G. B. Sowerby, Thesaurus Conchyliorum. Vol. 4, pis. 328, 329-41, 342-51 {Oliva, pis. 1-25). Martens, E. C. von. 1897. Conchologische Miscellen. Archiv f. Naturgesch., Bnd. 1, Heft 2:157. Martini, F. H. W. 1768-77. Neues Systema- tisches Conchylien-Cabinet. Vols. I— III. Nu- remberg. Melvill, J. C., and E. R. Sykes. 1896. Notes on a collection of shells from the Andaman Islands, with descriptions of new species. Proc. Malacol. Soc. London 2:166. — and Robert Standen. 1897. Notes on a collection of shells from Lifu and Uvea, Loyalty Islands. J. of Conchol. 8:404. Menke, C. T. 1828. Synopsis molluscorum generum omnium et specierum earum, quae in Museo Menkeano adservantur etc. Pyr- mont. Meuschen, F. C. 1787. Museum Geversianum, sive index rerum naturalium . . . comparavit A. Gevers publice distrahendam. Roterodami. Morch, Otto A. L. 1863. Catalogus Conchy- liorum quae reliquit . . . N. C. M. Lassen, etc. Hafniae. Olsson, Axel A., and S. Peter Dance. 1966. The Linnaean olives. Bull. Am. Paleol. 50 (227). Pease, William Harper. I860. Description of new species of Mollusca from the Sand- wich Islands. Proc. Zool. Soc. London I860: 145. Philippi, R. A. [1842] 1845-51. Abbildungen und Beschreibungen neuer oder wenig ge- kannter Conchylien. Cassel. Quoy, J. R. C, and J. P. Gaimard. 1825. Voy. Uranie et Phys., Zook, p. 432, pi. 72, figs. 4, 5. Paris. Ravenel, Edmund, 1834. Catalogue of Recent Shells, p. 19. E. Ravenel. Reeve, Lovell. 1850. Conchologia Iconica: or illustrations of the shells of molluscous ani- mals. Monograph of the genus Oliva, pis. 1- 30. London. Regenfuss, F. M. 1758. Auserlesne Schnecken Muscheln und andre Schaalthiere. . . . Copen- hagen. Roding, P. F. 1798. Museum Boltenianum. J. C. Trappii, Hamburg. Rumpf, Georg E. 1705. D’Amboinsche Rari- teitkamer, etc. Amsterdam. Schepman, M. M. 1911. The Prosobranchia of the Siboga Expedition. Leyden. Sowerby, G. B. (First of the name.) 1825. A catalogue of the shells contained in the col- lection of the late Earl of Tankerville. Lon- don. Sowerby, G. B. 1900. On some marine shells from Pondoland and the Kowie, with de- scriptions of seventeen new species. Proc. Malacol. Soc. London 4:3. Sowerby, G. B. 1914. New Mollusca of the genera Pleurotoma (Surcula) , Oliva, and Limopsis from Japan. Ann. Mag. Nat. Hist. 1914, ser. 8, 13:445, pi. 18, fig. 2. Tomlin, J. R. le B. 1934. Note on Omogymna Martens. Nautilus 48:70. 1953. In: W. K. Ford, Notes by J. R. le B. Tomlin on Marrat’s species of Oliva. Northwestern Naturalist 1953:442-449. Tryon, G. W., Jr. 1883. Manual of Conchol- ogy. Ser. 1, vol. 5. Philadelphia. Vanatta, E. G. 1915. Notes on Oliva. Nau- tilus 29:67-72. Weinkauff, H. C. 1878. Systematisches Con- chylien-Cabinet. Bnd. v, abth. 1. Die gattung Oliva. Nuremberg. Wood, W. 1825. Index Testalogicus; or a catalogue of shells, British and foreign, etc. London. 1828. Supplement to the Index Testa- logicus. London. Revision of the Genus Pandanus Stickman, Part 23 Three Australian Species of Pandanus Harold St. John1 The section Micro stigma of the genus Pan- danus has but three representatives in Australia. P. de-Lestangii Martelli was the first of these to be discovered and described. P. adscendens St. John belongs in the large section Pandanus. Like many others of its species, this new one with smooth-sided pha- langes has its habitat on the marine littoral. P. darwinensis St. John was described earlier and the details of its phalanges were given. Now habit photos are at hand and they show vegetative and fruiting structures. Pandanus de-Lestangii Martelli, Roy. Soc. Queensl. Proc. 38(5):57-58, pi. XI, 1926 (sect. Microstigma') . P. aquaticus F. Muell., Kew J. Bot. 8:324, 1856 (nomen provisorium) ; Fragm. Phytog. Austral. 5:40, 1865; and 8:220, 1874; Bentham, FI. Austral. 7:149, 1878; Warburg, Engler’s Pflanzenreich IV, 9:85, 1900; S. T. Blake, Austral. J. Bot. 2(1):130— 132, pi. 7, fig. 3, 1954. Figs. 240 and 241 diagnosis of holotype: Small and tree- like, forming dense clumps; trunk to 5 m in height, near the top 2-3 -branched; prop roots numerous; each soon producing a new stem; leaves 1.8-2. 7 m long, 7.5-8 cm wide near the base, 5.3 cm wide at the middle, bluish green, drooping, coriaceous, broadly channeled above the midrib, with 2 lateral pleats, at mid- section with 62 parallel secondary veins in each side, throughout the lower side the tertiary cross veins conspicuous, forming long oblong meshes, the blade sword-shaped and from the base gradually tapering to the trigonous subulate unarmed apex which is about 15 cm long, this at 10 cm down 3.5 mm wide; the base amplexi- caul and unarmed, but beginning about 6 cm up 1 B. P. Bishop Museum, Honolulu, Hawaii 96819. Manuscript received January 9, 1963. the margins with prickles 2. 5-3. 5 mm long, 5-12 mm apart, slender arcuate subulate, appressed ascending, reddish tipped; the midrib below in lower and outer thirds unarmed; at midsection the margins with prickles 1.5-3. 5 mm long, 10-17 mm apart, arcuate subulate, appressed ascending; the midrib below narrow, sharp, with prickles 2-3 mm long, 18-35 mm apart, slender subulate, closely appressed ascending; the apex almost unarmed; pistillate plant blooming in late October or early No- vember, with 1-2 terminal syncarps; when ripe the syncarps 10-13 cm in diameter, broadly ellipsoid, 3-sided, green, at maturity the core in a few days shrinks to a remnant 7 cm long, 2.5 cm in diameter, and the drupes fall in a mass; drupes very numerous, 31-35 mm long, the abundant 1 -celled ones 7-11 mm wide, 6-8 mm thick, narrowly oblanceoloid, upper lA free, 5-6-angled, the sides smooth, somewhat shiny, gently curving, the apex rounded pyramidal; stigma 1.5-2 mm long, broadly ellipsoid, creased, flush, oblique, excentric, brown, papillose; drupes with 1 cell are the normal, for on the holotype (fi) there are 136 such, and 21 with 2 cells, and 1 short basal one with 3 cells, while in the isotypic specimen (bri) there are 173 with 1 cell, to 32 with 2 cells, and 1 short basal one with 3 cells. Martelli recorded ones with 4 or 5 cells. Of the 2-celled ones, 6 appeared to have a third stigma, but this was actually a corky scar, and the fruits had only 2 cells. The 2-celled drupes are 10-14 mm wide, the apex shallowly lobed, the cleft 1-3 mm deep, the stigmas 0.8-1 mm long, ellipsoid to obovate, horizontal or oblique, centripetal; the single 3-celled drupe (a short basal, asymmetric one) 26 mm long, 14 mm wide, 12 mm thick, deltoid-oblanceoloid, the stigmas centripetal, placed in a triangle; the 1 -celled drupes with the endocarp central, bony, pale, the lateral walls 0.8-1. 3 mm thick, the apex subtruncate; seed subcuneate-barrel-shaped, 7-8 mm long; the 523 Fig. 240. Pandanus de-Lestangii Martelli, from isotype, a, b, 1-celled drupes, lateral view, X 1; c, drupe, longitudinal median section, X 1; d, e, drupes, apical view, X 1 \f,g, drupe apex, oblique view, X 4; h, i, 2- celled drupes, lateral view, X 1 ; j, drupe, longitudinal median section, X 1 ; k, l, drupes, apical view, X 1 ; m, drupe, apical view, X 4; n, 2-celled drupe, lateral view, X 1; o, drupe, longitudinal median section, X 1; p, drupe, apical view, X 1; q, 2-celled drupe, with corky scar imitating a third stigma, apical view, X 4; r, 3- celled drupe, transverse section, X 1; s, leaf base, lower side, X 1; /, leaf middle, lower side, X 1; u, leaf apex, lower side, X 1. 524 3 Cm, io cm. 525 Page 400: Revision of Pandanus , 23 — St. John Fig. 24l. Pandanus de-Lestangii Martelli, from type locality, 6 Nov. 1926, de Lestang. a, Staminate in- florescence, X y23 b, fascicle of stamens, X 10. io mm. 526 PACIFIC SCIENCE, Vol. XXI, October 1967 2-celled drupes with the seeds median, similar but slightly oblique and 4 mm in diameter; apical mesocarp with numerous transverse, pale membranes; basal mesocarp fibrous and fleshy. description of staminate plants: Grow- ing mingled with the pistillate plants; plants blooming in October and November, bearing 1-2 staminate inflorescences, these with nu- merous pale leafy bracts, the median ones 31 cm long, 3 cm wide, ligulate, acute, firm, foliaceous, veiny, the middle and upper margins with ascending prickles 0.2-0. 3 mm long, about 1 mm apart; spikes 3-6 cm long, 12-15 mm in diameter, finger-like, dense; fascicles of stamens about 15 mm long, divergent, the common filament base 7-10 mm long, bearing 4-7 stamens; free filament tips 0.5-2 mm long; anthers 2-2.9 mm long, oblong, bearing a subulate projection of the connective 0.5- 0.6 mm long. holotype: "Australia; growing under palms along perennial streams about 200 miles south- west of Burketown (Burke District), North- west Western Queensland,” Albert de Lestang (fi) ! Isotypes (bri, k) ! The label on the holotypic specimen has the additional data: "abundant, February 1925. The nuts sent although green are fully grown. These nuts are the favourite food of two species of Turtle which abound in the streams where this pandanus grows.” specimens examined: Australia, same data as above, but 6 Nov. 1926 (staminate) (bri) ; and ditto, 1927 (fi, k). discussion: P. de-Lestangii Martelli is a clearly distinct species. The collector, A. de Lestang, was an amateur naturalist who gathered abundant material and recorded good data. Drawing his description and figures from this material, Martelli published this easily recog- nizable species. He described the drupes as with 1 cell, or 2-3, or rarely 4-5 in one series. His illustrations show 1— 2-3-celled drupes in lateral and apical views, and 1— 2-celled ones in longi- tudinal median section. The holotype (fi) has been studied, as well as the abundant isotype (bri) ; together these contain 309 drupes that are 1 -celled, 53 of the 2-celled, and 2 of the 3-celied. On the angular shoulders leading to the apex of the broader drupes there are often pale, corky scars, very similar to stigmas, especially if the apex is partly eroded. For instance, when the writer first carefully sorted the drupes of the isotype, he separated 7 as with 3 cells, judged by the apparent stigmas. Later, by sectioning some, and by comparison, it was discovered that 6 of them had the body only 2-lobed and in section 2-celled, 2-seeded. In each case they had two stigmas in a line, and the third spot was actually a corky scar, not a stigma. The single remaining one was truly 3 -celled, but it was a shorter, asymmetric, basal one with the three stigmas in a triangle and centripetal. Count Martelli assigned P. de-Lestangii to section Hombronia , as its only representative on the Australian continent. On the contrary, it is now clear that when the stigmas are 2, they are in line but centripetaliy directed, flush, and elliptic to obovate or suborbicular. When 3, the stigmas are centripetal. In true Hombronia the several stigmas are arranged in a line or in several parallel lines, with the stigmas like flaps or teeth directed laterally at right angles to the line of carpels. In structure the fruits of P. de-Lestangii are quite at variance with this section. The species is here reassigned to the section Microstigma . A. de Lestang in later observations (in litt. ad W. D. Francis, 30/9/42) stated that the species succeeds well in cultivation as an orna- mental or for hedges, even in dry ground. "In the wild state thrives in bog and shallow water, loving best the fringe of deep pools where it anchor [s] itself with props extending to the bottom of the deepest water. In spring displays long spikes of yellowish flowers, male ses- sile, . . .” On his staminate sheet is his letter with many details, including: "Each grown tree blooms late in October and early November carrying one or two male inflorescences but only the older trees bear syncarps rarely more than two. Both male flowers and syncarps grow simultaneously on neighbouring limbs, and the specimens forwarded were from one tree.” — He seems to state that the trees are monoecious, but no such species is known. It is quite possible that he observed interlacing branches • from ad- jacent staminate and pistillate plants. Until Page 402: Revision of Pandanus, 23 — St. John 527 proven to the contrary, it will be assumed that this species is dioecious, like all other known species of Pandai7us. He continues: "When a syncarp is fully grown (if allowed) although still green, the balloon-like oval-shaped stem shrinks to nothing within a few days, as soon as the stem [or core] begins to shrink the drupes fall down, slipping in a bunch. "Thousands of White Cockatoos ( Cacatua galerita) systematically comb the Pandanus for syncarps, beginning in February, they tear down each drupe in quest of a kind of fly larvae which, I think, are solely associated with this fruit. The greater part of the drupes fall in the water below where herds of turtles gluttonously swallow whole the falling drupes; those falling upon the banks are not lost either, for when all the Pandanus are clean of syncarps the cockatoos search the ground carefully for the dry nuts and with their powerful beak crush and extract the edible parts.” Consideration must be given to P. aquaticus F. MuelL, which in 1856 was published as a provisional name, but in 1865 and 1874 was validated as a species. In these accounts von Mueller stated that the plant lacked aerial roots, was smaller and more slender, and had separate drupes. He gave no locality and cited no speci- mens. Legally P. aquaticus F. Muell. is valid. The holotype in Melbourne was studied in 1958. It is a single leaf 79 cm long, 3.7 cm wide. It was collected in December 1855 on the Upper Victoria River, labeled with the name von Mueller, but probably was collected by Leichhardt. No fruit was preserved. An isotype of this was sent to Kew, and Solms stated (Linnaea 42:69, 1878) that it was staminate. Warburg listed it (Engler’s Pflanzenreich IV, 9:85, 1900) with the "Species incertae sedis,” and could neither supplement the description nor cite additional collections. S. T. Blake (Austral. J. Bot. 2:131, 1954) reviews the history of von Mueller’s several publications of P. aquaticus and concludes correctly that his "remarks given are sufficient to validate the name.” He reduces P. de-Lestangii Martelli to its synonymy, believing that there is but a single Pandanus species with unicellular drupes in the area of northern Northern Territory and north- west Queensland. The present writer is in full sympathy with efforts to document and establish the identity of early described species. However, in this case, P. aquaticus F. Muell. rests upon a few non- diagnostic, descriptive words, and upon one leaf and a staminate inflorescence. It seems best treated as a valid name for a species so incom- pletely known that it should be left a species dubia, particularly as it is not safe to assume that only a single species of Pandanus can grow in one area. The next name, P. de-Lestangii Martelli, was based on good material from near Burketown, Queensland, and was published with an excellent diagnosis and illustration. This name is here adopted. Pandanus adscendens sp. nov. (sect. Pan- danus) Fig. 242 DIAGNOSIS HOLOTYPI: Arbor 7 m alta 18 cm diametro, cortice tuberculosa, radicibus ful- turosis 1-1.5 m longis 4 cm diametro verrucu- losis, foliis 1.3-1. 6 m longis proxima basem 7 cm latis in media 5.5 cm latis coriaceis in sectione oblate sinuose M-formatis gladi- formatis ex basi in apice subulato diminuentibus (apice non preservato) basi amplexicauli et inermi, ex 6-9 cm marginibus cum aculeis 1.8-3 mm longis 3-1 1 mm separatis crassiter subulatis adscendentibus brunneis vel cum apicibus brunneis, midnervo infra ex 15-20 cm cum aculeis 1.6-2. 2 mm longis 13-45 mm separatis graciliter subulatis adscendentibus, in sectione mediali marginibus cum subulato-serris 1-2 mm longis 4-10 mm separatis, midnervo infra cum serris simulantibus sed 10-32 mm separatis, proxima apicem marginibus et mid- nervo infra cum subulato-serrulis 0.2-0. 7 mm longis 2-7 mm separatis, pedunculo 13 cm longo bracteato, syncarpio solitario, nucleo 13 cm longo 4.5 cm diametro cylindrico-ellipsoideo obtuse deltoideo, phalangibus 5-5.6 cm longis 3. 3-4. 2 cm latis aurantiaco-luteis crassiter pyri- formatis vel cuneato-pyriformatis apice rotun- dato (rariter subtruncato) parte % supera libera, suturis lateralibus per y2_I parte libera distinctis 4-7-angulosis lateribus lateralibus laevibus lucidis in sicco palliditer brunneis sub- curvatis vel planis, sinibus centralibus apicalibus I -r- r- ^ ” r 1 O B cm.. 0 5 cm.. 1 i .j. __j i I Fig. 242. Pandanus adscendens St. John, from holotype. a, Phalange, lateral view, X 1; b, phalange, lon- gitudinal median section, X 1; c, phalange, apical view, % 1; d, apex of marginal carpel, oblique view, X 4; e, apex of central carpel, oblique view, X 4; f, leaf base, lower side, X 1 \ g, leaf middle, lower side, X 1; h, leaf apex, lower side, X 1. 528 Page 404: Revision of Pandanus , 23 — St. John 529 3-6 mm profundis V-£ormati§, carp.ellis 6-11 apicibus centralibus minoribus rotunda to- pyramidalibus vel subobiato-pyramidalibus illis marginalibus valde oblato-pyramidalibus vel semiorbicularibus et plemmque cum cavite parvo distali omnibus adscendentibus vel panels divergentibus, stigmatibus 1—1.5 mm longis apicalibus ovalibus vel orbicularibus obscuris sulcatis obliquis centripetalibus, sinibus proxi- malibus profundis l/2~2A a^ fondam extends, endocarpio supramediali 2-2.2 cm longo osseoso obscure brunneo lateribus lateralibus 3- 4 mm crassis, seminibus 12-15 mm longis ellipsoideis vel obliquiter ellipsoideis, meso- carpio in apice quaeque carpellae cavernam cum fibris fortibus longitudinalibus paucis et mern- branis medullosis formanti, mesocarpio basali fibroso et carnoso. DESCRIPTION OF ALL SPECIMENS EXAMINED: Tree 7 m tall, 18 cm in diameter; bark with prominent tubercles or warts; prop roots 1-1.5 m long, 4 cm in diameter, warty; leaves 1.3- 1.6 m long, 5-7 cm wide near the base, 3.6- 5.5 cm wide at the middle, coriaceous, in section depressed sinuous M-shaped, sword-shaped, tapering from the base to the subulate apex, but the actual tip not preserved, the very base amplexicaul and unarmed, but at 6-9 cm the margins with prickles 1.8-3 mm long, 3-11 mm apart, heavy subulate, ascending, brown or brown-tipped; the midrib below beginning at 15-20 cm up with prickles 1.6-2. 2 mm long, 13-45 mm apart, slender subulate, ascending; at the midsection the margin with subulate- serrae 1-2 mm long, 4-10 mm apart; those of the midrib below similar but 10-32 mm apart; near the apex the margins and midrib below with subulate-serrulations 0.2-0. 7 mm long, 2-7 mm apart; peduncle 13 cm long, bracted; syncarp solitary, the core 13 cm long, 4.5 cm in diameter, cylindric-ellipsoid, obtusely deltoid; phalanges 4. 8-5. 6 cm long, 2. 5-4.7 cm wide, 2.3-4. 1 cm thick, orange-yellow, pyriform or cuneate-pyriform, the apex rounded (rarely flattish), upper % free and in the free part the lateral sutures distinct from half to all its length, 4- 7-angled, the sides smooth, shining, when dried light brown, gently curving or plane; apical central sinuses 3-6 mm deep, V-shaped; carpels 6-13, the central apices somewhat the smaller, pyramidal or slightly oblate pyramidal to semiorbicular and most of them with a small distal concavity, the tips ascending or on a few slightly divergent; stigmas 1-2 mm long, apical, oval to orbicular, dark, creased, oblique, centri- petal; proximal sinus deep, running p|“% way to valley bottom; endocarp supramedian, 2-2.2 cm long, bony, dark brown, the lateral walls 3-4 mm thick; seeds 12-15 mm long, ellipsoid or obliquely so; apical mesocarp in each carpel forming a cavern with a few strong longitudinal libers and white medullary membranes; basal mesocarp fibrous and fleshy. holotypus: Australia, Queensland, Green Island, off Cairns, beach forest with Erythrina, Cordia suheordata, Morinda citrifolia, Feb. 9, 1958, H. St. John 26,266 (bish) . Fig. 243^. Pandanus darwinensis St. John, from holotype. Habit of mature trees. 530 PACIFIC SCIENCE, VoL XXI, October 1967 Fig. 2433. Pandanus darwinensis St. John, from holotype. Young trees and a detached syncarp. specimens examined: Australia, Queens- land, Green Island, off Cairns, beach forest with Erythrina, Cordia subcordata, Morinda citri- folia, Feb. 9, 1958, EL St. John 26,269 (bish); Percy I., Dec. 1870, McGeorge (mel); South Brooke L, G. Tandy (a); cult., Botanic Garden (Brisbane), C. T. White 3,332 (a). discussion: P. adscendens is a member of the section Pandanus , as is its closest relative P. Blakei St. John, also of Green L, a species with the phalanges with the central apical sinuses 1.5-3 mm deep; carpels 9-12; prop roots sparingly muriculate; leaves 8-8.5 cm wide, and the midrib below unarmed to beyond the middle. P. adscendens has the phalanges with the central apical sinuses 3-6 mm deep; carpels 6-13; prop roots warty; leaves 5-7 cm wide, and the midrib below beginning at 15- 20 cm up with ascending slender subulate prickles 1.6-2. 2 mm long, and 13-45 mm apart. The new epithet is the Latin participle ad- scendens, ascending, and is given with reference to the direction of the lower spines of the leaves. Pandanus darwinensis St. John (sect. Pan- danus) Fig. 243 a, h An isotype of this species is found in the collections of Martelli in Firenze. With it is a letter from the collector, C. E. F. Allen, Super- intendent of Agriculture, Darwin, Northern Territory, Australia, and two excellent photo- graphs. These are reproduced here by permis- sion of the Istituto Botanico, Firenze. From them the following additional details of descrip- tion can be derived. expanded diagnosis: Trees up to 6 m in height; trunk erect, simple, at length forking into erect branches; prop roots, if any, short; bark rather smooth; leaves ascending, then spreading, not becoming bent; infructescence with a single syncarp; peduncle about 63 cm long, recurving; syncarp 28 cm long, 21 cm in diameter, wide ellipsoid, bearing about 26 phalanges. The type locality is near Darwin, Northern Territory, and the photographs reveal that the species is littoral on marine shores. Revision of the Genus Pandanus Stickman, Part 24 Seychelles a New Section from the Seychelles Islands Harold St. John1 Seychelles sect. nov. (subgen. Pandanus ) Carpellis binis conmatis, stigmatibus binis arcuatis in lateribus oppositis cavitonis termi- nal! affixis, cavo cum stigmatibus falsis cordatis multis, dissepimento lato toto ex fibris longi- tudinalibus distinctis multis formatis. Arbor cum syncarpiis grandibus sphaericis. Carpels 2, connate throughout; stimas 2, like slender arcs on opposite sides of the terminal cavity which is filled with many elevated, cor- date, unequal structures that imitate stigmas; dissepiment with a pale, thin, cartilaginous tis- sue next to each seed, then all the rest of the broad interior solely of many, strong, separate, longitudinal fibers. Tree with large spherical syncarps. holotypus: Pandanus Hornei Balf. f., in Baker, FI. Mauritius and Seychelles 397, 1877. This species, when described, was not placed in a section. Later it was assigned to section V insoma by Warburg (Pflanzenreich IV, 9:54, 57, 1900), and with this placement Martelli concurred (Webbia 4(1) :94, 1913). In his redefinition of the sections, the present writer placed it in section Dauphinensia (Pacific Sci. 15(3) :34l, 1961). He now removes it from there and makes it the type of a new section. Like the Madagascar species of section Martelli- dendron, it is remarkable for having the dis- sepiment between the seed cavities formed not of a solid, bony tissue, but of a mass of distinct, longitudinal fibers. Also the two arclike stigmas, on opposite edges of the apical disc or con- cavity, are very similar. As for differences, the section Martelliden- dron, with 4 species from Madagascar, has the dissepiment solely of loose, longitudinal fibers; the phalange apex truncate, with an elliptic, firm disc, divided into quarters by deep valleys; the 2 stigmas lateral on opposite sides of the 1 B. P. Bishop Museum, Honolulu, Hawaii 96819. Manuscript received March 11, 1963. disc. The section Seychellea has the dissepiment of pale, thin, cartilaginous tissue next to each seed face, then all the rest of its broad interior made solely of many strong, separate, longi- tudinal fibers; the 2 arclike or parenthesis-like stigmas lateral on opposite sides of the rim of the apical cavity which is filled with many cordate pseudo-stigmas. Judging by Martelli’s illustration (Webbia 4(1) :t. 16, fig. 2, 1913) the fruit would be a 1-celled drupe. His specimen has been examined. He had a whole syncarp, but sectioned only one phalange. He made the cut between the two stigmas, just shaving the fibrous dissepiment, and cleaving only one of the two seed cavities. That the other seed cavity is there can be told by pushing a needle through the fibrous dis- sepiment. The phalanges seem to be always 2- celled. A new drawing is presented here to bring out the diagnostic structure of the dis- sepiment. Fig. 244. Pandanus Hornei Balf. f. in Baker. Transverse section of phalange, showing fibrous septum, two seed cavities, endocarp, fibrous mesocarp, and exocarp, X 1. Drawn from Isole Seychelles (fi). 531 532 PACIFIC SCIENCE, Vol. XXI, October 196? Pandanus Hornei Balf. f., in Baker, Fl. Mauri- tius and Seychelles 397, 1877; Warburg, Pflanzenreich IV, 9:57, 1900; Martelli, Webbia 4(1) :17, t 16, figs. 1-3, 1913; St. John, Pacific Sci. 15(3) : 341 , 345, figs. 20-21, 1961 (sect. Seychellea) . Fig. 244 supplementary description: Leaves at midsection with 104 parallel secondary veins in each half, the tertiary cross veins visible below toward apex, forming square or short oblong meshes. specimens examined: Isole Seychelles, dal Museo di Parigi (FI). The new section can be placed in the key to Pandanus (Pacific Sci. 14(3) :226, I960), by re- placing the first heading R, with the following: R. Phalange with the septum between the two cells of many loose fibers, i. Septum wholly of loose fibers; phalange apex an elliptic disc cut by valleys into quarters; stamens umbellate on the capitate apex of the column, and with abortive gyno- ecium Martellidendron 1. Septum of a thin cartilaginous tis- sue next to each seed, then its broad interior solely of loose fibers; pha- lange apex concave and filled with unequal, cordate pseudo-stigmas. . . Seychellea Revision of the Genus Pandanus Stickman, Part 25 Pandanus tectorius var. sinensis Warburg Harold St. John1 Pandanus tectorius Warburg var. sinensis Warb., Engler’s Pflanzenreich IV, 9:48, 1900. (sect. Pandanus ) The entire presentation of this variety was: "Folia minora angusta flagello longo terminata, spinis marginalibus quam in typo majoribus armata. Phalanges minores pauci-(5-6-) locu- lares. Siidchina. (Warburg, Naumann, Henry).” This characterization is wholly inadequate, and no type was indicated. In the Berlin her- barium the original specimens were still to be found in 1962. Warburg 3,482 from Macao con- sists of a single leaf 77 cm long, 3 cm wide; near the base the marginal prickles are alter- nately small and large, the latter 4-6 mm long, straight subulate, stramineous, ascending at 45°, the successive prickles 5-12 mm apart; the sub- ulate leaf apex at the point about 10 cm down 1.5 mm wide. Until complete material is known from Macao, Warburg 3,482 is undeterminable. The second specimen was from Hong Kong, Winter, 69/70, Dr. C. Naumann, and it con- tains parts of four leaves. It is probably P. remotus St. John. The third collection is more adequate. It is from Hainan, 1889, A. Henry. One sheet bears two young pistillate inflorescences. A second bears a pocket with several very young pha- 1 B. P. Bishop Museum, Honolulu, Hawaii 96819. Manuscript received March 11, 1963. langes, and one mature phalange 3.3 cm long, but this is a distorted, very asymmetric, 6-celled, basal one, not suitable for identification, but, since it is the only fruiting one in the series, it is here designated as lectotype of the var. sinensis. A third sheet with the same data is A. Henry 8,290. This bears parts of two good staminate inflorescences. These three sheets from Hainan represent a plant that differs from P. hainanensis St. John in its much broader leaves, etc. These specimens are not complete enough to identify. In conclusion, the var. sinensis Warb. is in two parts indeterminable, and in the third part probably synonymous with P. remotus St. John. Subsequent botanists, apparently content with Warburg’s description, have identified numerous collections as belonging to this variety. They have recorded the plant from the Mascarene Islands, India, Cambodia, Malaya, Tonkin, Philippines, Formosa, New Caledonia, Australia, Tonga, and Hawaii. Their collections are di- verse, and do not represent one species or variety. Since the original var. sinensis Warb. is ill-defined and incapable of identification, these later locality records can be rejected with- out further comment. Since the author considers the publication of P. tectorius by Parkinson (or Solander) invalid, he now attributes it to Warburg, who in 1900 first validly published the binomial with a description. 533 Reversal of Ethionine Inhibition by Methionine during Slime Mold Development Hans R. Hohl and Susan T. Hamamoto1 Dictyostelium discoideum , a cellular slime mold, has been shown by Filosa (I960) to produce abnormal fruiting bodies when grown on Es- cherichia coli in the presence of 1.2 x 10-3 * * * M ethionine. The fruiting bodies obtained un- der those conditions were described as short, with thick stalks and elongated sori. At a con- centration of 4.8 X 10-3 M no growth oc- curred. Starting from this initial observation we have attempted to answer the following questions : Which elements of development, such as morphogenesis or differentiation of spores and stalk cells, are affected at various concentrations of ethionine? Which phases of the life cycle are sensitive to ethionine? What is the mechanism of ethionine inhibition ? Answers to these questions might possibly pro- vide clues to the mechanism of development in these organisms, where an integrated mass of myxamoebae produces, in the absence of an external source of energy, a well-proportioned fruiting body composed of a cellulose-ensheathed stalk bearing a lobose sorus of spores. MATERIALS AND METHODS Myxamoebae of Dictyostelium discoideum NC-4(S2), a haploid strain, were grown at 24°C in the presence of Escherichia coli B/r on a medium containing 1% lactose, 1% pep- tone, and 1.5% Difco agar. After 44-46 hours, shortly before the cells reached the stationary growth phase, they were harvested with ice- cold Sorensen’s phosphate buffer (0.016 M, pH 6.0), washed, and their concentration ad- justed to 1. 5-2.0 X !08 cells/ml. A 0.5-ml sample of cell suspension containing 0.75- 1.0 X 108 cells was spread onto millipore filters 1 Pacific Biomedical Research Center and Depart- ment of Microbiology, University of Hawaii, Hono- lulu, Hawaii. Manuscript received September 26, 1966. This work was supported by a grant from the National Institutes of Health (GM-1 1758-03), United States Public Health Service. according to the method of Sussman and Lov- gren (1965). For this the millipore filters (48 mm diameter, black), resting on absorbent filter pads, were placed in plastic petri dishes (60 mm diameter) . Prior to the addition of myxamoebae the pads were soaked with 1.5 ml of phosphate buffer containing per ml 0.67 mg of strepto- mycin and 0.13 mg of sodiumlaurylsulfate to- gether with appropriate concentrations of the test solution. The sodiumlaurylsulfate appears to contribute to a more uniform development of the population and to enhanced effectiveness of at least some compounds, probably by in- creasing the cell permeability without having any obvious effect on normal development. All cultures were incubated at 24 °C in the dark; in some experiments when the plates were kept at room temperature overnight, the tem- perature varied between 21° and 25 °C. The plates were scored every 4 hours until no fur- ther developmental changes were observed. To test the effect of ethionine on growth and ag- glutination, the cells were grown on E. coli in shake cultures in the presence of the compound and the growth curves determined (Hohl and Raper, 1963), or washed cells were incubated in roller tubes and their pattern of agglutina- tion was observed (Hohl and Raper, 1964). For the growth in shake cultures, D. discoideum V-12 was used. RESULTS Ethionine completely inhibits growth of Dictyostelium at a concentration of 3.0 X IQ-2 M and shows no inhibitory effect at 1.0 X 10-3 M (Fig. 1). In no concentration, however, does it interfere with agglutination of myxamoebae as judged from the behavior of cells in roller tubes. Even at a concentration of 3.0 X 10_2M there was no sign of inhibition either in the rate of agglutination or in the size of the agglutinates formed. This indicates that 534 Reversal of Ethionine Inhibition by Methionine — Hohl and Hamamoto 535 Fig. 1. Growth of Dictyostelium discoideum on Escherichia coli in the presence of various concentra- tions of ethionine. The ordinate indicates cells/ml at the end of the growth period ( 46 hours ) . The inocu- lum consisted of 1.0 X 105 myxamoebae/ml. cell adherence or "stickiness” is not suppressed by ethionine. This result corroborates the ex- periments to be reported next, where it was found that even high concentrations of ethionine do not completely inhibit streaming and a type of massing of cells reminiscent of aggregation. In order to test the effect of ethionine on the developmental stages per se, the myxamoebae were grown in the absence of ethionine first and then washed populations were subjected to ethionine as described under materials and methods. Figure 2 presents the effects of various concentrations of ethionine on the development ETHIONINE DIFFERENTIATION INTO TYPICAL TYPE CONCENTRATION (M) SPORES SHEATH STALK FORMS NO. 0 + + CELLS + iil^ 1 7.5 xIO'4 + + + M 2 1.5 xIO'3 - + + ULAt. 3 3.0xl0'3 m + (IKAA 4 6.0xl0'3 - - - 5 1.5x10-2 - - 6 Fig. 2. Influence of increasing concentrations of ethionine on the development of Dictyostelium dis- coideum. The results represent the final stages of de- velopment regularly reached after 24 to 30 hours. of D. discoideum. Ethionine did not appreciably reduce the viability of the myxamoebae during the time of the experiment. Hence, the effects observed cannot be attributed to a partial killing of the cell population. A concentration of 1.5 X 10-2M completely inhibits morpho- genesis, although some streaming may occur resulting in the formation of vaguely defined clumps that usually disappear within 24 hours. At 6.0 X 10-3 M large, flat mounds of myx- amoebae are formed out of which tiny papillae may protrude, rudimentary signs of induction of polarity and fruiting body formation. The large number of these small papillae per mound indicates further that the critical mass, i.e., the mass of cells capable of integrated behavior (Hohl and Raper, 1964), has been drastically reduced. At a concentration of 3.0 X 10~3 M the myxamoebae collect into large aggregates that split up to form many finger-like pro- trusions. These protrusions are made up typi- cally of a heavy mass of cells at the base, a short, thick stalk-like structure oftentimes carrying at its apical end a lobose mass of cells some- what resembling a sorus. Up to this stage all the structures are made up of roundish cells without signs of spore differentiation or produc- tion of a cellulose sheath. Some large vacuolated cells, however, are present and we inter- pret them as representing stalk cells. At a con- centration of 1.5 X 10_3M stalks are formed and are ensheathed in a smooth cellulose envelope. The outside of the sheath is often covered with masses of undifferentiated cells, thus making the structures as a whole appear distorted. No spores are found in these fruiting bodies, but clumps of undifferentiated cells may occur at the tip of the stalk. Because of the still heavy base and the stalk without sorus the whole structure at this level often resembles a bowling pin. At a concentration of 7.5 X 10-4 M the fruiting bodies, apart from occasional distorted forms, are normal in ap- pearance though reduced in size. No apparent effect can be observed at lower levels of ethionine. For convenience the different levels of inhibition have been numbered from 1 to 6, as indicated in Figure 2. In general, spore differentiation is most sensitive to ethionine, followed by sheath production, stalk cell dif- ferentiation, and lastly morphogenetic move- 536 PACIFIC SCIENCE, VoL XXI, October 1967 ment. Of developmental stages, spore formation is also the most sensitive to treatment with 2- mercaptoethanol, as shown by Gerisch (1961). Spore formation and normal morphogenesis are closely linked, as spores have been found only in normal sori, and normal-appearing sori al- ways contained spores. Whereas no morphogenesis takes place in the presence of 1.5 X 10-2 M ethionine, the simultaneous addition of methionine in increas- ing amounts leads to the formation of pro- gressively more normal appearing fruiting bodies (Fig. 3). The structure of the resulting sorocarps follows the same pattern as described above for ethionine-treated populations. When the concentration of methionine approaches twice that of ethionine the fruiting bodies look normal but are somewhat smaller in size (equal to stage 2) compared with controls growing on methionine alone or on the buffer alone. If the experiment is repeated with 1/5 the con- centration of ethionine and a correspondingly lowered amount of methionine the results re- main unchanged. This demonstrates that the ratio of ethionine to methionine is the critical factor, rather than the absolute amounts of either compound. Ethionine clearly behaves like a competitive inhibitor of methionine. Next, ethionine sensitivity at the various de- velopmental stages (after cessation of vegetative growth) was determined. For this, populations were exposed to 1.5 X 10-2 M ethionine at selected points in their development for various periods of time. Some of the results are sum- ETHIONINE METHIONINE TYPICAL FORMS TYPE CONC. (M) CONC. (M) NO. 0 0 Hit* 1 0 1.5 XlO'2 HI 1 1.5 X I0'2 0 ^ 6 1.5 x I0"2 0.75xl0'2 4 I.5xl0-2 I.5xl0'2 JLASlh 3,2 I.5xl0"2 3.0x10" 2 ML 2,1 Fig. 3. Influence of ethionine and methionine, alone and in combination, on the development of Dictyostelium discoideum. The results represent the final stages of development regularly reached after 24 to 30 hours. Fig. 4. Sensitivity of various stages of slime mole development to the inhibitory action of ethionine. The time scale indicates hours after deposition of the washed myxamoebae populations together with the onset of aggregation (ag), migration (mig), and culmination (cul) of the control population. Solid lines denote the presence of ethionine, dotted lines of methionine. The resulting types of sorocarps are numbered according to Figure 1. They represent the final levels of development reached under the re- spective conditions. The delay in time to reach these final stages with respect to the controls is approxi- mately equal to the time the populations were exposed to ethionine. marized in Figure 4. Before and after the treatment with ethionine the cells were kept on equimolar concentrations of methionine. This was done in order to assure a quick removal of any possible free ethionine in the internal amino acid pool (Wright and Anderson, I960) that might obscure the results. In fact, it was found that when the cells were transferred (after treatment with ethionine) to buffer alone, instead of to buffer plus methionine, morpho- genesis was permanently inhibited. This strongly indicates the presence of a rather large internal pool where the ethionine can persist for hours, unless it is exchanged for exogenous methionine. The main conclusions to be drawn from Figure 4 are: (1) the entire morpho- genetic part of the life cycle up to the actual formation of the stalk is sensitive to ethionine, and (2) the inhibitory effect of ethionine is a gradual one, the damage becoming more severe the longer the cells are in contact with the substance. An important point is that even if ethionine is administered just prior to com- mencement of stalk production the inhibitory Reversal of Ethionine Inhibition by Methionine — Hohl and Hamamoto 537 effect is still very strong. Once the stalk is being formed the mass rises above the substrate and the further effects of ethionine are difficult to assess since the substances in the filter pad are no longer in direct contact with the cell mass. If ethionine is added for a certain period of time before aggregation, aggregation itself is also disturbed as described previously, i.e., only vaguely defined clumps form as the result of some streaming, or there is no reaction at all. Also, if ethionine is added after aggregation, then the proper polarization of the aggregated mass is disturbed and many small tips are formed as long as the contact with ethionine is maintained. The results of this experiment show that ethionine acts on several phases of the developmental sequence as well as on the vege- tative growth. Most significantly, however, it acts directly on the last stage, culmination. DISCUSSION Ethionine interferes with various stages of the life cycle of Dictyostelium, such as vege- tative growth, aggregation, and culmination. We have been able to show that the inhibition of the later stages of development is not neces- sarily a consequence of inhibition of earlier ones, since addition of ethionine as late as just prior to culmination is still inhibitory to a large extent. Spore differentiation is most sensitive to the action of ethionine, followed by cellulose sheath formation, stalk cell differentiation, and finally, morphogenetic movement. A similar sequence of sensitivity has been observed by Gerisch (1961) in the case of mercaptoethanol. Ethio- nine, then, seems to be primarily an inhibitor of differentiation. In turn, several aspects of disturbed morphogenesis, such as lack of a sorus or production of a short bulky stalk, can be indirectly attributed to this effect. More- over, how can the cell mass build a slender, evenly tapered stalk in the absence of any cellu- lose component for structural support? We have established that ethionine acts as a competitive inhibitor of methionine. This con- clusion is based on the observations that (1) methionine is able to reverse the inhibitory effect of ethionine when the two are added to the cultures simultaneously, and (2) the final product of development is a function of the ratio of the two components rather than of their absolute amounts. The proteins incorporat- ing ethionine instead of methionine are rendered biologically inactive and, therefore, normal development cannot proceed. It may be imag- ined that the various disturbances effected by ethionine have a common cause, but this is unlikely since ethionine is probably incorporated into a variety of proteins normally containing methionine. One point deserves mentioning: the ethionine has to be in contact with the cell population for at least 4-8 hours to produce any persisting inhibitory effect. At least two possible explanations might account for this: (1) ethionine slows down protein synthesis as long as it is in contact with the cells, so that only a small amount of ethionine-containing material is formed; and (2) the protein turn- over is high, so that the ethionine-containing proteins are rapidly broken down and replaced by the normal methionine-containing ones, as soon as the cells are switched from ethionine to methionine. The fate of the ethionine in the cell popula- tion has not been directly traced. However, from the work of Wright and Anderson (I960) on methionine metabolism in Dictyostelium we are able to get some information directly ap- plicable to our situation. These authors have shown that the endogenous "free” amino acid pool is a function of the stage of development only and is not influenced by exogenous methio- nine. However, exogenous methionine can ex- change with endogenous methionine, and the extent of this exchange is a linear function of the exogenous methionine present. This means that increasing the external methionine con- centration does not alter the size of the internal pool, but does increase the rate of exchange between exogenous and endogenous methionine. If we assume that ethionine behaves similarly in this respect, it then becomes clear that in- creasing the amount of ethionine in the medium does not simply add more ethionine to the existing endogenous amino acid pool, but re- sults in a higher exchange of ethionine for methionine between this pool and the environ- ment. In this way we have changed the ratio of methionine to ethionine, which is the essen- 538 PACIFIC SCIENCE, Vol. XXI, October 1967 tial determinant for subsequent developmental events. The effectiveness of ethionine up to the point of actual stalk formation indicates that at least some of the proteins essential for nor- mal development are formed right up to the point of culmination. Wright and Anderson (I960) have shown that certain ethanol-insolu- ble, methionine-containing proteins are pro- duced at an increased rate during the stage of pre-culmination. It will be of interest to de- termine whether these proteins are enzymes involved in cellulose synthesis or if they are structural proteins responsible for cell polariza- tion and differentiation. Evidence from experi- ments with colchicine in conjunction with elec- tron microscopy indicates the occurrence of such a proteinaceous cytoskeleton in Dictyostelium (Hohl and George, 1966). SUMMARY Ethionine progressively inhibits the develop- ment of Dictyostelium discoideum, from a con- centration of 7.5 X 10-4 M (which induces somewhat smaller but normal fruiting bodies) to a concentration of 1.5 X 1°-2 M (which results in complete inhibition). Intermediate concentrations produce a variety of distorted forms. With increasing concentrations the in- hibitory effect is first noticed in spore differen- tiation, then in cellulose sheath production, followed by stalk cell differentiation, and finally in morphogenetic movement. Simultaneous addition of methionine re- verses the effect of ethionine, the final result depending on the ratio of ethionine to methio- nine rather than on the absolute amounts of either substance administered. Ethionine exerts its effect at any time in the life cycle up to the actual formation of the stalk, the final ap- pearance of the fruiting bodies being a function both of the stage at which the ethionine was applied and of the period of time the cultures were in contact with it. The results indicate that, first, ethionine acts as a competitive in- hibitor of methionine, and, second, the pro- tein or proteins incorporating ethionine and thereby rendered biologically inactive are being produced continuously up to the time of actual stalk formation. REFERENCES Filosa, M. I960. The effects of ethionine on the morphogenesis of cellular slime molds. Anat. Rec. 138:348. Gerisch, G. 1961. Zellfunktionen und Zell- funktionswechsel in der Entwicklung von Dictyostelium discoideum. III. Getrennte Beeinflussung von Zelldifferenzierung und Morphogenese. Roux’ Archiv Entw. mech. 153:158-167. Hohl, H. R., and R. P. George. 1966. Col- chicine inhibition of cell polarization and cellulose synthesis in Dictyostelium. J. Cell Biol. 31(2) :47A-48A. and K. B. Raper. 1963. Nutrition of cellular slime molds. I. Growth on living and dead bacteria. J. Bacteriol. 85:191-198. 1964. Control of sorocarp size in the cellular slime mold Dictyostelium dis- coideum. Develop. Biol. 9:137-153. Sussman, M., and N. Lovgren. 1965. Prefer- ential release of the enzyme UDP-galactose polysaccharide transferase during cellular dif- ferentiation in the slime mold, Dictyostelium discoideum. Exptl. Cell Res. 38:97-105. Wright, B. E., and M. L. Anderson. I960. Protein and amino acid turnover during dif- ferentiation in the slime mold. II. Incor- poration of (S35) methionine into the amino acid pool and into protein. Biochim. Biophys. Acta 43:67-78. Comparative Decay Resistance of Twenty-five Fijian Timber Species in Accelerated Laboratory Tests Lynette D. Osborne1 ABSTRACT: Specimens from the heartwood of 2-5 trees of each of 25 species of Fijian rain forest timbers were tested by the laboratory soil-block method against two white-rot fungi, Fomes lividus (Kalch.) Sacc. and Pycnoporus coccineus (Fr.) Bond, and Sing., syn. Coriolus sanguineus (L. ex Fr.) G. H. Cunn. ; and against two brown-rot fungi, Lenzites trabea Pers. ex Fr. and Coni op bora olivacea (Fr.) Karst. The species most resistant to decay were Palaquium hornei, Intsia bijuga, Fagraea gracilipes, Syzygium spp. complex, and Dacrydium elatum. Most of the species tested were highly susceptible to decay. There was a tendency, both among species and within species, for the denser and less water-absorbent wood to be more resistant to decay. Also, the outer heart- wood was, in general, more resistant to decay than inner heartwood. A recent study was made of the decay re- sistance of a number of tropical rain forest timbers of New Guinea (Da Costa and Os- borne, 1967). Prior to this there was almost no information on the durability of the rain forest species of New Guinea, or of the East Asian and South Pacific areas in general. Be- cause more and more local timber is now being used in these countries, there is an increasing need for knowledge of the approximate dura- bility of these timber species so that efficient use can be made of the timber available. This situation applies in Fiji, as local timber has not previously been used extensively for perma- nent structures, and little is known of its performance in service. Unfortunately, although both New Guinea and Fiji have tropical rain forest vegetation, it appears that there are very few species common to both countries, so that information obtained for New Guinea species has little application in the use of Fijian tim- bers. This fact is a reflection of the great vari- ety of tropical rain forest timbers in the world and emphasizes the need for information on the durability of this large group of timbers. It is therefore desirable to study the decay resistance of the species occurring most com- monly in Fiji so that suitable timber can be 1 Division of Forest Products, CSIRO, South Mel- bourne, Australia. Manuscript received October 27, 1966. selected for a particular use. For example, it is desirable to use the most durable timbers for conditions of high decay hazard, such as for transmission poles, fence posts, sleepers, and bridge timbers. Less durable timbers may be suitable for external joinery, etc., that is, not in ground contact, whereas highly susceptible ones would be unsuitable for any external use in the humid climate without preservative treatment. One method of obtaining this information is by graveyard stake tests, and a few species are being studied in Fiji in this way (Alston, 1966). However, as these tests take some years to complete, an accelerated laboratory decay test was considered desirable. MATERIALS AND METHODS The methods used in this investigation fol- low closely those used in the study of 26 New Guinea timbers (Da Costa and Osborne, 1967) to which the reader is referred for more de- tailed information. Selection of Material The 25 Fijian timber species examined for decay resistance in these laboratory tests are listed in Table 1, together with local and family names. Although Swietenia macrophylla and Eucalyptus citriodora are not native to Fiji, 539 540 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE l Timber Species Tested TIMBER SPECIES TRADE NAME FAMILY Agathis vitiensis (Seem.) Drake Alphitonia zizyphoides (Spreng.) A. Gray Calophyllum spp.* Canarium spp.* Casuarina nodi flora Forst.* Dacrydium elatum Wall. Endospermum macrop hyllum (Muell. Arg. ) Pax et Hoffm. Eucalyptus citriodora Hook. Eagraea gracilipes A. Gray Garcinia myrtifolia A. C. Smith Gonystylus punctatus A. C. Smith Heritiera ornitbocephala Kosterm. Intsia bijuga (Colebr.) O. Kuntze My ri Stic a spp.* Palaquium fidjiense Pierre* Palaquium hornei (Hartog ex Baker) Dubard Parinari insularum A. Gray Podocarpus javanica (Burm. f.) Merr. Podocarpus neriifolius D. Don Podocarpus vitiensis Seem. Serianthes myriadenia Planch. Swietenia macrophylla King Syzygium spp. complex* Terminalia catappa L. Trichospermum richii (A. Gray) Seem. Eucalyptus microcorys F. MuelD Eucalyptus obliqua L’Herit.1 Pin us radiata D. Don1, Pseudotsuga menziesii (Mirb.) Franco1 Tectona grandis Lid dakua makadre Araucariaceae doi Rhamnaceae damanu Guttiferae kaunicina, kaunigai Burseraceae velau Casuarinaceae yaka Podocarpaceae kauvula Euphorbiaceae lemon scented gum Myrtaceae buabua Loganiaceae laubu Guttiferae mavota Gonystylaceae rosarosa Sterculiaceae vesi Leguminosae kaudamu Myristicaceae bauvudi Sapotaceae sacau Sapotaceae sa Rosaceae aumunu Podocarpaceae kuasi Podocarpaceae dakua salusalu Podocarpaceae vaivai-ni-veikau Leguminosae mahogany Meliaceae yasiyasi Myrtaceae tivi Combretaceae mako Tiliaceae tallowwood Myrtaceae messmate Myrtaceae radiata pine Pinaceae Douglas fir Pinaceae teak Verbenaceae * Groups consisting of more than one botanical species but regarded as one commercial species, t Reference timbers of known durability included for comparison. they have been listed with the other species, as they have been grown in plantations in Fiji for some years. Timber was collected in Fiji for each species, but often after testing had begun subsequent examination of botanical material showed that there were more than one botanical species within the "species” collected, two or more genera sometimes being represented. Wherever these species have been found to be very similar in appearance in the field, in anatomical struc- ture, physical and strength properties, and durability, the mixture has been regarded as one commercial species. These species mixtures are indicated in Table 1 and are discussed further under "Results.” In cases where a species differed appreciably from the main species of the group, it has been omitted. As these decay tests give comparative results only, five reference timbers, whose durability and performance in service are well known and which represent a wide range of durability, were included and are also listed in Table 1. As a general rule, specimens from five sep- arate trees were tested for each timber species. However, this number was not always avail- able and smaller numbers of trees were sam- pled for a few species (see Table 2). The timber was shipped in the form of green logs, and then a pith-to-bark billet measuring ap- proximately 24 inches longitudinally and 6 inches tangentially was cut from each tree, the Resistance to Decay of Fijian Timber — Osborne 541 TABLE 2 Properties and Decay Resistance of Species Tested* TIMBER SPECIES NO. OF TREES WOOD PROPERTIES PERCENTAGE WEIGHT LOSS** CAUSED BY TEST FUNGUS Basic Density (lb/cu ft) Water Uptake (%) Fomes lividus Pycnoporus coccineus Lenzites trabea Coniophora olivacea Palaquium 4 54.5 22.4 0.5 0.2 —0.1 0.1 hornet 51.0-57.0 21.4-23.7 0. 2-0.9 —0. 1-0.8 —0. 2-0.0 -0.1-0. 2 Eucalyptus 5 54.0 19.2 0.6 1.2 0.2 0.5 microcorys f 52.2-56.3 18.5-20.1 0. 3-0.9 0. 5-2.0 0.0-0. 6 — 3- 2-3.8 Intsia bijuga 4 46.0 28.3 2.4 0.2 0.0 0.2 43.0-50.8 23.6-31.4 0.0-8. 1 0. 0-0.4 0.0-0. 2 0.0-0. 4 Fagraea 2 50.8 17.5 0.9 1.2 1.6 1.8 gracilipes 49-1-51.9 17.3-17.8 0.7-1. 2 0.7-1. 7 1. 1-2.2 0. 3-4.0 Tectona grandis' 5 35.7 24.9 2.6 2.0 1.2 0.3 31.0-38.6 19.6-31.3 0. 2-6.1 0.4-5. 3 0.4-3. 0 0.0-1. 0 Syzygium spp. 15 47.7 22.5 6.3 3.1 0.6 5.7 39.4-55.1 18.8-34.1 0.2-24.3 -0.1-21.9 — 0.3-8.7 — 0.2-29.2 Eucalyptus 5 38.3 38.1 9.2 1.5 0.1 14.9 obliqued 35.3-43.6 30.7-43.3 2.1-18.2 0.4-2. 6 —0. 5-0.7 0.4-23.6 Dacrydium elaium 3 34.5 46.9 6.2 10.1 1.8 8.3 31.2-38.2 35.8-84.7 0.6-16.0 3.3-20.1 -0.5-8. 2 -0.3-22.6 Podocarpus 4 32.5 43.1 4.4 1.0 7.8 16.6 neriifolius 29-3-35.9 30.5-53.8 1.3-8. 3 -0.2-4. 5 0.0-13-9 4.3-22.5 Gave ini a 5 41.9 39.2 13.2 8.8 1.9 6.9 myrtifolia 37.3-46.2 35.8-46.6 2.1-29-9 0.9-27.5 0.4-6. 3 0.4-39.6 Heritiera 5 43.1 32.8 13.9 9.0 2.3 6.3 ornithocephala 33.2-50.4 23.0-51.0 4.2-28.9 2.3-27.0 0.4-6. 8 0.0-17.5 Swietenia 5 27.8 38.5 19.5 5.4 2.6 5.1 macrophylla 24.3-29.6 33.5-43.0 13.8-25.6 -0.2-12.7 —0.2-10.6 — 0.2-13.0 Palaquium 5 25.9 56.3 15.7 10.0 2.7 6.3 fidjiense 20.9-29-4 46.1-67.0 2.8-28.1 0.6-24.3 -0.3-14.5 -0.3-28.4 Serianthes 5 26.8 46.3 26.0 9.1 6.3 14.1 myriadenia 22.3-30.2 39.8-54.9 10.0-50.1 1.2-19.4 0.6-21.6 1.0-35.2 Pseudotsuga 5 26.0 41.5 10.6 4.4 19.0 32.2 menziesid 23.0-28.0 35.0-48.5 6.4-16.5 1.4-12.4 14.3-25.5 27.0-37.0 Calophyllum spp. 5 31.9 46.2 22.6 14.2 8.7 21.4 23.5-38.6 33.8-65.2 9-6-35.2 2.0-27.3 0.5-25.9 -0.7-41.0 Casuarina 4 52.9 31.8 19.7 26.5 5.1 18.2 nodiflora 51.0-54.0 24.6-40.4 13.6-26.8 14.4-39-4 —0. 5-9.1 5.9-28.3 Eucalyptus 4 41.5 42.4 24.4 27.9 7.8 18.6 citriodora 30.8-57.3 21.6-71.2 4.4-42.5 2.8-63.6 0.5-18.9 1.2-36.4 Podocarpus 5 24.2 72.2 13.3 21.2 17.8 27.8 vitiensis 22.6-25.7 44.4-99.5 8.9-18.8 14.7-28.0 8.2-21.3 23.6-31.0 Alphitonia 2 32.6 43.4 32.9 33.8 10.3 10.6 zizyphoides 30.3-34.8 39.3-50.6 28.5-41.8 28.1-39-7 6.2-12.7 0.0-25.6 Agathis 5 29.0 73.0 21.5 24.0 28.7 27.0 vitiensis 26.1-32.9 41.8-122.6 16.6-25.1 15.0-28.4 17.9-35.3 18.5-34.3 Parinari 5 38.8 51.7 29.6 31.7 16.9 27.4 insularum 34.0-43.7 42.0-62.1 19.6-41.3 20.3-40.6 9-2-25.8 21.7-30.1 T erminalia 4 25.1 57.0 40.6 32.6 17.2 25.0 catappa 16.9-32.8 43.8-67.7 28.7-58.1 24.0-43.5 4.7-26.3 10.8-34.3 542 PACIFIC SCIENCE, VoL XXI, October 1967 TABLE 2 ( continued ) WOOD PROPERTIES PERCENTAGE WEIGHT LOSS** TIMBER SPECIES NO. OF TREES Basic Density (lb/cu ft) Water Uptake (%) Fomes lividus Pycnoporus coccineus Lenzites trabea Coniophora olivacea Gonystylus 4 35.3 54.4 29.3 32.0 28.6 31.0 punctatus 29-9-41.6 48.4-60.4 25.1-33.9 23.1-37.7 19.8-38.6 25.0-36.7 Podocarpus 3 27.7 103.5 22.0 32.0 35.8 31.3 javanica 24.8-28.8 39.0-149.2 18.5-25.2 28.4-37.0 24.9-44.5 27.0-34.9 Pin us radiated 5 28.8 80.4 23.7 27.2 33.1 39.7 (sapwood) 25.6-32.3 75.3-84.9 21.4-27.5 20.8-36.0 24.4-39-7 38.2-41.0 Canarium spp. 5 29.4 49.7 48.0 28.9 24.7 30.4 26.1-33.8 35.0-77.4 32.1-60.1 21.0-47.4 9-9-35.6 17.4-40.0 Endospermum 5 25.1 65.0 41.9 44.0 33.8 36.4 macrophyllum 18.1-29-5 52.5-90.7 31.1-55.4 37.6-53.9 25.8-48.6 31.6-39.6 Myristica spp. 5 26.1 100.4 54.1 44.2 29.0 39.6 22.7-32.5 73.7-123 .6 49.8-60.1 36.4-63.3 19-0-39.2 33.6-46.0 T richospermum 4 17.4 198.2 60.7 55.7 31.4 46.1 richii 11.6-22.2 128.7-270.3 48.9-70.2 42.2-72.9 18.2-49.7 38.5-53.9 * Values represent the mean for two specimens (inner and outer heartwood) from each tree and the range. Species arranged in decreasing order of overall mean for four fungi. ** Incubation period of 8 weeks (12 weeks for P. coccineus) . I Reference timbers of known durability included for comparison. radial measurement varying with each tree. The material was air-dried before a quarter- sawn plank (y8 inch thick) was removed from each billet for testing. Two specimens were tested from each tree, for each fungal species. It has been shown in many timbers (Scheffer and Duncan, 1947; Findlay, 1956; Rudman and Da Costa, 1959; Rudman, 1964) that the outermost heartwood is the most durable wood in the tree, and so a sample was taken from this position, as well as one closer to the pith, representing the rest of the heartwood which would be used com- mercially. The test blocks measured % inch parallel to the fibres, 1 y8 inches radially, and y8 inch tangentially, the longest dimension being in the radial direction so as to sample the maximum variation in durability. The sap- wood was not normally tested, as it is usually nondurable and can be readily treated with preservatives if necessary. However, in some trees the sapwood-heartwood boundary could not be defined or there appeared to be little or no heartwood, and in these cases sapwood was tested as well as, or instead of, heartwood, in- asmuch as this would be the timber used com- mercially. Decay Tests A soil-block method was used in which cy- lindrical 8-oz glass jars (2% inches diameter; 3% inches high) with unlined metal screw caps were partly filled with 120 g of forest loam soil at 60% moisture content. Two feeder strips (1% X % X 1/16 inch) of beech ( Fagus sylvatica ) sapwood were placed on the soil and, after sterilization, were inoculated with the particular test fungus. After fumiga- tion with propylene oxide (Hansen and Snyder, 1947), two blocks, representing the inner and outer heartwood of the one tree, were placed in each jar, the largest face resting on the fungal mycelium. The percentage loss of weight, as compared with the air-dry initial weight, was used as a measure of the amount of decay. Four test fungi were used: Coniophora olivacea (Fr.) Karst, (dfp 1779) and Lenzites trabea Pers. ex Fr. (dfp 8845), both brown- rot fungi, and Pycnoporus coccineus (Fr.) Bond, and Sing. (syn. Coriolus sanguineus [L. ex Fr.] G. H. Cunn.) (dfp 2544) and Tomes lividus (Kalch.) Sacc. (dfp 7904), two white- rot fungi. The incubation period was 12 weeks Resistance to Decay of Fijian Timber — Osborne 543 for P. coccineus and 8 weeks for the other fungi. After completion of the main decay test, all blocks showing less than 10% weight loss after attack by C. olivacea or F. lividus were subjected to a further 16 weeks’ incubation with these two fungi. Measurement of Basic Density and Water Uptake It has been shown for 2 6 New Guinea timber species (Da Costa and Osborne, 1967) that there is a correlation between percentage weight loss and basic density and, more particularly, between percentage weight loss and water up- take. Therefore, measurements of these two properties were made on two specimens from each tree. The water uptake was measured by standing air-dry blocks, end grain down, in ]/g inch of water for 24 hours and calculating the increase in moisture content as a percentage of the oven-dry weight. The approximate basic density was calculated using the oven-dry weight and the "green” volume after blocks had been pressure-impregnated with water and allowed to swell for 48 hours. RESULTS The basic density and water uptake measure- ments, together with the decay figures, are presented in Table 2. As some timber species show considerable variation, both among trees and between the two radial positions within a tree, minimum and maximum values have been included, as well as the mean figure. As ex- pected, the outer heartwood was generally more resistant than was heartwood closer to the pith. In 70% of 41 6 relevant comparisons the percentage weight loss of the outer heart- wood was lower. Palaquium hornei proved extremely durable, being comparable in resistance with the very durable reference timber Eucalyptus microcorys. Intsia bijuga and Fagraea gracilipes also were durable, with several other timbers showing moderate durability. However, the remaining species showed poor resistance, most being highly susceptible. It can be seen from Table 2 that for each timber species there is a variation in the amount of decay depending on the particular test fungus, as well as a variation between trees. Because of these variations it is difficult to obtain a meaningful single-figure estimate of the relative decay resistance of the timber species. In Tables 2 and 3 the timber species are arranged in order of decreasing resistance based on the overall mean for the four fungi. However, other criteria may be used, such as the mean amount of decay caused by the most destructive fungus for each timber species, or the mean ranking for each timber (i.e., for each fungal species the timbers are ranked 1-30 in order of percentage weight loss, and the mean of these rankings for the four comparisons is obtained). Mean ranking is useful in cases where the test fungi show different rates of decay, L. trabea in particular producing almost consistently lower decay losses than the other three fungi. The advantages and disadvantages of the various methods have been discussed by Da Costa and Osborne (1967). It can be seen, however, that no matter which criterion is used the general order of the timber species does not alter appreciably (Table 3). Results for the second decay test of the more durable species are shown in Table 4. Pala- quium hornei and Fagraea gracilipes still proved to be durable, whereas the remaining timbers showed quite appreciable weight losses, at least against the white-rot fungus F. lividus. An interesting result is that Intsia bijuga showed a great increase in weight loss after a further 16 weeks’ incubation with F. lividus. As has been stated it was found, after testing had begun, that some timbers consisted of more than one botanical species (see Table 1). It is emphasized that these mixed groups in- clude only species which are regarded as being very similar in many respects, including natural durability. However, the following comments indicate the actual species tested. Of five trees tested of Calophyllum spp., three were identified as C. vitiense Turr. (mean percentage weight losses: 20.0, 15.7, 5.4) and two as C. leucocarpum A. C. Smith (weight losses: 20.4, 22.1%). The Canarium spp. group consisted of three trees of C. smithii Leenh., one tree of C. vitiense A. Gray, and one tree denoted as C. sp. aff. C. vitiense, all five trees 544 PACIFIC SCIENCE, Vol. XXI, October 1967 TABLE 3 Relative Decay Resistance by Various Criteria TIMBER SPECIES PERCENTAGE Overall Mean WEIGHT LOSS Mean for Worst Fungus MEAN RANKING* Palaquium hornet 0.2 0.5 1.0 Eucalyptus microcorys 0.6 1.2 3.5 Intsia bijuga 0.7 2.4 2.2 Fagraea gracilipes 1.4 1.8 4.8 Tectona grandis 1.5 2.6 5.2 Syzygium spp. complex 3.9 6.3 7.0 Eucalyptus obliqua 6.4 14.9 8.0 Dacrydium elatum 6.6 10.1 10.2 Podocarpus neriifolius 7.4 16.6 9-8 Garcinia myrtifolia 7.7 13.2 10.2 Heritiera ornithocephala 7.9 13.9 10.8 Swietenia macrophylla 8.2 19.5 10.5 Palaquium fidjiense 8.7 15.7 12.0 Seri ant he s myriadenia 13.9 26.0 15.5 Pseudotsuga menziesii 16.6 32.2 16.8 Calophyllum spp. 16.7 22.6 17.5 Casuarina nodiflora 17.4 26.5 16.0 Eucalyptus citriodora 19.7 27.9 18.5 Podocarpus vitiensis 20.0 27.8 18.0 Alphitonia zizyphoides 21.9 33.8 20.5 Agathis vitiensis 25.3 28.7 20.0 Parinari insularum 26.4 31.7 21.8 Terminalia catappa 28.8 40.6 22.8 Gonystylus punctatus 30.2 32.0 24.0 Podocarpus javanica 30.3 35.8 24.2 Pinus radiata (sap wood) 30.9 39.7 24.2 Canarium spp. 33.0 48.0 24.0 Endospermum macrophyllum 39-0 44.0 27.8 Myristica spp. 41.7 54.1 28.0 Trichospermum richii 48.5 60.7 29.2 * For each fungal species, the timbers were ranked 1-30 in order of increasing mean percentage weight loss and the mean of these rankings for the four comparisons was obtained. ! showing very similar durability (weight losses: 36.3, 34.9, 28.8; 31.5; 33.3%). Amongst the four trees of Casuarina nodi flora sampled was one tree identified as Gymnostoma vitiense L. A. S. Johnson, which showed a slightly higher mean percentage weight loss (23.1 cf. 13.0, 18.6, 17.4). However, no conclusions can be drawn from results for one tree of a species. Decay figures were similar for all trees of Myristica spp., which consisted of two trees of M. chartacea Gillespie (47.7, 39.8%), one tree of M. castanaefolia A. Gray (39.1%) and two trees of M. hypargyrea A. Gray (40.2, 41.8%). Of five trees tested of the Palaquium pdjiense group, two trees were identified as Palaquium n.sp., but all showed such similar decay losses that the two species could not be distinguished on durability in these tests. In the case of the Syzygium spp. complex, apparently a number of botanical species ap- pear very similar in the field, inasmuch as nine different species were received under the trade name of "yasiyasi.” On the basis of field char- acteristics and physical properties of the tim- ber, these species can be divided into two main groups, as shown in Table 5, which includes the mean percentage weight loss for all four fungi for each tree. As there is some variation in the mean percentage weight losses for different trees within a species, and within each group, and as only one or two trees were tested for many of the species, no differentiation Resistance to Decay of Fijian Timber — Osborne 545 TABLE 4 Decay Losses for Blocks of Durable Timbers Subjected to Second Decay Period TIMBER SPECIES MEAN PERCENTAGE WEIGHT LOSS Fomes lividus Coniophora olivacea 8 weeks 24 weeks 8 weeks 24 weeks Palaquium hornet 0.5 7.5 0.1 1.3 Eucalyptus microcorys* 0.6 4.2 0.5 1.0 Intsia bijuga 2.4 32.6 0.2 -0.3 Fagraea gracilipes 0.9 6.4 1.8 4.2 Tectona grandis* 2.6 22.6 0.3 2.7 Syzygium, spp. 4.0 34.2 2.1 14.3 Eucalyptus obliqua * 6.7 51.1 3.2 14.8 Dacrydium elatum 4.3 32.3 0.3 2.8 Podocarpus neriifolius 4.4 29-4 4.3 13.1 Garcinia myrtifolia 5.6 51.1 1.5 17.9 Heritiera ornithocephala 7.0 44.6 4.2 13.1 Swietenia macrophylla — — 3.9 19-2 * Non-Fijian timbers included for comparison. between the species or groups can be made from these durability results. Although Acicalyptus myrtoides belongs to the small-leaf group, tests carried out by the Division of Forest Products, csiro, show that this timber has different strength properties from all other species of yasiyasi. However, as far as natural durability is concerned, it is not possible from the present data to distinguish A. myrtoides from the other species, since only one tree was tested. It may be noticed that Eucalyptus citriodora showed rather low decay resistance. This is due to the fact that the specimens tested of two trees were sap wood (mean percentage weight losses 37.4 and 31.1 cf. 6.4 and 4.0 for trees where only heartwood was sampled). How- ever, since the sapwood in these trees extended for 4-6 inches from the bark, there would be sapwood present in most timber used for com- mercial purposes, and so the species will show low durability unless heartwood is carefully selected. The relationships between percentage weight loss, basic density, and percentage water up- take were investigated, first by using a mean value for each timber species, and then by ex- amining relationships within a species. Statis- tical analyses showed that the correlations be- tween these factors were very similar to those found by Da Costa and Osborne (1967) for TABLE 5 Decay Losses for Individual Trees of Syzygium Species Complex GROUP BOTANICAL SPECIES MEAN PERCENTAGE WEIGHT LOSS FOR EACH TREE Yasiyasi 1 Syzygium nidie Guill. 5.8 14.9 4.6 (Small-leaf group) Eugenia effusa A. Gray 1.8 0.6 6.0 0.9 Acicalyptus myrtoides A. Gray 0.0 GROUP MEAN 4.3 Yasiyasi 2 S. curvistylum (Gill) Merr. et Perry 2.8 (Medium-leaf group) S. fijiense L. M. Perry 3.5 S. brackenridgei (A. Gray) C. Muell. 3.8 Acicalyptus longiflora A. C. Smith 2.6 A. eugenioides (Seem.) Drake 6.0 A. elliptica A. C. Smith 4.0 1.5 GROUP MEAN 3-5 546 PACIFIC SCIENCE, Vol. XXI, October 1967 60 40 20 :• • POMES LIV1DUS • • • • • • J L 20 30 40 50 BASIC DENSITY (LB./ CU. FT.) 60 S40 20 PYCNOPORUS COCCINEUS • • -1 1 — * i - i» 20 30 40 50 BASIC DENSITY (LB^fcU. FT.) 60 40 20 LENZITES TRABEA • • • • * i • i •• l« 20 30 40 50 BASIC DENSITY (LB./CU.FT.) 60 40 20 CONIOPHORA OUVACEA • • % * J*. 20 30 40 50 BASIC DENSITY (IB./CU. FT.) Fig. 1. Relationship of decay resistance to basic density (species means). 26 New Guinea timbers. The scatter diagrams in Figure 1, using species means, indicate an inverse correlation between basic density and percentage weight loss for each of the four test fungi. It may be argued that, even if absolute losses in weight are identical for blocks of different densities, there would be a spurious inverse correlation of density with percentage weight loss. This possibility has been tested statistically, and it has been shown that the absolute weight loss was not constant for all species, and that there was a small (r = 0.32) but highly significant correlation between den- sity and absolute weight loss. As with the New Guinea timbers previously tested, there was a tendency for the more water-absorbent species to be more susceptible to decay (Fig. 2). In- asmuch as there was also a correlation between basic density and water uptake, multiple re- gression analyses were made. These showed that for all four fungi, the percentage water uptake was a better predictor of percentage weight loss than was basic density, as is indicated from a comparison of Figures 1 and 2. The effect of water uptake was significant at the 5% level for F. lividus and L. trabea, and at the 1% level for P. coccineus and C. olivacea, whereas the additional effect of basic density was not significant for any of the fungi. The relationship between basic density and decay resistance also held within the timber species for all four fungi. Because it is not practicable to give detailed results, data are presented for four timber species which showed wide ranges of basic density (Fig. 3). Statis- tical analyses again showed that the correlations between absolute weight loss (and hence per- centage weight loss) and basic density were highly significant in each case. The detailed data suggested a similar correlation between PER CENT. WEIGHT LOSS PER CENT. WEIGHT LOSS Resistance to Decay of Fijian Timber — Osborne 547 60 40 20 FOMES LIVIDU8 • • 9 99 •: 9 9 9. 9 40 80 120 180 PER CENT. WATER UPTAKE 200 60 40 <-> 20 Qc : LU CL PYCNOPORUS COCCINEUS • • • • • • I !• J J_ 40 80 120 180 PER CENT. WATER UPTAKE 200 60 40 i- 20 - LENZITES TRABEA • 9 . ** • 9 • 40 80 120 180 200 PER CENT. WATER UPTAKE 60 40 <-> 20 cr CONIOPHORA OLIVACEA 9 9 9 9 9 9 • •! • - - i 40 80 120 180 PER CENT. WATER UPTAKE 200 Fig. 2. Relationship of decay resistance to water uptake (species means). high water absorption and susceptibility to decay within each of these species. DISCUSSION From the data in Tables 2, 3, and 4 it can be seen that Palaquium hornei proved ex- tremely durable, even after a second, more severe decay test. Its resistance is comparable with that of the highly durable reference tim- ber, Eucalyptus microcorys, which is one of the timbers used in Australia for prolonged service in ground contact. Intsia bijuga and Fagraea gracilipes were also found to be durable, al- though I. bijuga did not show such high resis- tance after prolonged exposure to F. lividus. Specimens of L bijuga from New Guinea have been tested and shown to have comparable durability to the Fijian samples (Da Costa and Osborne, 1967), and also similar susceptibility to F. lividus during a second decay period. This timber has been widely used in ground contact in New Guinea, apparently with satisfactory results, and so the high susceptibility to F. lividus may be misleading. Syzygium spp. com- plex, Dacrydium elatum, Podocarpus neriifolius , Garcinia myrtifolia, Heritiera ornithocephala , Swietenia macrophylla, and Palaquium fdjiense all showed moderate durability, being slightly less resistant than Tectona grandis, which does not give extremely long service in the ground although it has an international reputation for durability. The remaining 16 timbers would probably be too susceptible for use in any situation of high decay hazard, such as ground contact, but a few less susceptible species could possibly give satisfactory service as exposed woodwork. 548 PACIFIC SCIENCE, Vol. XXI, October 1967 60 i- 50 - 40 - 30 - 20 - 10 - 20 PARINARI INSULARUM FOMES LMDUS % •*» 30 40 BASIC DENSITY (LB/CU. FT.) 50 50 - cn S a 40 - S 3o|- ^ 20 E 10 h HERITIERA ORNITHOCEPHALA FOMES LMDUS • • 1 20 30 40 50 BASIC DENSITY (LB./CU. FT) 60 50 - 40 - 30 - § O 20 b E 10 h TERMINAUA CATARRH FOMES LIVIDUS 20 30 40 BASIC DENSITY (LB/CU. FT.) 50 ENDOSPERMUM MACROPHYLLUM FOMES LIVIDUS # # 0 20 30 40 50 BASIC DENSITY (LB/CU. FT.) 60 r- 50 - 40 - 30 - 20 - 10 - ENDOSPERMUM MACROPHYLLUM PYCNQPORUS COCCINEUS •V 20 30 40 BASIC DENSITY (LB/ OJ FT) 50 i 40 i 30 - S’ 20 10 - ENDOSPERMUM MACROPHYLLUM LENZITES TRABEA 20 30 40 BASIC DENSITY (LE/CU. FT.) 50 Fig. 3. Intra-specific relationship of decay resistance to basic density (individual specimens). Resistance to Decay of Fijian Timber — Osborne 549 as does Pseudotsuga menztesit . However, those species which are comparable with the highly susceptible Pinus radiata sapwood would be un- suitable for external use in humid climatic conditions, unless they were impregnated with a preservative. Specimens of almost 20 of the species in the present test are, or have been, in graveyard stake tests in Fiji. Although these tests are not complete some comparison can be made be- tween the laboratory and field results (Alston, 1966). In general, there is good agreement be- tween the results of the two types of tests, the timber species ranking in approximately the same order, with only a few exceptions. Speci- mens of Garcinia myrtifolia, Palaquium fid- jiense , and Swietema macrophylla, when com- pared with the other timber species, all showed higher decay resistance in the laboratory tests than they did in the field tests. The reason for this discrepancy is not clear but could possibly be related to rate of wetting. S. macrophylla, al- though not native to Fiji, is an important plantation timber in Fiji, and it is therefore particularly important to note that for this species field test results are not as favourable as laboratory results. In conclusion, it should be stressed that the relationships obtained in the present tests for the tropical rain forest timbers of Fiji are very similar to those obtained for a group of com- parable timbers of New Guinea: notably, that less dense timber species tend to be more sus- ceptible to fungal decay, but, more particularly, that timbers which are highly water-absorbent are more susceptible. It is possible, therefore, that a knowledge of the density of a rain forest timber of which little else is known may be a rough guide to its durability. Again, a majority of trees was shown to have more durable heartwood in the outer zone than in the inner position, although the percentage was not as high as for the New Guinea timbers (70% cf. 86%). ACKNOWLEDGMENTS The author wishes to acknowledge the col- laboration of the Department of Forestry, Fiji in collecting material for this investigation, of the Royal Botanic Gardens, Kew in identifying specimens, and of Miss Nell Ditchburne in carrying out the statistical analyses. Thanks are due also to Mr. E. W. B. Da Costa for valuable discussion and to Mr. J. M. Stephenson and Miss Maureen A. Tighe for technical assistance. REFERENCES Alston, A. S. 1966. Natural Heartwood Dura- bility. Fiji Timbers and Their Uses, No. 2. Dept, of Forestry, Fiji. Da Costa, E. W. B., and Lynette D. Os- borne. 1967. Comparative decay resistance of 26 New Guinea timber species in accel- erated laboratory tests. Commonw. For. Rev. 46(1) : 63-74. Findlay, W. P. K. 1956. Timber decay — a survey of recent work. For. Abstr. 17:317— 327, 477-486. Hansen, H. N., and W. C. Snyder. 1947. Gaseous sterilization of biological materials for use as culture media. Phytopathology 37 (5) : 369—371. Rudman, P. 1964. The causes of natural dura- bility in timber. Pt. 16. The causes of varia- tion in decay resistance in jarrah (Eucalyptus marginata Sm.). Holzforschung 18:172-177. and E. W. B. Da Costa. 1959. Varia- tion in extractive content and decay resistance in heartwood of Tectona grandis L.f. J. Inst. Wood Sci. 3:33-42. Scheffer, T. C, and Catherine G. Duncan. 1947. The decay resistance of certain Central American and Ecuadorian woods. Trop. Woods No. 92:1-24. NOTES Notes on the Hawaiian Flora Benjamin C. Stone1 In this report on various Hawaiian plants are gathered taxonomic and nomenclatural notes which have accumulated over several years. Cer- tain new taxa, some nomenclatural adjustments, comments on noteworthy collections, and dis- tribution records are presented. CYPERACEAE An Additional Species of Cyperus in Kauai Cyperus haspan L. Kauai: Wahiawa Bog, 2 January 1957, Stone 1665 (bish). Det. T. Koyama. New to Kauai; known also (collected once, recently) from Hawaii. Evidently an ad- ventive species; its distribution is very broad. PAPAVERACEAE Authority for the Hawaiian Argemone In his monograph of the genus Argemone in South America and Hawaii, G. B. Ownbey cites the Hawaiian species as Argemone glauca L. ex Pope, Man. Wayside PI. Haw., p. 71, pi. 32, 1929, adding the remark "as 'Argemone glauca Linnaeus’ in error.” This was indeed as Pope published the species. It is manifest from Pope’s writing that he believed that A. glauca had pre- viously been described by Linnaeus, and he pre- sumably did not check on this assumption. In fact, Linnaeus published no such species. Pope clearly did not intend a new epithet. He also seems to have been unaware of the valid vari- etal name glauca published by Prain, in J. Bot. 33:329, 1895, with Nuttall indicated as the source. However, Pope may have seen this pub- lication, and, retaining in memory the epithet, forgot its rank and place of publication; or 1 Department of Botany, University of Malaya, Kuala Lumpur. Manuscript received June 15, 1966. perhaps the glaucous appearance of the plants in question simply suggested the same epithet. We also find Degener, in Plants of Hawaii National Park, p. 164, 1930, using the name Argemone glauca , but as a provisional name and hence not a formal nomenclatorial usage; and later in Flora Hawaiiensis (31 July 1958) where, with a long list of synonyms, the name is given as " Argemone glauca (Prain) Deg. & Deg. comb, nov.” Ownbey considers that this transfer is contrary to Article 32 of the Inter- national Code of Botanical Nomenclature (1956 ed.). There seems to be no good reason to in- voke Art. 32, however, which in itself offers no reason to consider as incorrect the author citation as suggested by the Degeners. Ownbey himself seems to perpetuate the idea that Pope was "attributing” the name to Linnaeus. This is a practice that has been used at times, but it has nothing to recommend it and in this case is clearly not supported. We are forced to conclude that Pope’s de- scription, although definitely not intended as a proposed new name and species, can be taken as if he had proposed a new species. He is defi- nitely the first to publish the binomial Arge- mone glauca. He does furnish a good descrip- tion; since it was published in 1929, he did not have to include a Latin description. He cited no specimens and no holotype; but citation of type is required only after 1 January 1958. He does furnish an illustration. There is no diffi- culty at all in interpreting his meaning. Conse- quently we can accept his publication as if it were describing a new species, and thus the correct citation of the name is: Argemone glauca Pope, Man. Wayside PL Haw. p. 71, pi. 32, 1929 (attributed in error to Linn.). A neotype should now be chosen for this species. The new combination by the Degeners, 550 Hawaiian Flora — Stone 551 though technically correct, is superfluous, as priority is by rank. MALVACEAE A Synonym for the Hawaiian Hibiscus clayii In Sister Roe’s study of the Hawaiian species of Hibiscus (Pacific Sci. 15:3-32, 1961) a new species from Kauai is described and named Hi- biscus newhousei for its discoverer, W. Jan Newhouse. It was first located in the foothills of the Moloaa Forest Reserve at near 500 ft altitude, and has been collected by I. E. Lane (no. 58-44, 10 November 1958) and by the writer ( Stone 3420, 3421, 15 April I960), in company with Tadayuki Kato. I find, however, that this species is synonymous with Hibiscus clayii Deg. and Deg. (Flora Hawaiiensis fam. 221, 20 March 1959), which was described from cultivated plants grown from cuttings long since taken from Haiku, Kauai, by Albert Duvel. Hibiscus clayii Deg. and Deg. Hibiscus newhousei Roe, syn. nov. Kauai: Moloaa, mountains ssw of Moloaa, Moloaa stream and waterfall, alt. about 700 ft, 15 April I960, Stone 3420 (flowering), 3421 (fruiting) (bish). A small grove of perhaps six plants was found, on vertical rocky banks just above a small waterfall. The trees approached 25 ft in height; some were in flower, and one bore sev- eral mature fruits, each in the form of a five- pointed star, the segments follicular, and bear- ing normal seeds. The flowers were dark-red in color. Associated species included Cordyline fruticosa, Aleurites moluccana, Pleomele aurea, and Eleocharis sp. Two Other Recent Collections of Hibiscus Hibiscus saintjohnianus Roe Kauai: Na Pali Coast, trail to Kalalau Valley, rim of Hanakoa Valley at about 800 ft alt., 14 August 1961, Stone, Stern and Carlquist 3748 (bright orange flowers), 3749 (darker reddish- orange flowers) (bish). Hibiscus sp. (perhaps H. arnottianus, forma) Oahu: Waianae Mountains, Palikea trail, alt. 1,500 ft, dry gully, 5 May I960, Stone 3471 (bish). This collection, taken from a largish tree about 30 ft high, has baffled certain deter- mination because it lacked flowering branches. However, it may be a form of H. arnottianus, probably f. parviflora Skottsberg. The leaves had light-magenta-colored midribs and veins, purplish and sparingly puberulent petioles (the pubescence stellate), and small subulate reddish soon caducous stipules. The importance of this collection is in showing the species in such a dry locality among almost nothing but weeds. EUPHORBIACEAE Tivo New Taxa in Aleurites The kuktii or candlenut tree, Aleurites mo- luccana, is a familiar plant in the Hawaiian landscape, its pale foliage distinct even at long distances and indicating the groves and isolated trees so common at moderate elevations on the Hawaiian mountains and in valleys. It is gen- erally agreed that the plant is one of the aboriginal introductions of the Hawaiians, since it figured largely in the Polynesian culture throughout the high islands of Polynesia, and indeed is a valuable tree for the people of many other Pacific islands. It is known to be planted in groves around many Hawaiian village and temple sites, along trails, and around present- day houses as well. Although surely widespread in Hawaii through natural means, it is also dis- tributed deliberately (or has been in the past), and perhaps also accidentally, since the seeds may be carried easily and perhaps dropped. In other words, it is a plant that is marginally a cultivated plant. Because it was of some im- portance (for torches, made from the oily seeds; for medicinal purposes, the seeds being some- what purgative in small doses and violently so in larger ones; and for food, either raw or pref- erably cooked), the early Hawaiians no doubt took an interest in the trees just as they did in their selections of taro varieties (Colocasia) , ti varieties (Cordyline) , bananas, and other plants. This would lead to an observance and to a deliberate selection of unusual forms among the ktikui trees, as it did with other plants of cultural interest. This selection would tend to perpetuate forms which might otherwise dis- appear (as for instance at the demise of a par- ticular tree with a remarkable recombination 552 PACIFIC SCIENCE, Vol. XXI, October 1967 type), and it would account for the spread of such a remarkable type from island to island. A few years ago just such an unusual kukui was described as a species, Aleurites remyi Sherff (Field Mus. Bot. ser. 17:558, 1939) from specimens collected long ago by Jules Remy (1851-55), apparently somewhere near Kona, Hawaii. Additional collections and good photographs of the leaves were published by Sherff in a later paper (Am. J. Bot. 31:157, pis. 1-3, 1944), and definite localities then were known: Holualoa-Kailua road, North Kona; and a garden in Hilo. A plant was grown from a seed by Dr. Sherff in Chicago. How- ever, according to Mrs. Thomas Jaggar, the Kona tree (or trees) were supposed to have been brought to Hawaii from Kauai (as a nut) and was known locally as the "Kauai” or "mango-leaved” kukui or, because of the Kona locality, as the "Kona” kukui . More recently, another unusual kukui has been discovered, nearly simultaneously, by Dr. Otto Degener and by Tadayuki Kato on Kauai. Like the "Kona” form, it differs from the com- mon kukui in its strikingly different leaves. An individual of this form may be seen on the grounds of Kauai High School in Lihue. In fact, these two forms have a common ten- dency, i.e., a relative lengthening of the leaf and reduction or loss of the lateral lobes. In A. remyi the lateral lobes are very narrow, the terminal lobe much elongated; in the Kauai High School form the lateral lobes are reduced or absent. In reasonably typical A. moluccana proper, the lateral lobes are present and rather broad, and the terminal lobe is not particularly longer. For comparison the sketches in Figure 1 are given. It will be noticed that even in A. moluccana proper there is a marked variation in extent of lobing. In describing A. remyi Sherff states: "Several staminate inflorescences have been seen. They appear different from those on A . moluccana , with which A. Remyi will stand in the section Camirium. . . . However, the floral characters of A. moluccana are so lacking in sharp delimi- tations that much reliance upon them for a dis- tinction from A. Remyi seems for the present unwise. A . Remyi has slightly smaller petals (for its staminate flowers) and these are often sharply 1 -denticulate on each margin close to their expanded distal portion, which in turn is very often irregularly obtuse-denticulate or -lobulate. In A. moluccana the tendency is for Hawaiian Flora — Stone 553 the petals to be entire or essentially so.” While agreeing that some minor variation in petals does occur it should be added that this is not of any great importance. In short, the discrimi- nation of species is made on the leaf characters alone. Knowing the somewhat cultivated nature of the Hawaiian kukui, and having now a reason- ably accurate radio-carbon date from archaeol- ogy that shows that the Hawaiian Islands were probably populated rather less than 2,000 years ago, I find it unlikely that A. remyi can be con- sidered an endemic native Hawaiian species. In addition, the examination of flowering material of many specimens of A. moluccana, including the Kauai High School form, fails to reveal any additional differentiating characters. In conclu- sion, it appears preferable to treat these forms as subspecific taxa and to suggest that, in fact, they are of aboriginal Hawaiian selection. The Hawaiian kukui trees, then, can be ar- ranged as follows: Aleurites moluccana (L.) Willd. (syn. A. tri- loba Forst.) (1) var. moluccana. This is the common form not only in Hawaii but elsewhere in the range of the species. (2) var. remyi (Sherff) B. C. Stone, stat. nov. A. Remyi Sherff, Field Mus. Bot. ser. 17:558, 1939; Am. J. Bot. 31:157, pis. 1-3, 1944. holotype : Remy 600, pro parte (3 sheets, Paris). Although this may be from Kauai, the locality at present known is North Kona, Ha- waii. As a common name "Remy’s kukui” is suggested. (3) var. katoi Degeners and Stone. The formal description of this variety will appear in the Flora Hawaiiensis, vol. 7. It is named for Mr. Tadayuki Kato of Kauai High School, who has been very helpful to me and to other visiting botanists. The holotype speci- men, taken from the tree on the grounds of Kauai High School in Lihue, is at the Bishop Museum ( Stone 3427, collected on 15 April I960). A further specimen collected by Dr. Degener is also available. A suitable common name would be "Kato’s kukui” or, alternatively, the "mango-leaved kukui.” ARALIACEAE A Recent Collection of Munroidendron racemosum The genus Munroidendron Sherff (Bot. Leafl. 7:21, 1953; Am. J. Bot. 43:47 6, 1956) is of particular interest because it is endemic in Ha- waii, consisting of a single species which is very rare. It was rediscovered a few years ago by Tadayuki Kato, of Kauai, and a small col- lection was made later by a party (the writer, with William Stern and Sherwin Carlquist) that found Munroidendron in the Nonou Range not far from the Wailua River, Kauai, on the west side at about 700 ft alt. in the second valley from the south end of the range. The trees were leafless at the time (16 August 1961), but were in flower, the long racemose inflorescences hanging from the rather thick bare branches. A photograph was taken but is not particularly good (Fig. 2). The specimen ( Stone 3768) is in the Bishop Museum. The single tree seen was on a relatively steep arid slope facing west, Fig. 2. Munroidendron racemosum (Forbes) Sherff; habit of a tree on Nonou Mountains, Kauai (Stone 3268). 554 PACIFIC SCIENCE, Vol. XXI, October 1967 associated with such species as Canavalia ga- le ata, Si da sp., Plectranthus australis, Mucuna gigantea, Aleurites moluccana , Cordyline fruti- cosa, Psidium guajava , Lantana camara, and other aridity-tolerant plants, many of them weedy introduced species. The tree, with its leafless appearance and numerous dry, decayed branches, had an unhealthy aspect. It should be sought in a season when fruits are ripe so that seeds may be gathered for the preservation of this very interesting species, now perhaps nearly extinct. APOCYNACEAE The ”Kalaipahoa ” Tree of W ah aula Heiau, Haivaii , Is Rauvolfia During a survey performed by members of a Bishop Museum expedition for the U. S. Na- tional Park Service in Kalapana, Puna, Hawaii in the summer of 1959, a single tree of Rau- volfia remoti flora Degener and Sherfif was found at Wahaula Heiau, near the coast and not far from Kalapana village. This tree, called "kalai- pahoa,” was discovered by S. Konanui and J. Halley Cox. Supposedly in earlier times, during the ascendancy of this heiau, a grove of trees existed, and this one is a remnant of the grove, which was said to contain many kinds of plants useful to the priests. Another collection was made higher on the dry slopes of Kealakomo in native forest ( Stone and Pearson 3016, alt. 1,400 ft. 9 July 1959, bish). The same species has previously been recorded from two localities in the Kau District (at Waiohinu and near Kaa- lualu) in SherfPs treatment of 1947 (Field Mus. Bot. sen). The possibility that the heiau’s priests used this plant suggests that they might have had some knowledge of the medicinal properties of the milky sap, which in some spe- cies (especially R. serpentina of India) provides an important drug now well known as reser- pine. It would be of interest, therefore, to have the Hawaiian species (which number seven) investigated for this material. SOLANACEAE The Identity of Solanum carterianum Rock Rock’s description (Indig. Trees Haw. Is. p. 423, 1913), and specimens collected since his discovery of this plant in Hawaii, accord well with a variety of another species: Solanum verbascifolium L. var. auric ulatum (Aiton) O. Kuntze. This appears to have a natural distribu- tion in tropics of both hemispheres, in such regions as Borneo, Sumatra, Java, and Amboina, and has been reported from Tonga. In Hawaii it has been found only in the Waiahole-Waianu Valley on the windward (east) side of the Koo- lau Range of Oahu. Rock listed a vernacular name ("pua-nanahonua”) for the plant, which evidently he took as an indication that it was an indigenous species. This is doubtful, how- ever. The very restricted occurrence of this plant in Hawaii, the fact that so many exotic species have been introduced (not all by known persons or at known times), and the remark- able, even reprehensible, lengths to which intro- ducers of foreign plants have gone in distrib- uting alien species, all tend to support the conclusion that this Solanum is an exotic, not a native, species. It may have been introduced about 1900. It is easily distinguished from the truly indigenous species of Solanum by its ar- borescent habit, dense fulvous tomentum, and rather large bluish-lavender flowers. Endemic species, such as S . kauaiense Hillebr., are shrubby, and bear white or purplish flowers. Other introduced species are also mostly herbs or shrubs, and inhabit disturbed areas. S. ver- bascifolium is illustrated, under Rock’s desig- nation, in Degener’s Flora Hawaiiensis. A New Variety of Nothocestrum ( Solanaceae ) Nothocestrum longifolium Gray var. rufo- pilosum B. C. Stone, var. nov. Folia magna elliptica usque ad 16-17 X 6.3 cm, laminis costis nervisque pilis rufis tomen- tosis sed ultime glabrescentibus, ramulis glabris. Leaves large, elliptic, up to 17 X 6.3 cm, the blades, costa, and nerves beneath tomentose with rufous hairs, but at last becoming glabrous; branches glabrous. holotype: Hawaii: between Glenwood and Twenty-nine Miles, in wet forest, 24 June 1929, O. Degener 7434 (us). For its species this is an unusually large- leaved plant, the blades densely rufous-tomen- tose beneath, especially along the midribs and lateral nerves, but in age becoming somewhat glabrate except on the midrib. Hawaiian Flora — Stone 555 medium-sued leaf x 1 Fig. 3. Lipochaeta acris Sherff; leaf, florets (disc- and ray-) and achenes (disc- and ray-). All from Stone 825, Kauai. 556 PACIFIC SCIENCE, Vol. XXI, October 1967 compositae the University of Hawaii, the extremely rare species Bidens cuneata Sherff was rediscovered. Rediscovery of Bidens cuneata Previously this had been known only from the In 1959, while I was botanizing on Diamond original description, based on a single type col- Head, Oahu, with Dr. Charles Lamoureux of lection, and had been declared extinct (Dege- Hawaiian Flora — Stone 557 ner, J. Pan-Pacif. Inst 7(2) :3, 1932). But several years ago, Lamoureux and E. T. Ozaki collected it a second time, and the plant ap- peared to be persisting, not common but not at all extinct. On the last trip about half a dozen individual plants were observed, all within a small area on the northwest rim of Diamond Head (Leahi). Illustrations of the achenes of two recently collected species of Lipochaeta (Compositae) . A number of photographs of herbarium sheets were illustrated in Sherffs monograph (Bish. Mus. Bull. 135, 1935) of the genus Lipochaeta, and many good line drawings are found in De- gener’s Flora Hawaiiensis. To these may be added the following: Lipochaeta acris Sherff Kauai: Na Pali Coast, Hanakapiai, 500 ft alt., 18 June 1955, Stone 825 (bish). Habit illustrated in Degener, Flora Hawaii- ensis; the additional illustration here (Fig. 3) shows disc and ray florets and disc and ray achenes. Lipochaeta alata Sherff Kauai: Na Pali Coast, Hanakapiai, 500 ft alt., 17 June 1955, Stone 760 (bish). Figure 4 shows habit, achene, and palea. Notes on the Ecology of the Pogonophoran Genus Galathealinum Kirkegaard, 1956 Oluwafeyisola S. Adegoke1 The pogonophoran genus Galathealinum Kirkegaard, 1956 has a rather wide areal and latitudinal range in the northern hemisphere. It has been recorded from the Arctic Ocean, latitude 69° 32'N (Southward, 1962) to the Celebes Sea, about 1° 50' north of the Equator (Kirkegaard, 1956). This wide latitudinal range makes the genus an exceptionally good one for examining the ecological factors which control pogonophoran distribution, and one would like to know what ecologic factors may be found to interpret the wide latitudinal range. Four species have been described within the genus. The first, Galathealinum hruuni (the type species) was described by Kirkegaard (1956) from the Celebes Sea in the western Pacific. Ivanov (1961) next described G. hrachiosum from the Pacific coasts of Canada and Oregon. The most northerly known spe- cies, G. arcticum was described by Southward (1962) from Thetis Bay, Herschell Island, north of Yukon, Alaska. The writer recently (Adegoke, 1967) described the fourth species, G. mexicanum from collections made in the Gulf of Tehuantepec, Mexico. The specimens from the same region earlier listed as G. hruuni (?) by Parker (1963:86) belong to this latter species. Apart from these four species, Hartman and Barnard (I960) listed the occur- rences of a few large-sized fragments (3-4 mm diameter) of pogonophoran tubes from West Cortes, East Cortes, and Long basins, and from the San Diego Trough off the coast of southern California. These fragments were later referred to the genus Galathealinum by Hart- man (1961:546). Although specifically inde- terminable, these fragments are significant be- cause they are the largest reported tubes of members of this genus. 1 Division of Geological Sciences, California Insti- tute of Technology, Pasadena, California. Present address: Department of Geology, University of Ife, Ibadan Branch, Ibadan, Nigeria, West Africa. Manu- script received September 29, 1966. ACKNOWLEDGMENTS The author wishes to thank Professor Heinz A. Lowenstam of the California Institute of Technology for his critical reading of the manu- script and for his helpful suggestions. The figures were prepared by Mrs. Ruth Talovich. ECOLOGY Because the first records of pogonophoran species were from great depths it was initially assumed that pogonophorans were exclusively inhabitants of abyssal and hadal depths (Kirke- gaard, 1956:80). Subsequent records (Ivanov, 1963; Jagersten, 1956; Kirkegaard, 1958; A. J. Southward, 1958; E. C. Southward, 1962; Southward and Southward, 1958, 1963) of pogonophoran species from extremely shallow waters (for example, Sihoglinum caulleryi Ivanov from 22 m in the Sea of Okhotsk; and Galathealimim arcticum Southward from 36 m in Thetis Bay, Herschell Island, Yukon, Alaska) clearly showed that absolute depth is not neces- sarily a limiting factor in pogonophoran ecology. It is now known that, although a majority of pogonophoran species inhabit abyssal and hadal depths, only a few are characteristically confined to such habitats (Ivanov, 1963:123-126; Southward, 1962:385). Many species are en- countered at comparatively shallow depths and a few forms are also known to dwell at bathyal or even sub-littoral depths. A relatively large number, however, thrive in shallow as well as in deep waters. Thus, S. caulleryi Ivanov has been recorded from depths ranging from 22 m in the Sakhalin Gulf to depths of about 8,164 m in the Kuril-Kamchatka Trench (Ivanov, 1963:221). According to D. B. Carlisle (see Ivanov, 1963:123), this is the greatest known bathymetric range for any known species of marine organism. In the light of these presently known depth distributions, Kirkegaard (1958:1087) and 558 Ecology of Galathealinum — Adegoke 559 Southward (1962) concluded that the limiting factor in pogonophoran distribution is low water temperature rather than absolute depth. The geographic, latitudinal, and bathymetric distribution of the recorded species of Galathealinum discussed below corroborates the views of Kirkegaard and Southward. The geographic and bathymetric distribu- tion of the four species of Galathealinum and of the undetermined species from southern Cali- fornia (Hartman and Barnard, I960; Hartman, 1961) are shown in Table 1. Figure 1 shows a plot of the minimum depth of occurrence of the species against latitude of occurrence. The data show that there is a direct correlation between these two parameters. The species oc- cupy increasingly greater depths the lower the latitude. The approximate values of bottom tempera- tures at the localities where the pogonophoran species discussed here were collected are shown also in Table 1. Temperature values were ob- tained from published sources, especially from the work of Emery (1954, I960), Emery and Rittenberg (1952), Kramp (1957), Manner F \ \ \ \ h \ 50 40 20 <0 \ + Habitable \ \ \ Depths. \ + 4#^ \ Uninhabitable Depth: J L X+2 +\ \ \ 0 >1000 2000 3000 4000 5000 Depth in Meters Fig. 1. Relationship between latitudinal position and minimum depth range of species of Galatheali- num. TABLE 1 Geographic and Bathymetric Distribution of Species of Galathealinum Kirkegaard, 1956 DEPTH IN BOTTOM TEMP. SPECIES LOCATION LATITUDE LONGITUDE METERS °C Galathealinum Thetis Bay, Herschell 69°32'N 138°57'W 36 ca. 0° arcticum Island, Yukon, Alaska Galathealinum brachiosum west coast of Canada west coast of Oregon 54°23'N 42°40'N 134°4TW 124°29'W 1233-2605 0.61°- 0.72° Galathealinum East Cortes Basin, 32°2TN 1 18°40'10"W 1872 3-13° sp. indet. southern California 32°16'30"N 118°27'55"W 1801 West Cortes Basin, 32°21'N 119°14'W 1924 southern California 32°l4'N 1 19°15'W 1923 3.3° 32°1TN 1 19° 18'W 1668 San Diego Trough 32°19'N 117°26'55"W 1420 — Long Basin, 31o55'09"N 119°10'W 1833 2.70° southern California Galathealinum Gulf of 14°28'N 95°09'W 3529-3557 1.5° mexicanum Tehuantepec, Mexico 12°20'N 91051'W 3596-3642 1.5° Galathealinum Celebes Sea 1°50'N 119°30'E 5090-5110 o 00 fO bruuni 560 PACIFIC SCIENCE, Vol. XXI, October 1967 0L 0 _j | L_ | | i 10° 20° 30° 40° 50° 60° 70° Latitudinal Position °N Lot. Fig. 2. Relationship between bottom temperatures (°C) and the latitudinal position of species of Galathealinum. (1930), Parker (1963), Sverdrup et al. (1942), and from the Oceanographic Atlas of the Polar Seas, Part II (U. S. Navy Hydro- graphic Office, 1957). The distribution of bot- tom temperatures in relation to the latitudinal position of the pogonophoran species is shown in Figure 2. The graph shows that the species closest to the Equator live in comparatively warmer water than those from more northern latitudes. The total temperature range for all the four species, however, is less than 4°C. This narrow range in temperatures, and the fact that these temperatures are from 0°— 4°C, seem to explain why these species occupy con- tinuously deeper waters from high to low lati- tudes. At high latitudes, the temperatures of the water from near the surface to great depths are similar and are within the range of value known at present for species in this genus. Hence, one could expect that species at high latitudes may occupy the total depth range of the genus provided that other ecological condi- tions, such as adequate food supplies (as indi- cated by Kirkegaard, 1956; 1956^), are satis- fied. The occurrence of the undetermined species from southern California at relatively shallower depths than expected (Fig. 1) and in relatively warmer water (Fig. 2) than expected is no- table. These basinal occurrences are rather anomalous. More data are needed to explain satisfactorily this particular case. The diameter of the tubes and the thickness of coarse (external) fibers of the tube walls for the species of Galathealinum were exam- ined and found to differ widely. The data are shown in Table 2. The tube diameter and coarse fiber thickness do not show a direct relationship to latitude. Their values increase from the Equator northward, reaching maxi- mum dimensions between 31°55/N and 32° 21' N, and decreasing from there toward high latitudes. It is interesting to note that the largest species and the one with the coarsest fibers occur at mid-latitudes (about 3 1 °- 32 °N) and not at high or low latitudes. But whether this is related to any ecological factors cannot be determined with the data at hand. Additional data are needed on species dis- tribution, temperature, and nature of bottom conditions to determine whether the indicated relationships are real or merely fortuitous. TABLE 2 Relationship of Tube Diameter and Coarse Fiber Thickness to Latitudinal Position of Species of Galathealinum Kirkegaard, 1956 Species Latitude Diameter of Tube (mm) Thickness, of Coarse Fibers M Galathealinum arcticum 69°32'N 1.33-1.95 1-2 Galathealinum brachiosum 54°23'N to 42°40'N 2. 0-2. 6 7-12 Galathealinum sp. indet. 32°2EN to 31°55'09"N 3. 0-4.0 — Galathealinum mexicanum 14°28'N to 12°20TST 1.96-2.5 15-22 Galathealinum bruuni 1°50'N 0. 8-2.0 2-4 Ecology of Gdldthedlinum — Adegoke 561 REFERENCES Adegoke, O. S. 1967. Pogonophora from the Northeastern Pacific: First Records from the Gulf of Tehuantepec, Mexico. Pacific Sci. 21(2) :188— 192. Bruun, A. F. 1957. General introduction to the reports and list of deep-sea stations. Galathea Rept. 1 :7-48. Emery, K. O. 1954. Source of water in basins off southern California. J. Mar. Res. 13:1- 21, 6 figs. I960. The Sea off Southern California. John Wiley and Sons, Inc., New York, London. 366 pp. and S. C. Rittenberg. 1952. Early diagenesis of California basin sediments in relation to origin of oil. Bull. Am. Assoc. Petrol. Geologists 36:735-806, 30 figs. Hartman, Olga. 1961. New Pogonophora from the eastern Pacific Ocean. Pacific Sci. 15(4) : 542-546. and J. L. Barnard. 1958. The benthic fauna of the deep basins off southern Cali- fornia. Allan Hancock Pacific Exped. 22(1) : 1-67, chart 1, pis. 1-2. I960. The benthic fauna of the deep basins off southern California. Al- lan Hancock Pacific Exped. 22(2) :217-284. Ivanov, A. V. 1961. New pogonophores from the eastern part of the Pacific Ocean. 1. Gdldthedlinum brdchiosum sp. n. Zool. Zh. 40:1378-1384. 1963. Pogonophora. (Translated from the Russian and edited by D. B. Carlisle with additional material by Eve C. Southward.) Consultant Bureau, New York. 479 pp. Jagersten, G. 1956. Investigations on Sibog- Unum ekmdni n. sp., encountered in Skager- rak, with some general remarks on the group Pogonophora. Zool. Bidr. Uppsala 31:211- 248. Kirkegaard, J. B. 1956. Pogonophora. Gd- ldthedlinum bruuni n. gen., n. sp., a new representative of the class. Galathea Rept. 2:79-83, 2 text- figs. 1956^. Pogonophora. First records from the eastern Pacific. Galathea Rept. 2:183-186. 1958. Records of the group Pogonoph- ora in the Skagerrak. Nature, London 181 (4615) :1086-1087. Kramp, P. L. 1957. Stephdnoscyphus (Scypho- 20a). Galathea Rept. 1:173-188. Marmer, H. A. 1930. The Sea. D. Appleton and Co., New York, London. 312 pp. Parker, R. H. 1963. Zoogeography and Ecology of Some Macro-Invertebrates, Par- ticularly Mollusks, in the Gulf of California and the Continental Slope off Mexico. Vi- densk. Medd. fra Dansk naturh. Foren. Bd. 126, 178 pp., 15 pis. Southward, A. J. 1958. Abundance of Pogo- nophora. Nature, London 182(4630) :272. and Southward, E. C. 1958. Pogo- nophora from the Atlantic. Nature, London 181(4623) :1607. 1963. Notes on the biology of some Pogonophora. J. Mar. Biol. Assoc. U.K. 43:57-64. Southward, E. C. 1962. A new species of Gdldthedlinum (Pogonophora) from the Canadian Arctic. Can. J. Zool. 40:385-389. Sverdrup, H. U., M. W. Johnson, and R. H. Fleming. 1942. The Oceans. 1087 pp. Prentice-Hall, Inc., Englewood Cliffs, New Jersey. U. S. Navy Hydrographic Office. 1957. Oceanographic Atlas of the Polar Seas, Part II. Arctic. H. O. Publ. no. 705. Washington, D. C. Notes on the Systematic Status of the Eels Neenchelys and Myroconger Gareth J. Nelson1 Neenchelys Bohlke (I960) suggested that eels of the genus Neenchelys possibly have overlapping branchiostegal rays and that, if they did, they should be assigned to the family Ophichthidae. Nelson (1966^) described the osteology of Neenchelys buitendijki, confirming the presence of overlapping branchiostegal rays, and for this and other reasons referred the genus Neenchelys to the family Ophichthidae, subfamily Eche- linae. The present report, based on an ex- amination of the holotype of Neenchelys micro- tret us, confirms the presence of overlapping branchiostegal rays in the type species of Neen- chelys. Like those of N. buitendijki (Nelson, 1966a, fig. 2 A), those of N. microtretus in- clude six rays articulating with the dorsal por- tion of the ceratohyal and more than 25 others widely overlapping in the midline. Myroconger This genus and the family it represents ap- parently are known only from the holotype of Myroconger compressus. The specimen had been partly dissected, leaving the gill arches exposed, which allowed the following observa- tions to be made: third and fourth upper pharyngeal tooth plates separate; first and sec- ond pharyngobranchials absent, the third sup- porting the tooth plates; basibranchials absent; independent rodlike hypobranchials in arches one-three, those of the third cartilaginous; fourth ceratobranchials not extended anteriorly, not separating the third arches of either side; fifth ceratobranchials apparently absent; ventral parts of the arches not meeting in the midline. Myroconger has the frontal bones separated by a suture and therefore belongs to the anguil- loid lineage of Regan (1912), including the Heterenchelidae, Anguillidae, Moringuidae, 1 Department of Paleozoology, Swedish Museum of Natural History, Stockholm 50, Sweden. Manuscript received January 20, 1967. Present address: Ameri- can Museum of Natural History, New York, N.Y. Xenocongridae, Dysomminidae, and Muraeni- dae (Nelson, 19 66b). In completely lacking basibranchials, the arches of Myroconger differ from those of Heterenchelys, Anguilla, and Moringua, but resemble those of xenocongrids, Dysommina, and muraenids. In lacking a second pharyngobranchial they are unlike xenocon- grids, but resemble Dysommina and muraenids. Like that of Dysommina the fourth arch of Myroconger is not appreciably enlarged and "pharyngeal jaws” like those of muraenids do not occur. Thus, the arches of Myroconger are most like those of Dysommina. The most nota- ble differences include the presence in Myro- conger of third hypobranchials (a primitive feature) and the apparent absence of fifth ceratobranchials (an advanced one). What could be learned of the pharyngeal musculature also suggests a relationship with the more advanced eels of the anguilloid lineage, for a subpharyngealis occurs, as it does at least in Moringua, Kaupichthys, and muraenids, and retractor muscles have a small area of origin on the vertebral column, foreshadowing the large area of origin in some muraenids (Nelson, 1967). These observations of Myroconger complete a review of gill arch structure for the families of anguilloid eels (Nelson, 1966b) . Within this group, on the basis of gill arch structure there seem to be three main lines of specializa- tion, each characterized by reduction of the gill arch skeleton: one leads toward the Morin- guidae, another toward the Muraenidae, the other toward the Cyemidae. If the anguilloid eels are given the status of a suborder, these lines of specialization could be given the status of superfamilies. However, on the basis of gill arch structure alone it is difficult to distinguish between generalized members of these different lines, or to decide which if any Recent forms can be considered generalized muraenoids. Consequently, the following synopsis is offered more as a working hypothesis than as a final classification: 562 Neenchelys and Myro conger — Nelson 563 Caudal fin continuous with dorsal and anal; no pelvic fins; frontal usually paired suborder Anguilloidei a. Jaws not produced; gill arch skeleton in- cluding at least rudimentary basibranchials superfamily Anguilloidae (including families Heterenchelidae, An- guillidae, Moringuidae) b. Jaws not produced; gill arch skeleton with- out basibranchials .................... .............. superfamily Muraenoidae (including Xenocongridae, Dysomminidae, Myrocongridae, Muraenidae) c. Jaws produced; gill arch skeleton with or without basibranchials ............ superfamily Nemichthyoidae (including Serrivomeridae, Nemichthyidae, Cyemidae) Observations on type specimens were made through the courtesy of Dr. P. H. Greenwood at the British Museum (Natural History) while I was on an nsf postdoctoral fellowship. REFERENCES Bohlke, J. E. I960. A New Ophichthid Eel of the Genus Pseudomyrophis from the Gulf of Mexico. Notulae Naturae 329, 8 pp., 2 figs. Nelson, G. J. 1966^. Osteology and relation- ships of the eel, Neenchelys buitendijki. Copeia 1966(2) :321— 324, 2 figs. 1966A Gill arches of teleostean fishes of the order Anguilliformes. Pacific Sci. 20(4) :391-408, 58 figs. 1967. Branchial muscles in repre- sentatives of five eel families. Pacific Sci. 21(3): 348-363. Regan, C. T. 1912. The osteology and classi- fication of the teleostean fishes of the order Apodes. Ann. Mag. Nat. Hist., Ser. 8, 10: 377-387, 2 figs. Record of a Lancelet from Hawaii L. G. Eldredge1 The species Epigonichthys maldivensis (For- ster Cooper, 1903) was originally described from 45 Maidive Island specimens as Hetero- pleuron maldivense. Parker (1904) also re- corded 12 other specimens from the same island group. Both Forster Cooper’s and Parker’s specimens were collected from depths of 15-20 fathoms. No further specimens have been found. The generic status is somewhat confused, for in Franz’s (1922) generic revision of the cephalochordates, the species appears as Asym- metron maldivense. However, in the same year Hubbs (1922) listed the species in a world- wide review of the group as Epigonichthys mal- divensis, establishing it in his asymmetrical- form family, Epigonichthyidae. (In neither paper is reference given to the other one.) The single specimen from Hawaii was col- lected alive by R. E. Johannes and the author via the "Pele” dredge through coarse sand off Barber’s Point, Oahu, at a depth of 16-20 fathoms on March 18, 1962. This report con- stitutes the first record of a lancelet (amphi- oxus) in Hawaiian waters. With two exceptions its characteristic mea- surements fall within the ranges of Forster Cooper’s specimens as analyzed by Punnett (1903) as well as within those given by Franz (1922), who used a combination of the Forster Cooper-Punnett and the Parker characteristics. The following table compares the characteristics of these specimens. The two features which differ from those previously described are the greater number of myotomes from the anus to the tip of the tail, and the smaller number of gonads, a peculiarity 1 Department of Biology, College of Guam, Agana, Guam. Manuscript received April 3, 1967. CHARACTER PUNNETT (1903) FRANZ (1922) OAHU SPEC. Length (mm) 18-30 16-30 23 No. Myotomes Total 70-76 70-76 73 Head to atriopore 42-46 45 43 Atriopore to anus 15-17 16 15 Anus to tip of tail 11-14 12 15 Gonads 23-30 25 18 which might be explained by the intermittent location of the 18 gonads lying irregularly beneath at least 22 myotomes. Forster Cooper’s original figure shows what looks like at least one gonad per myotome. REFERENCES Forster Cooper, C. 1903. Cephalochorda. I. Systematic and anatomical account, pp. 347- 360. In: J. S. Gardiner, ed., The Fauna and Geography of the Maidive and Laccadive Archipelagoes, Vol. 1. Cambridge Univ. Press. Franz, V. 1922. Systematische Revision der Akranier. Jena Z. Naturwiss. 58:369-452. Hubbs, C. L. 1922. A list of the lancelets of the world with diagnoses of five new species of Branchiostoma. Occas. Pap. Mus. Zook Univ. Mich. 105:1-16. Parker, G. H. 1904. Maidive cephalochordates with the description of a new species from Florida. Bulk Mus. Comp. Zook 46(2) :39- 52. Punnett, R. C. 1903. Cephalochorda. II. Note on meristic variation in the group, pp. 361- 367. In: J. S. Gardiner, ed., The Fauna and Geography of the Maidive and Laccadive Archipelagoes, Vol. 1. Cambridge Univ. Press. 564 News Note1 THE ADDITION OF seven tide stations in Chile will give Alaska, Oregon, Washington, Hawaii, and many Pacific Basin nations up to two or more hours additional warning time when a de- structive tsunami is generated off the Chilean coast. This was disclosed by Mark G. Spaeth of the Environmental Science Services Administration’s Coast and Geodetic Survey. Spaeth, a geophysicist in the Office of Seismology and Geomagnetism, made the disclosure following his return from an 18-day trip to South America designed to improve the Tsunami Warning System in the Pacific. The System is operated by the Coast and Geodetic Survey in Honolulu. Spaeth conferred in Valparaiso with officials of the Chilean Hydrographic Institute, and agree- ment was reached to supply reports on tide conditions from seven additional tide stations to the C&GS Honolulu Observatory, which issues the warnings. Until now, there were only two participating tide stations in the Tsunami Warning System along the entire coast of South Amer- ica— at Valparaiso, Chile, and Callao, Peru. "This should greatly reduce the time needed to confirm the existence of tsunamis originating along the South American coast," said Spaeth, "and provide up to two or more hours additional warning time to people throughout the Pacific Basin when a destructive wave is generated." A I960 tsunami off the Chilean coast caused deaths and extensive damage in Chile, Hawaii, the Philippines, Japan, and Okinawa. In spite of six hours’ warning, 61 persons were killed in the Hawaiian city of Hilo. All Chilean coastal towns between the 36th and 44th parallels were destroyed or severely damaged by earthquake and the ensuing tsunami. In California, a half- million dollars’ damage was caused to harbor installations and ships at Los Angeles, San Diego, Cresent City, and Half Moon Bay. Crescent City was also the scene of widespread destruction from a tsunami generated by the March 1964 earthquake in Alaska. Eleven persons lost their lives in this community. Spaeth also conferred with scientists in the Panama Canal Zone, Peru, and Ecuador. As a re- sult, further steps to strengthen the Warning System may be taken in these areas. "Tsunami" is the Japanese name for a seismic sea wave, the destructive oceanic offspring of earthquakes and volcanic eruptions. Although frequently, and mistakenly, called tidal waves, they are not related to the tides. The phenomenon is a series of ocean waves, not unlike those made when a pebble is dropped into a pond, which travel up to 600 miles or more per hour. In the open ocean the crests of the waves are sometimes hundreds of miles apart and always only a few feet high. As the tsunami nears the coast, the speed diminishes and the waves increase in height. They cannot be seen from the air, nor felt aboard ships in deep water. This is why tide stations are essential in the Warn- ing System — to report to the System Headquarters at Honolulu any change in water level, following large earthquakes, that could indicate a tsunami has been generated. The Honolulu Office in turn alerts Civil Defense agencies in communities the tsunami may strike so the low-lying areas can be evacuated. The Tsunami Warning System was established by the Coast and Geodetic Survey following the disastrous Aleutian tsunami of April 1, 1946, which struck the Hawaiian Islands without warn- ing, leaving 159 persons dead and about $25 million property damage. But this was the last tsunami to strike Hawaii without warning, because the Coast and Geodetic Survey developed and installed a network of seismographs around the Pacific to detect the earthquakes which cause tsunamis. 1 From: Earthquake Information Bulletin 1(3): 1-4 (1967). 565 566 PACIFIC SCIENCE, Vol. XXI, October 1967 Since only a very few of the thousands of earthquakes which occur annually cause tsunamis, it was also necessary to include tide stations in the Warning System to detect the waves. The Coast Survey’s network of tide stations was expanded and a tsunami detector was developed which filtered out tidal and wind waves. Tsunami traveltime charts were prepared to provide a rapid and easy method of estimating tsunami arrival times at tide stations in the Warning System. The military services and Federal Aviation Agency initially provided communication facilities. As the Warning System expanded, facilities of the Department of Defense and foreign governmental agencies were utilized. The Warning System began operation in August, 1948. Initially it consisted of the C&GS seis- mological observatories at College and Sitka, Alaska; Tucson, Arizona; and Honolulu, Hawaiif and nine tide stations. Today the System has 17 participating seismic observatories and 30 tide stations (see map). In addition to the United States and Chile, the Tsunami Warning System now provides warnings to Tahiti, Japan, Taiwan, Fiji, Hong Kong, New Zealand, Western Samoa, Canada, and the Philippines. Index to Volume XXI Author Index Adegoke, Oluwafeyisola S.: Pogonophora from the Northeastern Pacific: First Records from the Gulf of Tehuantepec, Mexico, 188-192 Notes on the Ecology of the Pogonophoran Genus Galathealinum Kirkegaard, 1956, 558-561 Alvarino, Angeles: A New Siphonophora, Vogtia kuruae n. sp., 236- 240 Bathymetric Distribution of Chaetognatha, Sipho- nophorae, Medusae, and Ctenophorae off San Diego, California, 474-485 Arnold, B. C: A Hitherto Unrecorded Midge Gall of Myrsine australis (A. Rich.) Allan, 115-118 Bowman, Thomas E.: The Planktonic Shrimp, Lucifer cbacei sp. nov., (Sergestidae: Luciferinae), the Pacific Twin of the Atlantic Lucifer faxoni, 266-271 Boyd, Carl M.: The Benthic and Pelagic Habitats of the Red Crab, Pleuroncodes planipes, 394-403 Breese, Paul: see Hunsaker and Breese Bruska, Gary J.: The Ecology of Pelagic Amphipoda I. Species Accounts, Vertical Zonation and Mi- gration of Amphipoda from the Waters off Southern California, 382-393 II. Observations on the Reproductive Cycles of Several Pelagic Amphipods from the Waters off Southern California, 449-456 Burch, J. B., and R. Natarajan: Chromosomes of Some Opisthobranchiate Mollusks from Eniwetok Atoll, Western Pacific, 252-259 Burch, John Q., and Rose L. Burch: The Family Olividae, 503-522 Burch, Rose L.: see Burch and Burch Campbell, Richard D.: Monobrachium parasitum, a One-Tentacled Hy- droid, Collected at Vancouver Island, 431 Chamberlain, Theodore K.: see Stearns and Chamberlain Costlow, John D., Jr., and Elda Fagetti: The Larval Development of the Crab, Cyclograp- sus cinereus Dana, under Laboratory Conditions, 166-177 Eldredge, L. G.: Record of a Lancelet from Hawaii, 564 Fagetti, Elda: see Costlow and Fagetti Fellows, David P., and A. Earl Murchison: A Noninjurious Attack by a Small Shark, 150-151 Furumoto, Augustine S.: A Study of the Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964 Part II: Analysis of Rayleigh Wave, 310-316 Gooding, Reginald M., and Magnuson, John J.: Ecological Significance of a Drifting Object to Pelagic Fishes, 486-497 Hamamoto, Susan T.: see Hohl and Hamamoto Hines, Judith, and Ron Kenny: The Growth of Arachnoides placenta (L.) (Echi- noidea), 230-235 Hohl, Hans R., and Susan T. Hamamoto: Reversal of Ethionine Inhibition by Methionine during Slime Mold Development, 534-538 Holden, John C: Late Cenozoic Ostracodes from the Drowned Ter- races in the Hawaiian Islands, 1-50 Hunsaker, Don, and Paul Breese: Herpetofauna of the Hawaiian Islands, 423-428 Jerde, Charles W.: A Comparison of Euphausiid Shrimp Collections Made with a Micronekton Net and a One-Meter Plankton Net, 178-181 Johnson, Charles G.: see Krivoy, Johnson, and Koyanagi Kenny, Ron: see Hines and Kenny Knight, Margaret D.: The Larval Development of the Sand Crab Emerita rathbunae Schmitt (Decapoda, Hippidae), 58-76 Knudsen, Jens W.: Trapezia and Tetralia (Decapoda, Brachyura, Xanthidae) as Obligate Ectoparasites of Pocil- loporid and Acroporid Corals, 51-57 Komaki Yuzo: On the Surface Swarming of Euphausiid Crusta- ceans, 433-448 Koyanagi, Robert Y.: see Krivoy, Johnson, and Koyanagi Krejsa, Richard J.: The Systematics of the Prickly Sculpin, Cottus asper Richardson, a Polytypic Species Part I. Synonymy, Nomenclatural History, and Distribution, 241-251 567 568 PACIFIC SCIENCE, VoL XXI, October 1967 Part II. Studies on the Life History, with Espe- cial Reference to Migration, 414-422 Krivoy, Harold L., Charles G. Johnson, and Robert Y. Koyanagi: An Unusual Example of Pseudoseisms Resulting from Military Exercises, 119-128 Lanzing, W. J. R.: A possible Relation between the Occurrence of a Dendritic Organ and the Distribution of the Plotosidae (Cypriniformes), 498-502 Little, Georgiandra: Chromatophore Responses in Relation to the Pho- toperiod and Background Color in the Hawaiian Ghost Crab, Ocypode ceratophthalma (Pallas), 77-84 Magnuson, John J.: see Gooding and Magnuson Malahoff, Alexander: Gravity and Geological Studies of an Ultramafic Mass in New Zealand, 129-149 Mueller-Dombois, D., and C. H. Lamoureux: Soil-Vegetation Relationships in Hawaiian Kipu- kas, 286-299 Murchison, A. Earl: see Fellows and Murchison Natarajan, R.: see Burch and Natarajan Nelson, Gareth J.: Branchial Muscles in Representatives of Five Eel Families, 348-363 Notes on the Systematic Status of the Eels Neen- chelys and Myroconger , 562-563 Osborne, Lynette D.: Comparative Decay Resistance of Twenty-five Fijian Timber Species in Accelerated Laboratory Tests, 539-549 Pararas-Carayannis, George: A Study of the Source Mechanism of the Alaska Earthquake and Tsunami of March 27, 1964. Part 1. Water Waves, 301-310 Provenzano, Anthony J., Jr.: The Zoeal Stages and Glaucothoe of the Tropical Eastern Pacific Hermit Crab Trizopagurus mag- ni ficus (Bouvier, 1898) (Decapoda; Diogenidae), Reared in the Laboratory, 457-473 Rehder, Harald A.: A New Genus and Two New Species in the Fami- lies Volutidae and Turbinellidae (Mollusca: Gastropoda) from the Western Pacific, 182-187 Roger, Claude: Note on the Distribution of Euphausia eximia and E. gibboides in the Equatorial Pacific, 429-430 Rogers, Terence A.: see Sather and Rogers Rosenblatt, Richard H.: The Osteology of the Congrid Eel Gorgasia punc- tata and the Relationships of the Heterocon- grinae, 91-97 St. John, Harold: Revision of the Genus Pandanus Stickman Part 21. The Pandanus monticola Group in Queens- land, Australia, 272-281 Part 22. A new Species (Section Hombronia) from New Caledonia, 282-285 Part 23. Three Australian Species of Pandanus, 523-530 Part 24. Seychellea, a New Section from the Seychelles Islands, 531-532 Part 25. Pandanus tectorius var. sinensis Warburg, 533 Salmon, Michael: Acoustical Behavior of the Menpachi, Myripristis berndti, in Hawaii, 364-381 Sather, Bryant T.: Studies in the Calcium and Phosphorus Metabolism of the Crab, Podophthalmus vigil (Fabricius), 193-209 Sather, Bryant T., and Terence A. Rogers: Some Inorganic Constituents of the Muscles and Blood of the Oceanic Skipjack, Katsuwonus pelamis, 404-413 Schwab, Robert G.: Overt Responses of Polychoerus carmelensis (Tur- bellaria: Acoela) to Abrupt Changes in Ambient Water Temperature, 85-90 Stearns, Harold T., and Theodore K. Chamber- lain: Deep Cores of Oahu, Hawaii and Their Bearing on the Geologic History of the Central Pacific Basin, 153-165 Steele, Carol Wright: Fungus Populations in Marine Waters and Coastal Sands of the Hawaiian, Line, and Phoenix Islands, 317-331 Stone, Benjamin C: The Flora of Romonum Island, Truk Lagoon, Caroline Islands, 98-114 Notes on the Hawaiian Flora, 550-557 Strasburg, Donald W.: Observations on the Biology of the Lousefish, Phtheirichthys lineatus (Menzies), 260-265 Swartz, L. G.: Distribution and Movements of Birds in the Bering and Chukchi Seas, 332-347 Unnithan, R. Viswanathan: On Some Gastrocotyline (Monogenoidean) Para- sites of Indian Clupeoid Fishes, Including Three New Genera, 210-229 Index 569 Subject Index acoustical behavior of Myripristis herndti, 364-381 Alaska earthquake and tsunami of March 27, 1964, source mechanism of, 301-316 Rayleigh wave and analysis of, 311—316 water waves, 301-310 Amphipoda from southern California, species ac- counts, vertical zonation, and migration of, 382- 393 reproductive cycles of, 449-456 Anguilloidei, branchial muscles in, 348-363 Arachnoides placenta, growth of, 230-235 Australia, three Pandanus species in, 523-530 Benthovoluta gracilior n. sp., 185-186 Bering Sea, birds in, 332-347 birds in the Bering and Chukchi seas, distribution and movements of, 323-347 branchial muscles in five eel families, 348-363 calcium and phosphorus metabolism of Podophthal- rnus vigil, 193-209 Cenozoic ostracodes from the Hawaiian Islands, 1-50 Central Pacific Basin, geologic history of, 153-165 Chaetognatha, distribution of, 474-485 chromatophore responses in Hawaiian ghost crab, 77-84 chromosomes of opisthobranchiate mollusks from Eniwetok, 252-259 Chukchi Sea, birds in, 332-347 clupeoid fishes, parasites of, 210-229 congrid eel, osteology of, 91-97 corals, obligate parasites of, 51-57 Cottus asper, 241-251, 414-422 crustaceans, euphausiid, surface swarming of, 433— 448 Ctenophorae, distribution of, 474-485 Cyclograpsus cinereus, larval development of under laboratory conditions, 166-177 decay resistance of Fijian timber, 539-549 dendritic organ, relation of to distribution of Ploto- sidae, 498-502 drifting objects, ecological significance to fish, 486- 497 earthquake, see Alaska earthquake and tsunami Earthquake Information Bulletin, news note from, 565-566 ecological significance of drifting objects to fish, 485- 497 ecology of Galathealinum, 558—561 eels, branchial muscles in, 348-363 systematic status of Neenchelys and Myroconger, 562-563 Enter it a rathbunae, larval development of, 58-76 Eniwetok, chromosomes of mollusks from, 252-259 Epigonichthys maldivensis from Hawaii, 564 ethionine, in slime mold development, 534-538 Euphausia eximia, distribution of, 429-430 Euphausia gibboides, distribution of, 429-430 Euphausia pacifica in Japanese waters, 433-448 euphausiid crustaceans, surface swarming of, 433— 448 euphausiid shrimps, comparison of collection meth- ods for, 178-181 Fijian timber, decay resistance of, 539-549 first records of Pogonophora from the Gulf of Tehuantepec, Mexico, 188-192 flora of Romonum Island, Caroline Islands, 98-114 fungi, marine, of various Pacific island groups, 317- 331 Galathealinum, ecology of, 558-561 Galathealinum mexicanum n. sp., 188-192 gastrocotyline parasites of Indian clupeoid fishes, 210-229 geologic history of Central Pacific Basin, 153-165 glaucothoe of hermit crab, 457-473 Gorgasia punctata, osteology of, 91-97 gravity and geology of an ultramafic mass in New Zealand, 129—1 49 Gulf of Tehuantepec, Pogonophora from, 188-192 Hawaii, lancelet from, 564 Hawaiian flora, taxonomic and nomenclatural notes on, 550-557 Hawaiian ghost crab, chromatophore responses in, 77-84 Hawaiian Islands, fungi in coastal waters and sands of, 317-331 herpetofauna of, 423-428 late Cenozoic ostracodes from, 1-50 hermit crab, larval development of, 457-473 herpetofauna of the Hawaiian Islands, 423-428 Heterocongrinae, relationships of, 91-97 hydroid, one-tentacled, from Vancouver Island, 431 India, parasites of clupeoid fishes from, 210-229 inorganic constituents in muscles and blood of oceanic skipjack, 404-413 Katsuwonus pelamis, inorganic constituents in, 404- 413 kipukas, Hawaiian, soil-vegetation relationships in, 286-299 lancelet recorded from Hawaii, 564 larval development of Cyclograpsus cinereus, 166- 177 of Emerita rathbunae, 58-76 of Trizopagurus magnipcus, 457-473 Line Islands, fungi in coastal waters and sands of, 317-331 lousefish, biology of, 260-265 Lucifer chacei n. sp., description of and comparison with L. faxoni, 266-271 Lucifer faxoni, comparison with L. chacei n. sp., 266-271 marine fungi, populations in Pacific islands, 317-331 Medusae, distribution of, 474-485 menpachi in Hawaii, acoustical behavior of, 364-381 570 PACIFIC SCIENCE, Vol. XXI, October 1967 methionine, in slime mold development, 534-538 micronekton net, used in collection of euphausiid shrimps, 178-181 midge gall of Myrsine australis, 115-118 military exercises, pseudoseisms resulting from, 119- 128 mollusks, opisthobranchiate, chromosomes of, 252- 259 Monobrachium paras i turn from Vancouver Island, 431 Monogenoidea parasitic on Indian clupeoid fishes, 210-229 muscles, branchial, in five eel families, 348-363 Myripristis berndti, acoustical behavior of, 364-381 Myroconger, systematic status of, 562-563 Myrsine australis, mide gall of, 115-118 Neencbelys, systematic status of, 562— 563 New Caledonia, new species of Pandanus from, 282- 285 new genera of gastrocotyline parasites, 210-229 new genus of Volutidae, 182-183 new species of gastrocotyline parasites, 210-229 of Pandanus, 279-281, 282-285, 527 of planktonic shrimp, 266-271 of Pogonophora, 188-192 of Siphonophora, 236-241 of Turbinellidae, 185-186 of Volutidae, 182-183 New Zealand, ultramafic mass in, 129-149 news note, from Earthquake Information Bulletin, 565-566 northeastern Pacific, Pogonophora from, 188-192 Oahu, deep cores of and their bearing on geologic history, 153-165 Olividae, 503-522 opisthobranchiate mollusks, chromosomes of, 252- 259 osteology of congrid eel, 91-97 ostracodes, Cenozoic, from Hawaii, 1-50 Pandanus, of Australia, 523-530 from New Caledonia, 282-285 in Queensland, 272-281 from Seychelles Islands, 531-532 tectorius var. sinensis, 533 Pandanus, revision of, Part 21, 272-281 Part 22, 282-285 Part 23, 523-530 Part 24, 531-532 Part 25, 533 pelagic Amphipoda, ecology of, 382-393 reproductive cycles of, 449-456 pelagic fish, ecology of, 486-497 Phoenix Islands, fungi in coastal waters and sands of, 317-331 phosphorus and calcium metabolism of Podophthal- mus vigil, 193-209 photoperiod and background color, chromatophore responses to, 77-84 Phtheirichthys lineatus, biology of, 260-265 plankton net, one-meter, used in collection of euphau- siid shrimps, 178-181 Pleuroncodes planipes, habitats of, 394-403 Plotosidae, distribution of, 498-502 Podophthalmus vigil, calcium and phosphorus me- tabolism of, 193-209 Pogonophora from the northeastern Pacific, 188-192 Polychoerus carmelensis, responses to changes in water temperature, 85-90 prickly sculpin, synonymy, nomenclatural history, and distribution of, 241-251 life history and migration of, 414-422 pseudoseisms resulting from military exercises, 119- 128 Queensland, Pandanus monticola in, 272-281 Rayleigh wave, in study of Alaska earthquake, 311- 316 red crab, benthic and pelagic habitats of, 394-403 reproductive cycles of pelagic amphipods, 449-456 Romonum Island, flora of, 98-114 sand crab, larval development of, 58-76 Seychelles Islands, new section of Pandanus from, 531-533 shark, attack by, 150-151 shells of family Olividae, 503-522 shrimps, euphasiid, collection methods for, 178-181 Sigaluta n. gen., 182 pratasensis n. sp., 182-183 Siphonophora, new species of, 236-240 Siphonophorae, distribution of, 474-485 skipjack, constituents in muscles and blood of, 404- 413 slime molds, effect of ethionine and methionine on development of, 534-538 soil-vegetation relationships in Hawaiian kipukas, 286-299 southern California, pelagic Amphipoda from, 382- 393, 449-456 surface swarming of euphasiid crustaceans, 433-448 Petr alia, as parasite of corals, 51-57 tide stations, new, in Chile, 565-566 timber, Fijian, decay resistance of twenty-five species of, 539-549 Trapezia, as parasite of corals, 51-57 Trizopagurus magni ficus, larval development of, 457- 473 tsunami, see Alaska earthquake and tsunami Tsunami Warning System, 565-566 Turbinellidae, new species in, 183—186 ultramafic mass in New Zealand, 129-149 Vancouver Island, one-tentacled hydroid from, 431 Vogtia kuruae n. sp., 236-240 Volutidae, new genus and new species in, 182-183 water temperature, responses of Polychoerus to changes in, 85-90 water waves associated with Alaska earthquake, 301- 310 zoeal stages of hermit crab, 457-473 Manuscript Form. Manuscripts should be typed on one side of standard-size, white bond paper and double-spaced throughout. Pages should be consecu- tively numbered in upper right-hand corner. Sheets should not be fastened together in any way, and should be mailed flat. Inserts should be either typed on separate sheets or pasted on proper page, and point of insertion should be clearly indicated. Original copy and one carbon copy of manuscript should be submitted. The author should retain a car- bon copy. Although due care will be taken, the editors cannot be responsible for loss of manuscripts. Introduction and Summary. It is desirable to state the purpose and scope of the paper in an introductory paragraph and to give a summary of results at the end of the paper. Dictionary Style. It is recommended that authors fol- low capitalization, spelling, compounding, abbrevia- tions, etc., given in Webster's New International Dic- tionary (unabridged), second edition; or, if desired, the Oxford Dictionary. Abbreviations of titles of pub- lications should, if possible, follow those given in World List of Scientific Periodicals. Footnotes. Footnotes should be used sparingly and never for citing references (see later). When used, footnotes should be consecutively numbered by supe- rior figures throughout the body of the paper. Foot- notes should be typed in the body of the manuscript on a line immediately below the citation, and sepa- rated from the text by lines running across the page. Citations of Printed Sources. All references cited should be listed alphabetically by author at the end of the paper, typed double-spaced. References to books and to papers in periodicals should conform to the following models: Batzo, Roderick L., and J. K. Ripkin. 1849- A Treatise on Pacific Gastropods. Rice and Shipley, Boston, vii + 326 pp., 8 figs., 1 map. Crawford, David L. 1920^. New or interesting Psyllidae of the Pacific Coast (Homop.). Proc. Hawaii. Ent. Soc. 4(l):12-l4. 1920&. The sandalwoods of Hawaii. Proc. Hawaii. Ent. Soc. 4(2) :374-375, 13 pis. In the text, sources should be referred to by author, date, and page, as follows: ‘'It was noted (Rock, 1916:18) that . . ” or "Rock (1916:21-24) says ...” Quotations. Quoted matter of fewer than five printed lines (about 200 characters) should be given in the text in the usual form, using double quote marks. Longer quotations should be set flush with left mar- gin. The author is responsible for the accuracy of quoted material. Numbers. Decimals, measurements, money, percent- ages, time; enumerations in which any figure is 10 or over; and isolated enumerations of 10 and over should be given in Arabic figures, rather than spelled out, except when the number begins a sentence. ILLUSTRATIVE MATTER Only the minimum number of illustrations required to supplement the text will be accepted by the editors. Reproduction costs of illustrations in excess of the number allowed by the editors will be paid by the author. Artwork for illustrations should be 8x/2 x 11 inches or smaller, and it should accompany manuscript, on separate sheets. Often more valuable than a photo- graph is a good line drawing. Figures and Graphs. Copy for figures and graphs should always be drawn large enough to allow for at least one-third reduction by the engraver. Copy should consist of carefully prepared line drawings in one color only, drawn in India ink on plain white draw- ing paper or tracing cloth. Co-ordinate paper with lines in light blue (a color which will not show in a photograph) may be used; but co-ordinates which should appear in the finished graph must be drawn in India ink. If original figures may not be conven- iently submitted with manuscript, duplicate rough sketches or photographic prints may be furnished to aid the editors in their decisions. It is strongly urged that an indication of scale be incorporated as a part of all drawings in which mag- nification and size are critical considerations. Photographs. Photographs should be chosen for clarity in portraying essential information. They should be printed for contrast, on glossy paper, and should be sent unmounted. They should be identified with serial number written in soft pencil on the back to corre- spond with list of captions. Illustrations will be returned to the author. Tables. Tabular matter should be kept to a minimum. Each table, prepared to conform with Pacific Science style, should be typed on a separate page, and its posi- tion indicated on the manuscript. Mathematical Formulas. Complicated formulas cannot be set by the printers. Authors should submit them as illustrations. Captions. Readily identifiable captions for figures, graphs, photographs, and other illustrative matter should be supplied on a separate page. PROOF Proof should be corrected immediately and returned at once to John S. Howe, Jr., assistant to the editors. Authors are reminded that the editors will allow only a minimum number of corrections on galley proof. Additions to the printed text and changes in style and content are not permitted. All queries on proof should be answered. If cor- rected proof is not received within four weeks after being sent to the author, author’s changes cannot be accepted. REPRINTS Reprints or separates should be ordered on the form provided and returned with author’s proof. All correspondence concerning separates must be directed to the printer, HefFernan Press Inc., 35 New Street, Worcester, Massachusetts 01605. The publisher does not keep a reprint file.