PROC. BIOL. SOC. WASH. 102(4), 1989, pp. 826-865 A REVISION OF THE GENUS ASPIDOSIPHON (SIPUNCULA: ASPIDOSIPHONIDAE) Edward B. Cutler and Norma J. Cutler Abstract. — The 64 putative species of the sipunculan genus Aspidosiphon and the morphological characters used to differentiate them are critically reviewed. The monograph of Stephen & Edmonds (1972) is used as a starting place and all changes made in the intervening years are reiterated here. All available type material was studied and new collections of Hawaiian and Caribbean material are used to analyze within-deme variation. Hook and anal shield morphology are determined to be broadly useful at the species level, four characters (lon- gitudinal muscle layer, retractor muscle origins, caudal shield, nephridia length) in a more restricted manner to separate subgroups, and three (introvert/trunk angle, bifurcated anterior spindle muscle, loosely wound gut coil) are useful in special cases. A new subgenus, Aspidosiphon (Akrikos), is proposed for those five species lacking hooks in rings. A key to, and a discussion of, the 19 remaining species (plus one reduced to subspecies) with the newly designated synonyms are presented. A brief statement of the distribution of each species is given. An overall summary of the zoogeography and habitat shows more endemic species are found in the warm water regions of the Atlantic Ocean than in the Indo-West Pacific, and that only 42% of Aspidosiphon species live in coral or rock. This continues our revisionary series on the species of sipunculan worms (e.g., Cutler & Cutler 1985a, b, 1986, 1987b, 1988). With this work we complete our examination of all the genera in this phylum except Phas- colosoma, which is in preparation. The monograph of Stephen & Edmonds (1972) is the starting place for this work (48 species names). Also included (Table 1) are the 1 1 species erected since that time, the two species transferred into this genus, and the three resurrected names. The genus Aspidosiphon was erected by Diesing in 1851, and was placed in its own family, Aspidosiphonidae, by Baird (1868) in the order Aspidosiphoniformes Cutler & Gibbs (1985). The name Par aspidosiphon was proposed by Stephen in 1 964 as a genus for those species with the longitudinal mus- cle layer separated into bundles. Cutler (1973) reduced it to subgeneric rank and this has been followed by most other authors (see discussion below). Whenever possible we have obtained type material to verify the original descriptions. In several cases we have made detailed ob- servations on series of recently collected in- dividuals to evaluate better the traditionally used morphological characters. Recent col- lecting trips to Hawaii, Curacao and Ven- ezuela (Cumana and Isla de Los Roques) have greatly facilitated this effort. The op- portunity to observe living material is in- valuable. Parts of these collections will be deposited in the National Museum of Nat- ural History, Washington, D.C. as reference material. We first discuss the morphological char- acters in light of our recent analyses, then discuss those taxa not clearly belonging to this genus. Following are a key to all the species we consider valid, a section where VOLUME 102, NX'MBER 4 827 each of these species is discussed including a synonomy. a discussion of any newly added junior synonyms, and a summary of the known distribution of each species. A short zoogeographical summary of the genus con- cludes this work. For clarity in the Morphological Char- acters section, the recent work of Saiz Sa- linas (1984) needs to be mentioned here. His redescription of Quatrefages' 1865 species from the Paris Museum has led to the elevation of A. coyi and A. laevis as se- nior synonyms of the more familiar A. trun- catus for the former and the large A. cum- ingii klunzingeri complex for the latter. In both cases holotypes are now available to science, which is not the case for the more familiar names. In some ways this action is analogous to that of Rice & Stephen (1970) where they resurrected the older and long unused names of Gray and Baird. The following abbreviations are used in the text for the museums from which we borrowed material: American Museum of Natural History, New York (AMNH); Brit- ish Museum (Natural History). London (BMNH); Museum National d'Histoire Na- turelle. Paris (MNHN): Museum fur Na- turkunde der Humboldt-Universitat zu Berlin (MNHU); Musee Oceanographique Monaco (MOMV); Naturhistoriska Riks- museet. Stockholm (NHRS): Royal Scottish Museum, Edinburgh (RSME); National Museum of Natural History, Washington (USNM); Zoologisk Museum. Copenhagen (UZMK); Zoological Institute. Academy of Science. Leningrad (ZIAS): Zoological In- stitute. Tohoku University. Sendai (ZITU); Zoologisch Museum, Universiteit van Am- sterdam (ZMUA); Zoological Museum, University Bergen (ZMUB): Zoologisches Museum. Universitat Hamburg (ZMUH). Morphological Characters 1. Introvert hooks and spines. —As in most genera, the introvert bears (in all but three species) some array of specialized structures usually referred to as hooks. Voss-Foucart et al. (1977) have shown these to lack chitin but consist of a horny protein. Many hooks are arranged in regular rings around the dis- tal portion of the introvert and may have either one or two points (uni- or bidentate). On some species scattered hooks also are found proximally and in two species only scattered hooks are found. Additionally, epidermal structures of varying sizes and shapes called spines are arranged in a ran- dom manner on the proximal portion of the introvert. Examination of the literature re- veals that the term "spine" meant different things to different authors, sometimes being used as a synonym for what others would term "unidentate hook." Since hooks and spines come in a wide variety of arrange- ments, sizes, and shapes, and often grade into one another, it is easy to understand the genesis of this problem that is unique to Aspidosiphon. We will attempt some clar- ification and definitions. In the phylum Sipuncula. the term "hook" has been applied to structures having a wide variety of shapes, sizes, and arrangement. It is clear that a Themiste hook is different from a Xephasoma hook and that both dif- fer from Phascolosoma hooks. Structures that have been called "spines" are similar to "hooks" of Themiste and of some Phas- colion species. We now propose calling all of these introvert structures hooks. The term "spine" will be restricted to conical pointed anal shield units. The hook's apex points posteriorly (away from the mouth) with the convex curvature being anterior. The different types of hooks are defined as follows: Type A: Compressed hooks.— Usu- ally arranged in rings, occasionally scattered, laterally compressed, and having, in a side view, a distinct pos- terior curve. These may be unidentate or bidentate (Fig. 1A. B). When a sec- ondary tooth is present it may be vari- able in size, sometimes reduced to a 828 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Table 1.— Original and proposed names of the Aspidosiphon species. Subgenus Aspidosiphon s.s. Aspidosiphon albus Murina, 1967 Aspidosiphon brocki Augener, 1903 Aspidosiphon carolinus Sato, 1935 Aspidosiphon cylindricus Horst, 1899 Aspidosiphon elegans (Chamisso & Eysenhardt, 1821) Aspidosiphon exhaustus Sluiter, 1912 Aspidosiphon exiguus Edmonds, 1974 Aspidosiphon exilis Sluiter, 1886 Aspidosiphon gerouldi ten Broeke, 1925 Aspidosiphon gosnoldi Cutler, 1981 Aspidosiphon gracilis (Baird, 1868) Aspidosiphon hartmeyeri Fischer, 1919 Aspidosiphon hispitrofus LiGreci, 1980 Aspidosiphon homomyarius Johnson, 1964 Aspidosiphon imbellis Sluiter, 1902 Aspidosiphon inquilinis Sluiter, 1 902 Aspidosiphon jukesii Baird, 1873 Aspidosiphon kovaleskii Murina, 1964 Aspidosiphon longirhyncus Cutler & Cutler, 1980 Aspidosiphon macer (Sluiter, 1891) Aspidosiphon mexicanus (Murina, 1967) Aspidosiphon misakiensis Ikeda, 1904 Aspidosiphon muelleri Diesing, 1851 Aspidosiphon ravus Sluiter, 1886 Aspidosiphon spinalis Ikeda, 1 904 Aspidosiphon spinosus Sluiter, 1902 Aspidosiphon spiralis Sluiter, 1 902 Aspidosiphon thomassini Cutler & Cutler, 1979 Aspidosiphon tortus Selenka, de Man & Biilow, 1883 Aspidosiphon venabulum Selenka, de Man & Biilow, 1883 Aspidosiphon zinni Cutler, 1969 Subgenus Par aspidosiphon Aspidosiphon ambonensis Augener, 1903 Aspidosiphon angulatus Ikeda, 1904 Aspidosiphon brasiliensis Cordero & Mello-Leitao, 1952 Aspidosiphon coyi Quatrefages, 1865 Aspidosiphon cumingii Baird, 1868 Aspidosiphon exostomus Johnson, 1964 Aspidosiphon fischeri ten Broeke, 1925 Aspidosiphon formosanus Sato, 1939 Aspidosiphon gigas Sluiter, 1884 Aspidosiphon grandis Sato, 1939 Aspidosiphon havelockensis Haldar, 1978 Aspidosiphon insularis Lanchester, 1905 Aspidosiphon johnstoni Edmonds, 1980 Aspidosiphon klunzingeri Selenka, de Man & Biilow, 1883 Aspidosiphon laevis Quatrefages, 1865 Aspidosiphon levis Sluiter, 1886 Aspidosiphon major Vaillant, 1871 Aspidosiphon makoensis Sato, 1939 Aspidosiphon ochrus Cutler & Cutler, 1979 Aspidosiphon pachydermatus Wesenberg-Lund, 1937 Aspidosiphon parvulus Gerould, 1913 no change* A. elegans A. elegans species inquirenda no change A. muelleri no change A. elegans A. misakiensis no change no change A. misakiensis A. muelleri A. elegans A. muelleri A. muelleri A. muelleri A. muelleri A. mexicanus* species inquirenda no change* no change no change A. elegans A. elegans A. elegans no change no change* A. muelleri no change* no change* A. tenuis A. laevis A. laevis no change A. laevis A. steenstrupii no change A. tenuis A. laevis A. laevis A. tenuis Phascolosoma perlucens A. laevis A. laevis no change A. tenuis A. laevis A. steenstrupii A. steenstrupii A. laevis no change VOLUME 102, NUMBER 4 829 Table 1.— Continued. Aspidosiphon planoscutatus Murina, 1968 Aspidosiphon quatrefagesi Saiz Salinas, 1984 Aspidosiphon pygmaeus Fischer, 1921 Aspidosiphon schnehageni Fischer, 1913 Aspidosiphon semperi ten Broeke, 1925 Aspidosiphon speciosus Gerould, 1913 Aspidosiphon speculator Selenka, 1885 Aspidosiphon spinososcutatus Fischer, 1922 Aspidosiphon steenstrupii Diesing, 1859 Aspidosiphon tenuis Sluiter, 1886 Aspidosiphon trinidensis Cordero & Mello-Leitao, 1952 Aspidosiphon truncatus (Keferstein, 1867) Golfingia mokyevskii Murina, 1964 no change A. laevis A. muelleri A. gracilis schnehageni A. steenstrupii A. laevis A. steenstrupii A. parvulus no change no change A. steenstrupii A. coyi Antillesoma antillarum * Now in new subgenus A. (Akrikos). small knob. A transition zone in some species exists at the proximal end of the rings of hooks where one may find a gradual widening of the anterior base of the unidentate hooks. Sometimes these scattered hooks are rounded at the anterior-lateral corners but still compressed posteriorly looking like a ship's stout mast and sail. Type B: Pyramidal hooks.— Have triangular bases, the anterior side of which is shorter than the lateral sides, are usually less curved than Types A or C, are variably pigmented (dark to light), and translucent (Fig. IE, F). The borderline between Types A and B is not clear in all species. Type C: Conical hooks.— Have a nearly circular cross section (cone shape), a gentle posterior curve, and are usually opaque and dark colored (Fig. 1C, D). This type is found on the dorsal side of A. elegans' introverts. When introvert skin is removed and placed on a slide in a drop of glycerin for closer examination, the orientation of these hooks can add to the confusion. If viewed from the anterior or posterior (instead of laterally), scattered unidentate compressed hooks look very much like pyramidal hooks (Fig. 1G). However, a compressed hook has a narrower base. Further distortion can be caused if the hooks are not lying flat on the slide. Scanning electron micrographs can help reveal the natural configurations and the three dimensionality of these structures. The shape of the clear area (less dense to transmitted light) in the hook has limited taxonomic value. In most species there is an ill-defined triangular area, but in A. steenstrupii and A. elegans there is a thin, posteriorly directed, tongue-like extension (Fig. 2). The height of the hook has sometimes been used as a diagnostic character, but this has limitations. The unidentate A. laevis complex (as defined below) shows a clear correlation between trunk size and hook size (5-10 mm worms have 20-40 iim hooks, 20-30 mm worms have 40-60 fim hooks, and 40-70 mm worms have 60-80 fj,m hooks). The same pattern is shown by an analysis of 32 specimens of A. steenstrupii ranging in trunk length from 7-50 mm. Hook size varied from 30-90 ^m with larger worms having larger hooks (a linear regres- sion of these data gave a positive slope of 0.89). Therefore, hook size should not be considered in isolation from trunk size. De- spite this a clear pattern does appear with certain species (e.g., the members of the new subgenus proposed here) always having small hooks (under 30 /zm) and some species 830 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON ->">- ^Mppi! s to ^^j Fig. 1 . Introvert hooks: A, Compressed, bidentate (Type A) from A. muelleri, posterior view; B, Compressed unidentate (Type A) from A. misakiensis; C-D, Corneal (Type C) from A. elegans; E, Pyramidal (Type B) from A. parvulus; F, Pyramidal (Type B) from A. steenstrupii viewed from above; G, Pyramidal (Type B) from A. steenstrupii, different angles on light microscope. Scale line =10 /mi, for G = 20 /mi. with only larger hooks. The central problem is that many species have both large and small hooks. In general, some species lack unidentate hooks, some lack bidentate hooks, some have both Type A and B hooks while others have only Type A hooks. Certain species will have both kinds of compressed and Type B hooks. In summary, if one examines the most distal rings of hooks and differentiates between unidentate Type A and Type B hooks, hook morphology can be useful to the systematist in almost all cases and hook size can help in some cases. 2. Anal shield (degree of development, na- ture of units, grooves). —At the anterior end of the trunk, horny protein (not chitin) forms an array of cuticular units varying in degree of development (Voss-Foucart et al. 1977). At one extreme is A. mexicanus or A. tho- massini with a collection of small scattered units sometimes looking more like an area VOLUME 102, NUMBER 4 831 Fig. 2. Internal hook structure: A, A. elegans, B, A. steenstrupii, C, A. tenuis. Note absence of tongue-like extension in C. Scale lines = 20 ^m. of rough skin. At the other end of the con- tinuum is A. laevis or most A. muelleri where the units are compacted to form a thick, dark, solid mass. These two species are among those that have shields with well de- veloped longitudinal and/or transverse grooves. When a shield has aggregations of units separated by grooves we refer to these aggregations as plates. The shield nearest the mid-dorsal anus is the dorsal part while that nearest the introvert is the ventral part. The nature of the shield units may undergo slight changes with age (see A.jukesii in Cut- ler & Cutler 1979a:970) and may be mod- ified by the size/shape of the shell (gastro- pod vs. scaphopod) or other space occupied by the worm. When the introvert is retract- ed this shield functions as an operculum. While some within-deme variation does ex- ist, the morphology of the anal shield is con- sistent enough and distinct enough to be useful to the systematist. 3. Caudal shield (degree of develop- ment).— At the posterior end of the trunk there is a epidermal structure (horny pro- tein) present in most (but not all) species. This shield assumes various forms in living worms and, therefore, when preserved, can also vary from rather flat to pointed to pa- goda shaped. They have a variable number of radially arranged grooves or furrows, but this attribute is not species specific. Even within a deme the degree of development (thickness) varies (Fig. 3). In a species with extensive historical data indicating a "nor- mal" shield, one may find individuals with very weakly developed shields. In such an- imals the shield may be reduced to some papillae and only a vague suggestion of a chitinoid layer. It is probable that in many species the genetic potential is there, but its expression is variable and responds to en- vironmental stimuli or age. Some species {A. laevis) always have a shield and others (e.g., A. mexicanus or A. zinni) never do. When a shield is present, there is little to distinguish one species from another using this character. Therefore, aside from pres- ence/absence, the caudal shield has limited value to the systematist. 4. Introvert retractor muscles (origin, de- gree of fusion).— The single pair of long muscles used to retract the introvert insert at the anterior end and originate from the ventral trunk wall in the posterior third of the worm. The placement of these origins (attachment) is usually included in species descriptions, often with a significant lack of precision (near the posterior end, in the last quarter), but sometimes it is used as a di- agnostic (or key) character. While in other genera we have found that retractor origins appear to shift anteriad as the animal grows (zone of trunk growth being posterior to the origins), it is difficult to generalize about this in Aspidosiphon. If one calculates the dis- 832 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. 3. Caudal shields of A. elegans. Three 17 mm worms from one deme in Ponape to show variation in degree of development. Scale line = 1 mm. tance as a percentage of trunk length, there do seem to be some species where this value remains reasonably constant over a wide range of sizes (e.g., A. muelleri, 95-100%). However, other species exhibit more vari- ation (see Table 2 for A. elegans, 65-85% and A. steenstrupii, 70-90%). This is not clearly correlated with size. Therefore, while this attribute can be useful for differentiat- ing two subsets, it does not have value at the species level. The second aspect of this complex that is sometimes noted is the degree of fusion be- tween these two muscles. In the older lit- erature there is considerable confusion on this matter, such that some descriptions al- lege there to be only one or as many as four, or one muscle with one, two or four "roots." From what we know about the ontogeny of this functional complex in sipunculans, it is probable that they all begin life with two pairs of retractors (Rice 1976). The dorsal pair is lost and the ventral pair fuses to vary- ing degrees. However, quantifying this is ex- tremely difficult because extension of the introvert will stretch out the muscle and the point of fusion will move away from the posterior end of the trunk. While one gets the impression that there are different sized muscles and different degrees of fusion, no pattern can be discerned. We are unable to suggest how this information can be uni- versally applied even if species specificity was evident. 5. Spindle muscle attachment. —This thin muscle extends from the body wall at the posterior end of the trunk, through the gut coils, and along the rectum to the body wall just anterior to the anus. While there have been a few reports of this muscle merging with the body wall posterior to the anus, the only cases of this kind we have been able to confirm are in A. laevis and A. coyi in which the muscle sends a large branch to the mid-dorsal body wall posterior to the anus. It is not always easy to trace the course VOLUME 102, NUMBER 4 833 Table 2.— Selected morphological attributes of A. elegans and A. steenstrupii. Aspidosiphon elegans Aspidosiphon steenstrupii Trunk length in mm Retractor origin; % trunk Nephridia length; % trunk Nephridia attachment; % nephridium Trunk length in mm Retractor origin; % trunk Nephridia length; % trunk Nephridia attachment; % nephridium 6 83 75 89 7 71 42 67 7 86 70 95 7 79 57 75 7 — 57 95 10 75 50 50 8 63 63 99 11 82 45 75 9 67 50 99 11 73 99 67 10 70 60 67 12 75 50 67 10 70 70 93 12 75 99 99 11 82 64 99 14 79 54 75 14 86 57 88 15 77 50 — 15 60 67 50 15 80 60 56 15 80 40 83 15 73 67 50 16 63 53 85 15 73 60 50 16 81 63 99 17 82 41 71 16 88 88 95 19 79 58 50 17 76 35 99 19 68 47 67 17 65 65 82 19 84 47 90 19 79 58 86 20 85 75 40 19 74 47 99 21 76 57 — 21 90 52 64 22 77 68 50 22 86 86 95 24 71 42 75 23 78 57 99 25 80 48 58 25 84 60 67 26 81 69 50 32 78 56 99 27 70 81 99 28 75 65 99 33 79 67 50 33 85 52 53 34 85 50 50 50 76 70 33 50 80 54 75 of this muscle, especially in smaller worms or where the spindle muscle seems to merge with the wing muscle. Except for the above species, the muscle originates anterior to the anus and, therefore, in this genus, it is of limited usefulness as a diagnostic character. 6. Fixing muscle number.— In most si- punculan taxa fine thread-like muscles an- chor some part of the anterior intestine to the body wall. The maximum number re- corded in this genus is one and its presence (or absence) has been considered system- atically important by some biologists. Our review of the literature reveals a lack of con- sistency on this point. Our own studies strongly suggest that either: A. Genetic poly- morphism exists within populations or B. This fragile structure is placed where it can be easily damaged during dissection. While it may be true that some species totally lack this muscle and others usually have one, the possibility of any one worm deviating from the norm is too great to give weight to this character. 7. Nephridia (length, attachment, level of nephridiopore).— Both the nephrostome and the nephridiopores are located at the ante- rior end of the nephridia, a pair of tubular sac-like ventro-lateral organs. These open to the outside at the anterior end of the trunk. Three attributes have been recorded and are sometimes treated as diagnostic. 834 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON First is the position of the nephridiopores relative to the anus. In the literature most species of A. (Aspidosiphon) and a few A. (Paraspidosiphori) are reported as having nephridia at the level of the anus but slightly posterior to it in the remaining species. In this latter group our own data show this distance to be 3-8% of the trunk length. However, even in these populations a few animals have the nephridiopores and anus at the same level. While there do seem to be other species where this distance is al- most always zero, careful examination of more than ten worms will probably reveal one or two where these openings are not at the same level. It may be possible to say: "80-90% of species X have these openings at the same level while in species Y only 5- 10% do." But as these kinds of data are not always available (small sample sizes), the information has limited value to the sys- tematist. A second attribute is the length of these organs. The literature includes statements like "very long, long, reach to the base of the retractors, two-thirds/half/one fourth as long as the trunk," etc. Our observations show a range of lengths within a population, not correlated with trunk size (see Table 2), but there are some differences among species. Six species have nephridia half the trunk length or less. Many exhibit a broad range (e.g., 25-100%, 45-85%, 50-100%) and a few have only long nephridia (over 85% of the trunk length). With a few ex- ceptions, nephridia length can only be used in a limited manner. A third attribute often mentioned is the attachment of the nephridia to the body wall by a membranous sheet of connective tis- sue. Occasionally this takes the form of fil- amentous strands of tissue scattered along the nephridia that bind them less tightly. While nephridial attachment often appears in keys to Aspidosiphon, its constancy is overrated. While the original description may state that the nephridia are attached for a particular length (one-half, two-thirds, 100%, etc.), subsequent authors either ig- nore it or do not verify it critically. The connective tissue is easily torn, and within a single worm the attachment can differ for each organ. In most species a wide range of values is observed, commonly in the 50- 100% range (see Table 2 for A. steenstrupii, 50-80% and A. elegans, 80-100%) while A. muelleri ranges from 0-100% attached in different reports. Therefore, this attribute has restricted systematic value. 8. Rectal caecum (presence, complexi- ty).— In many sipunculans there is a small caecum on the rectum near the intestinal coil. Certain species of Aspidosiphon are re- ported to lack a caecum while others are said to have one. Two problems exist: If one looks at all of the published reports of any frequently recorded species, one finds inconsistencies, e.g., in A. elegans several authors say it lacks a caecum, others make no mention (this could mean that it is absent or possibly that they did not look for it), and a few report seeing the caecum. Alter- natively, A. muelleri has been said to have a caecum by many authors but a few assert that it lacks one. In the A. muelleri we have recently examined, 25% have a caecum. In A. albus the original describer (Murina 1 967c) asserts that the caecum is absent. Subse- quent workers looking at other members of the population (Cutler 1973, Migotto & Di- tadi 1988) saw one. The second problem is related to this and it is illustrated by our finding a caecum in two often A. elegans we examined. Here we have one observer looking at one deme and finding dimorphism. In 30 worms that Mig- otto & Ditadi (1988) dissected they found a caecum in 1 8 (two of these being "large, villous"). In other words, 40% of this pop- ulation of a species described as having a caecum was found to lack one. As this is true for one population, it may be true for all. We conclude that it is an error to con- sider the presence or absence of a rectal cae- cum as species specific. In this genus there is one special case, i.e., VOLUME 102, NUMBER 4 835 a rectal structure described as: ". . . the last part of the rectum densely bordered with long villi-like structures" (Selenka et al. 1883 in A. cumingii) or "rectum with caecum plus many blind tubes attached to both sides" (Sato 1939 for A. grandis) or "rectum with a larger caecum, many lobed on both sides" (Selenka et al. 1883 in A. klunzingeri) or "rectal caecum with lappets" (Edmonds 1956 in A. klunzingeri) or "a large cluster of blind sacs as ramified intestinal appendages" (Wesenberg-Lund 1937 in A. pachyderma- tus). Only one specimen of A. pachydermatus and two of A. grandis (100, 70 and 100 mm trunks) exist. Of the 12 records of A. klun- zingeri that include morphological com- ments, only the original material (3 worms) has this type of rectal elaboration. For sub- sequent authors, its absence seemed incon- sequential. Similarly, in the nine reports of A. cumingii with morphological comments, only Selenka et al. (1883) record this struc- ture in five worms (the words are used by other authors but only when referring to Selenka et al.). So, while these species sup- posedly have an elaborate rectum, several authors have used this name for animals lacking said structure. In museum collec- tions, fewer than 10 worms exist (Baird's two have been lost) with this condition. The four worms in our 1985 Hawaii material that we are calling A. laevis have trunk lengths of 7, 10, 11 and 14 mm. The largest and the smallest bear a single structure with 6-8 short branches or lobes coming off each side. The 10 and 1 1 mm worms show small lobes, but the precise structure is less clear due to the fragility of the rectum. In our 1988 Venezuela collections we have five A. laevis (10-30 mm trunks), but none of these has a caecum of any kind. Migotto & Ditadi (1988) report large villous, simple and no caeca in a single population. The question posed above reappears here: If an author did not mention this structure, does that mean it was not present? In those cases when an author specifically says that the complex caecum was not present, what significance does that have? Historically the position that this is insignificant and vari- able within a population has implicitly pre- vailed. Our recent data confirm that and suggest that A. laevis has the genetic poten- tial (perhaps polygenic), but this potential is not always expressed, and when it is ex- pressed it may be overlooked by an observ- er. 9. Intestinal coils (tightness).— In some other genera the number of gut coils has been used as a systematic character, but we have elsewhere shown this to be size de- pendent and not useful. In Aspidosiphon the interest is restricted to the nature of the coil- ing. In most species a regular, compact dou- ble helix is present, but A. misakiensis ex- hibits a loose, less regular helix (Fig. 4). This has also been reported by a few authors for A. elegans. The helix is maintained by the fine strands of muscle linking the coils to the axial spindle muscle. In those animals having a looser gut coil the linkage is not continuous and the strands may be longer. 10. Longitudinal muscle (bundles, anas- tomosing, fracturing).— The body wall of sipunculans has an internal layer of lon- gitudinal muscle. In one subgenus {A. Aspidosiphon) this is an undivided sheet, almost. In the other subgenus {A. Paraspi- dosiphon) this layer is divided into separate bundles. However, it is not always a clear- cut dichotomy. Intermediate conditions of two general types exist. First, ten putative species of A. {Aspido- siphon) have been described as having frac- turing of the muscle layer in the anterior dorsal trunk. Commonly this is restricted to the area under the anal shield, but in some species the fractures continue out be- yond these borders for a small (10-20% of the trunk length) but variable distance. The second variation occurs in species of A. {Par aspidosiphon) where the longitudinal muscle bands are not distinct. The degree of variation is much greater than in other genera with these bundles. While several have distinct rarely anastomosing bundles, 836 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. 4. Intestinal coil of A. misakiensis showing irregular loose arrangement (A— anal shield, M— retractor muscle, N— nephridium, R— rectum). Scale line = 2 mm. many exhibit a modest degree of anasto- mosing and others show frequent cross link- ages. In these latter the layer appears like a continuous sheet that has split or fractured and not like distinct bundles (e.g., A. fi- scheri). The number of muscle bands varies con- siderably within a population and within an individual (25-35 anteriorly and 15-25 posteriorly). It is difficult in a few cases, especially in small worms, to know whether one is looking at an A. {Par aspidosiphon) with much anastomosing of bundles or at an A. (Aspidosiphon) with some fracturing of a layer. While this character state may be used for separating subgenera, it is neither species specific nor discriminating at that level. Since the subdivision of a continuous layer appears to be an homoplastic, apomorphic condition (Cutler & Gibbs 1985) it is pos- sible that it has arisen more than once with- in this genus. Therefore, using it as the single attribute to separate subgenera may mask actual phylogenetic relationships. 11. Angle of introvert to trunk.— Tn most Aspidosiphon the extended introvert pro- trudes at an angle of 75-90° to the main axis of the trunk at the ventral edge of the anal shield. However, in at least three species, all with very weakly developed anal shields, this angle is reduced to 45-60°. While this is not broadly useful, it can help in these special cases. VOLUME 102, NUMBER 4 837 Summary.— Two characters that have broad taxonomic usefulness are the hook and anal shield morphology. Four charac- ters useful for separating the genus into dif- ferent subsets are: (1) longitudinal muscle layer continuous or divided, (2) retractors originating in the most posterior 5% of the trunk or in the 70-80% range, (3) caudal shield developed or not, and (4) nephridia length (less than 50% of trunk length, more than 75%, or a broad range). An introvert/ trunk angle of less than 75° separates three species, while a bifurcated anterior spindle muscle and a loosely wound gut coil each characterizes one species. The presence/ab- sence of fixing muscles or caecum, the place- ment of the nephridiopores, and the attach- ment of the nephridia to the body wall are too variable to have any systematic value in this genus. Systematic Section Aspidosiphon Diesing, 1851 Diagnosis. —Introvert usually longer than trunk. Recurved hooks in numerous rings (absent in three, only scattered in two species). Trunk with anal shield composed of hardened units (occasionally inconspic- uously developed). Introvert protrudes from ventral margin of shield. Body wall either with continuous longitudinal muscle layer or with longitudinal muscle layer gathered into anastomosing, sometimes ill-defined, bundles. Oral disk with tentacles enclosing dorsal nuchal organ but not mouth. Con- tractile vessel without villi. Two introvert retractor muscles sometimes almost com- pletely fused. Spindle muscle attached pos- teriorly. Two nephridia. One species may exceed 100 mm but most less than 40 mm long. We herein create a new subgenus for a set of five species as defined below. The major character state we focus on is the absence of compressed hooks in rings, a plesio- morphic character state for this entire class (Cutler & Gibbs 1 985). One could argue that these species represent primitive transition forms from a very early ancestral stock, but we propose instead that this trait has been secondarily lost through subsequent evo- lution, i.e., a type of reversal. In support of this, note the atypical ecology of these taxa (e.g., interstitial or abyssal, none boring in coral or rock) suggesting an ecological spe- cialization. The following four taxa are not consid- ered valid members of this genus and are discussed first. After the key, the remaining species are presented alphabetically within subgenera. Aspidosiphon cylindricus Horst, 1899 Aspidosiphon cylindricus Horst, 1899:195— 198, text-jgs. 3-4. -Stephen & Ed- monds, 1972:222-223. — Not Sluiter, 1902:18-19. Material examined.— ZMU A, Sluiter's material (V. Si. 26.8). Discussion. —This species was based on a single worm that disappeared from the Lei- den museum prior to 1930 (van der Lund, pers. comm.). Enough questionable but now unverifiable features exist (especially the hook morphology) that we place this name on the list of species inquirenda pending fu- ture clarification. Sluiter's (1902) material was reexamined and is herein referred to A. elegans. Aspidosiphon insularis (Lanchester, 1905) Aspidosiphon insularis Lanchester, 1905b: 40, pi. 2, fig. 4. -Gibbs & Cutler, 1987: 56. Par aspidosiphon insularis. —Stephen & Ed- monds, 1972:247. Material examined. — BMNH, type (1924.3.1.80). Discussion. — This worm is in poor con- dition, but, as noted by Gibbs & Cutler (1987), it is clearly a Phascolosoma and based on hooks and papillae is a junior syn- onym of P. perlucens. 838 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Aspidosiphon macer (Sluiter, 1891) Phascolosoma macer Sluiter, 1891:114-115, pi. 2, figs. 13-14; 1902:34. Golfingia macra.— Stephen & Edmonds, 1972:149. -Cutler & Murina, 1977:183. Aspidosiphon macer. —Cutler & Cutler, 1986:568. Material examined.— ZMU A, type and only specimen (V. Si. 65). Discussion. —This putative taxon is based on a single specimen that has been thor- oughly dissected over the years. The generic status is not altogether firm since the anal shield is very poorly developed and the in- trovert does not appear to be ventrally dis- placed. The gut is missing and while Sluiter asserted that the spindle muscle is not at- tached posteriorly, there is a muscle coming from the center of the caudal shield that we interpret as the broken posterior portion of this muscle. Sluiter said there were no hooks but there are about ten distinct rings of sharply pointed, unidentate hooks. The lon- gitudinal muscle layer is undivided. While Sluiter asserted that the tentacles surround the mouth, the introvert is not extended so one cannot really tell how the tentacles are arranged. They appear to us to be clustered towards one side as in this genus. Therefore, until additional material is obtained to clar- ify this uncertainty, we place this name on the list of species inquirenda. Golfingia mokyevskii Murina, 1964 Golfingia mokyevskii Murina, 1 9 64a: 256- 259, figs. 4-5. ^Aspidosiphon mokyevskii.— Gibbs et al., 1983:302. Material examined.— 71 AS, type mate- rial. Discussion.— Gibbs et al. (1983) suggest- ed that Murina' s species might be an As- pidosiphon; however, it is now clear that this was an error. The 50 long tentacles, absence of hooks, anastomosing longitudinal muscle bands, large dark papillae at the base of the introvert, and especially the presence of contractile vessel villi all support placing this name in the synonomy of Antillesoma antillarum. Key to Aspidosiphon species 1 . Hooks not present, or if present, not in rings A. (Akrikos) 2 - Hooks arranged in rings on distal portion of introvert 6 2. Introvert hooks absent 3 - Scattered introvert hooks present 5 3. Anal shield of tightly packed, uni- form sized, pale, flat units with dis- tinct angular margin A. albus Murina - Anal shield of dispersed, often dark units, sometimes very poorly de- veloped, with indistinct margin . . 4 4. Anal shield units distinct, dark; those around margin usually pointed cones . . . A. venabulum Selenka, de Man & Biilow - Anal shield units indistinct, widely spaced, flat, sometimes arranged in indistinct rows A. thomassini Cutler & Cutler 5. Anal shield ill defined and diffuse, trunk usually more than 5 mm, shallow warm water A. mexicanus Murina - Anal shield well defined and com- pact, trunk usually less than 5 mm, deep cold water A. zinni Cutler 6. Longitudinal muscles in continu- ous layer (except under anal shield) A. {Aspidosiphon) 7 - Longitudinal muscle layer divided into separate (or anastomosing) bundles . . A. {Par aspidosiphon) . . 13 7. Anal shield with extensive array of furrows present, not just around margin 8 Anal shield with randomly distrib- uted hardened units, lacking ex- tensive grooves/furrows 9 VOLUME 102, NUMBER 4 839 8. Individual units form into longi- tudinal ridges over dorsal half of anal shield A. muelleri Diesing - Individual units arranged in offset squares or rectangles A. spiralis Sluiter 9. All hooks unidentate, ill-defined anal shield A. gracilis (Baird) - Distal hooks bidentate, anal shield distinct 10 10. All compressed hooks bidentate followed by dark conical hooks . . . . . A. elegans Chamisso & Eysenhardt - Distal bidentate compressed hooks followed by proximal unidentate ones 11 1 1 . Interstitial, introvert 2-5 times the trunk length, nephridia 25-33% trunk A. exiguus Edmonds - Occupies coral or shells often sub- tidal, introvert 1-3 times trunk, nephridia more than 50% of trunk length 12 12. Normal gut helix, lives in gastro- pod shells, anal shield units square, arranged in rows, and each made up of smaller granular subunits . A. gosnoldi Cutler - Gut coils loose or absent, does not occupy gastropod shells, anal shield units more solid and randomly ar- ranged A. misakiensis Ikeda 13. Anal shield ungrooved or, if pres- ent, only as short marginal ones (Fig. 5A) 14 - Anal shield with extensive grooves or furrows present (Fig. 5B) .... 18 14. Distal hooks bidentate 15 - All hooks unidentate A. planoscutatus Murina 15. Compressed hooks bidentate fol- lowed by dark pyramidal hooks . A. steenstrupii Diesing - Compressed hooks of both types, pyramidal hooks pale, if present 1 6 16. No pyramidal hooks, longitudinal muscle bands distinct, compressed hooks over 30 mm tall, retractor origins 75-88% A. tenuis Sluiter - Pale pyramidal hooks present, lon- gitudinal muscle bands indistinct, hooks less than 30 ^m tall, retrac- tor origins 95-100% 17 17. Anal shield marginally becomes diffuse forming cones or spikes, nephridia more than l h trunk . . . A. parvulus Gerould - Anal shield with distinct margins, no cones or spikes, nephridia less than l h trunk . . . A. fischeri ten Brocke 18. All hooks unidentate, retractor origins not at posterior end (60- 80%) A. laevis Quatrefages - Distal hooks usually have very small secondary tooth, retractor origins at posterior end (95-100%) A. coyi Quatrefages Aspidosiphon (Akrikos), new subgenus Diagnosis. —Aspidosiphon with com- pressed hooks not in rings, i.e., either scat- tered and small (less than 30 jttm), or absent; caudal shield absent or very diffuse; longi- tudinal muscle layer continuous. Not known to bore in coral or rock. The name is from the Greek meaning "without rings." The spelling is a literal transliteration according to the classical method. Type species: Aspidosiphon albus Murina, 1967. Aspidosiphon albus Murina, 1967 Aspidosiphon albus Murina, 1967a: 1330- 1331, fig. 2 (1H3).- Stephen & Ed- monds, 1972:219-221. -Cutler, 1973: 174-1 75. -Cutler & Cutler, 1980a:4.- Migotto & Ditadi, 1988:247-248. -Not Cutler etal., 1984:307. Aspidosiphon hartmeyeri. — Wesenberg- Lund, 1957a:7-8; 1959a: 197; 1959b:212. Material examined. — ZIAS, type; USNM, cataloged as A. cumingii 066214-222 off Louisiana; material from Brazil (Cutler & 840 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. 5. Anal shields: A, Ungrooved type of A. elegans; B-C, Grooved type as in A. muelleri; D, Ventral cone- shaped units near the ventral margin present in some A. muelleri. Scale lines on A, B, C = 0.5 mm, D = 0.1 mm. Cutler 1980a), U.S. east coast (Cutler 1973), and new Ivory Coast material from 5°N, 4°W, at 20 m; A. hartmeyeri, UZMK, Wes- enberg-Lund's Niger worms. Discussion. — This hookless species also seems to lack tentacles. The recorded body length ranges from 2 to 45 mm, the introvert is 3-5 times the trunk length, and the anal shield is fine grained with small furrows around the margin but without grooves. Often there is a median stripe made up of darker units. The nephridia are 50-75% of the trunk. The Cutler et al. (1984) record was based on one incomplete worm and was not a pos- itive identification. This should not be in- cluded within this species unless additional material is found in Japanese waters. Distribution.— Cape Hatteras, northern Gulf of Mexico (unpublished), Cuba, Brazil continental shelf (10-123 m), and the east Atlantic from the Gulf of Guinea. It is the most common member of this genus on the Brazilian shelf. Aspidosiphon mexicanus (Murina, 1967) Golfingia mexicana Murina, 1967c: 1333- 1334, fig. 3. VOLUME 102, NUMBER 4 841 Aspidosiphon mexicana Cutler et al., 1983: 673. Aspidosiphon longirhyncus Cutler & Cutler, 1980a:4-6, figs. 4-5. Material examined. — ZIAS, type; A. lon- girhyncus, AMNH, type (4022); other ma- terial of both species from the type locali- ties. Also, 12 unpublished specimens from the Azores (33°N, 16°W, 320 m) and seven from three stations off Florida and North Carolina (25-34°N, 85-190 m). Discussion. — This species does not have the typical aspidosiphonid appearance in that the anal shield is very weakly devel- oped with scattered, ill-defined platelets. Also, the introvert is on an atypical angle (45-60°) with the main trunk axis. The cau- dal shield is almost nonexistent. The intro- vert is 4-5 times the trunk length, bearing scattered, small (less than 30 jum), uniden- tate compressed hooks. The nephridia are 35-75% of the trunk. The decision to combine these two names was discussed in Cutler et al. (1983) as part of a review of the taxon containing Murina's species. Distribution.— Southern Brazil, Cuba, and southeastern U.S. at shelf depths (80-200 m), and the Azores at 320 m. Aspidosiphon thomassini Cutler & Cutler, 1979 Aspidosiphon thomassini Cutler & Cutler, 1979a:971-973, figs. 3-14. Material examined. — MNHN, type ma- terial (AH 406-408). Discussion. — This is another small (1.5- 7 mm) hookless species whose tentacles (if present) have yet to be observed. The re- tractor muscles are fused for most of their length and the nephridia are around 50% of the trunk. The introvert is 2-4 times the trunk length. These have no caudal shield and the anal shield is very poorly devel- oped. This, together with the smaller angle between the trunk and introvert axis (40- 45°) make it possible to mistakenly identify this as a Nephasoma species. Distribution. — Intertidal coral sands in Madagascar and French Polynesia. Aspidosiphon venabulum Selenka, de Man & Biilow, 1883 Aspidosiphon venabulum Selenka et al., 1883:123, pi. 14, figs. 202-204. -Fischer, 1895:18; 1914a:68-69.-Wesenberg- Lund, 1957c:5-7; 1959a: 196-1 97; 1959c: 212. — Longhurst, 1958:85. — Stephen, 1960a:5 19. -Cutler, 1977a: 148. Aspidosiphon venabulus. — Stephen & Ed- monds, 1972:237. -Cutler & Cutler, 1979a:971. Material examined. — MNHU, type (644); UZMK, Wesenberg-Lund's Atlantide ma- terial; worms from Madagascar (Cutler & Cutler 1979a) and West Africa (Cutler 1977a). Discussion. —This species lacks hooks and the anal shield is ungrooved, being made up of dark, pointed, conical units more widely scattered than in many species. It resembles the anterior end of certain Phascolion species that have large anterior papillae. The re- tractor muscles extend to the posterior end and under the anal shield, the longitudinal muscle layer splits into a few bundles. The nephridia are 60-95% of the trunk, the latter being reported from 5-30 mm. Distribution. — Subtidal depths (10-55 m with one intertidal report and one at 960 m). Most records are off West Africa with one report from southern Madagascar. Aspidosiphon zinni Cutler, 1969 Aspidosiphon zinni Cutler, 1969:209-211, fig. 1. -Cutler, 1973:176-178.-Cutler& Cutler, 1979a:968; 1980b:457; 1987a:73. Material examined. — USNM, type ma- terial (38242, 38243); Atlantic Ocean ma- terial (Cutler & Cutler 1987a). Distribution. —This small (most less than 5 mm), deep-water worm commonly lives in foraminiferan tubes. The anal shield is made up of very fine grained, pale units and the caudal shield is nonexistent. The small 842 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON (15-20 /mi), scattered, unidentate, com- pressed hooks and introvert not longer than the trunk are unusual in this genus. The nephridia are less than 25% of the trunk. Distribution.— Common in north Atlan- tic Ocean (1100-4400 m), two stations around 9°S off the Congo River (1535 and 2700 m), and one in the Mozambique Channel (25°S) at 132 m. Subgenus Aspidosiphon (Aspidosiphon) Diagnosis.— Introvert with compressed hooks in rings, longitudinal muscle layer continuous except near anal shield. Most do not bore in coral or rock. Aspidosiphon elegans (Chamisso & Eysenhardt, 1821) Sternaspis elegans Chamisso & Eysenhardt, 1821:351-352, pi. 24, figs. 5a-e. Sipunculus elegans.— de Blainville, 1827, pi. 26, fig. 2. Loxosiphon elegans. — Diesing, 1851:70.— Quatrefages, 1865:605. Phascolosoma (Aspidosiphon) elegans. — Griibe, 1868a:645-647. Aspidosiphon elegans. — Selenka et al., 1883: 124-1 26. -Shipley, 1898:471; 1899b: 153.-Whitelegge, 1899:393. -Sluiter, 1891:116; 1902:1 9. -Herubel, 1904: 564.-Lanchester, 1905a:33; 1905b:40.- Fischer, 19 14b: 14. -Gravely, 1927:87.- Sato, 1935:316; 1939:426-427. -Wes- enberg-Lung, 1954:10-11; 1957a:198- 199; 1957b:7-8; 1959c:68. -Cutler, 1977b:154. -Cutler & Cutler, 1979a: 968.-Edmonds, 1 980:44-46.- Cutler et al., 1984:304.-Migotto&Ditadi, 1988: 248-250. Aspidosiphon elegans elegans. — Stephen & Edmonds, 1972:223. Aspidosiphon elegans var. yapense Sato, 1935:316-318, pi. 4, fig. 18, text-figs. 12- 15. Aspidosiphon elegans yapensis. — Stephen & Edmonds, 1972:224. Aspidosiphon brocki Augener, 1903:328- 330, figs. 9-13.-Murina, 1967b:42.- Rice, 1970:1618-1620; 1975:44-45.- Stephen & Edmonds, 1972:221.— Rice & Macintyre, 1979:311-319. Aspidosiphon carolinus Sato, 1935:31 8-3 1 9, pi. 4, fig. 19, text-figs. 16-17. — Stephens & Edmonds, 1 972:222. - Cutler & Cutler, 1981:77-78. Aspidosiphon cylindricus. —Sluiter, 1902: 18-19. Aspidosiphon exilis Sluiter, 1886:497, pi. 3, figs. 11-12; 1891:116; 1902:18.-Leroy, 1 942:39-40. - Stephen & Edmonds, 1 972: 224-225. -Edmonds, 1980:44-46. Aspidosiphon homomyarium Johnson, 1 964: 332-334, pi. 8. Aspidosiphon homomyarius. —Stephen & Edmonds, 1972:227. Aspidosiphon ravus Sluiter, 1886:495-496, pi. 3, figs. 9-10; 1891:116; 1902:18.- Shipley, 1899a:56; 1899b: 153. -Stephen & Edmonds, 1972:234. Aspidosiphon spinalis Ikeda, 1904:47-49, text-figs. 12, 81-85; 1924:37. -Sato, 1939:428.-Stephen & Edmonds, 1972: 234-235.-Cutler & Cutler, 1981:79-81. Aspidosiphon spinosus Sluiter, 1902:28, pi. 2, figs. 17-19.— Stephen & Edmonds, 1972:235. Material examined. — ZMUA, Skater's from Indonesia (V. Si. 3); worms from the western Pacific Ocean (Cutler et al. 1984), 1988 collections from the southern Carib- bean, a few from Brazil (Migotto & Ditadi 1988); A. brocki, MNHU, type material (6954-5); USNM, Caribbean material iden- tified by M. Rice (USNM 48924-5); A. cy- lindricus, ZMUA, two of Sluiter' s (26.8); A. exilis, BMNH and ZMUA (V. Si. 4), parts of type material in both places, all of it badly dried out; A. homomyarius, RSME, two worms from India presumably deposited by Johnson (1965.32.1); A. ravus, ZMUA (V. Si. 13) and BMNH, type material. Discussion. — Like A. {Par aspidosiphon) steenstrupii in its subgenus, A. elegans is the most common and widespread tropical VOLUME 102, NUMBER 4 843 member of this subgenus with many junior synonyms. Additionally, both species have ungrooved anal shields, bidentate com- pressed hooks in rings, and dark scattered hooks. Edmonds (1980:44-45) presented a detailed description of A. elegans and some interesting comments on this complex of related (or identical) taxa. What he illus- trates as introvert spines (his figs. 77-78) are, by our definition, conical hooks. The variation in the shape (real and due to ori- entation on slide) of the bidentate hooks has led to the creation of different species. Fig- ure 6 illustrates some of this variation (de- gree the hook is bent and sharpness of the point) that we now know to be within-deme variation. Six to twelve short stubby nuchal tentacles are present. The caudal shield is weakly developed, barely discernable in many worms. Internally the longitudinal muscle layer may subdivide in the area of the anal shield. Less than 20% of the worms dissected have a caecum and a fixing muscle was seen in only 4% of the worms dissected. The ne- phridia open at the level of the anus or just posterior to it. Table 2 shows our obser- vations on the nephridia and retractor origins. The gut has the normal helical coil, but about half the worms show a degree of looseness in part of the coil. The putative species A. carolinus and A. spinalis were reduced to junior synonyms by Cutler & Cutler (1981) and A. exilis by Edmonds (1980). We reaffirm those deci- sions. Aspidosiphon brocki (Augener 1903) is a name used by only two authors since it was described: Murina (1967b) and Rice (1970, 1975). Neither of these authors has used the name A. elegans in their writings. Augener' s paper was the first of only two articles he wrote about this phylum, and he made no reference to any other member of this sub- genus nor was there a differential diagnosis. The original report was from Malaya but the four subsequent records are all Carib- bean, and all of this latter material is less Fig. 6. Variation in shape of bidentate compressed hooks within a single specimen of A. elegans. Note difference in degree of bend and sharpness of point. Scale line = 20 jum. than 10 mm long. Our recent Caribbean collections (340 worms) ranged from 4 to 22 mm long. The pictures of hooks in Ste- phen & Edmonds (1972:230, fig. 27B, F, G) illustrate the extremes of a continuum and could be misleading. This much variation occurs within demes. Comparison of the type, Rice's material, and our own Carib- bean and Pacific material convinces us that this putative species is conspecific with A. elegans. Rice (1970) reported asexual reproduc- tion by budding in A. brocki, something not recorded elsewhere in this phylum. Our Ca- ribbean material showed this (even in 4 mm worms), but it is also present in our collec- tions from Majuro, Marshall Islands, and in French Polynesian worms collected by Peyrot-Clausade. Sluiter's (1 902) Siboga report included A. cylindricus, Horst, but not A. elegans. Our examination of his material showed no dif- ferences from A. elegans. When Johnson (1964) described A. ho- momyarius, he did not include a differential diagnosis except to contrast it with the other new species in that paper, so we do not know how he thought it differed from A. elegans. 844 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON He seemed to think that the dorsal array of nuchal tentacles was unique in this genus. Our examination of his two worms in Edin- burgh confirmed our analysis of his article, i.e., meaningful differences from A. elegans are lacking. Aspidosiphon ravus was described by Slui- ter (1886) as having unidentate hooks. Our reexamination of the material (the Amster- dam worms are in good condition but the London worm is not) revealed bidentate compressed and conical hooks. The intro- vert is retracted in all three worms. This material is clearly conspecific. Aspidosiphon spinosus Sluiter, 1902, was based on a single worm that has subse- quently been lost. His description is indis- tinguishable from A. elegans. He compared it to A. exilis differentiating it by the form of the hooks and a few other features. Plac- ing this name on the list of incertae sedis was considered since it cannot be examined (the hypothesis cannot be tested) but re- ducing it to a junior synonym of this species is a rational alternative given what we now know about within-deme variation. The subspecies A elegans yapensis, which Sato (1935) described as a variety, was dif- ferentiated on the basis of hook morphol- ogy: sharply pointed apex, not blunt. The within-deme variation we find makes it clear that this subspecies lacks biological signif- icance. Distribution.— Widespread and common in the Indian and western Pacific Ocean (from southern Japan to northern Australia out to Hawaii); the Red Sea and Israel. In the Caribbean from northern Brazil to the Florida keys and Bermuda. Aspidosiphon exiguus Edmonds, 1974 Aspidosiphon exiguus Edmonds, 1 974: 187— 192, figs. 1-7. Material examined. — BMNH, holotype (1975.22.1). Discussion.— The largest worm reported is less than 4.5 mm long, but it does bear bidentate hooks in rings as well as a few scattered unidentate compressed hooks. The introvert is very long (2-5 times the trunk) and no tentacles or gametes have yet been observed. The anal shield consists of small, pale units with no grooves. At the anterior/ ventral border 1-4 cone-like papillae are present. It is similar to A. albus but has hooks and shorter nephridia (25-33%). In the anterior 1 0% of the trunk the longitu- dinal muscle layer is divided into bands vis- ible through the body wall. Aspidosiphon {Par aspidosiphon) parvulus is common in this area and is similar in several ways de- spite having longitudinal muscle bands (weakly developed). Distribution.— Cuba, intertidal, intersti- tial. Aspidosiphon gosnoldi Cutler, 1981 Aspidosiphon gosnoldi Cutler, 1981:445— 449, figs. l-4.-Migotto & Ditadi, 1988: 253-254. Aspidosiphon spinalis.— Cutler, 1973:175- 176. -Cutler & Cutler, 1979b:107. Material examined. — USNM, type (61624-5) and additional material from the western Atlantic Ocean (Cutler 1981). Discussion. — The anal shield is com- posed of randomly arranged flat units of relatively uniform size. The borders are usually distinct but dark skin papillae may be present at the anterior end of the trunk. The introvert is 1 . 5-3 times the trunk length and bears distal rings of 20-30 nm, biden- tate hooks (the secondary tooth may be small). Scattered, pale, pyramidal hooks cover much of the proximal part of the in- trovert. This is in the group of species where the longitudinal musculature commonly splits into irregular bundles under the anal shield and the retractors originate from the caudal shield. The intestine forms a normal helical coil, and the nephridia are 50-90% of the trunk length. Distribution.— Shelf waters (7-185 m) from Cape Hatteras to Florida and Brazil (to 23°S) living in gastropod shells. VOLUME 102, NUMBER 4 845 Aspidosiphon gracilis gracilis (Baird, 1868) Pseudoaspidosiphon gracile Baird, 1868:103, pi. 10, figs. 1, la. Aspidosiphon gracilis— Selenka et al., 1883: 122-123. -Sluiter, 1902:1 7. -Augener, 1903:31 9-32 l.-Herubel, 1904:564.- Rice & Stephen, 1970:69. -Stephen & Edmonds, 1 972:225-226. -Gibbs, 1978: 85. -Edmonds, 1980:46-47. Material examined. — BMNH, syntypes (43.5.1 5. 58a/b); MNHN, two of Herubel's specimens (V20). Discussion. — This species has an under- developed anal shield composed of non- contiguous brown papillae surrounded by darker platelets. These units are arranged in irregular longitudinal rows. The introvert comes off at about a 60° angle, is up to 1.5 times the trunk length, and the slender trunk (up to 1 5 times the width) is coarsely pap- illated all over. The assertion that bidentate hooks are present (Stephen & Edmonds 1972) must be a flawed translation since we cannot find any other reference to them. The unidentate hooks are in rings, are broader than high (up to 40 /mi), and are followed by a proximal area of pyramidal hooks. The retractor muscles originate very near the posterior end and the nephridia are as long as the trunk. Herubel (1 904) had three worms from the Gulf of Aden that look different but are damaged and desiccated such that we can- not confirm or refute his identification. Distribution. —Australia, Indonesia, Phil- ippines, Gulf of Aden and an unpublished record from the Andaman Islands. Aspidosiphon gracilis schnehageni Fischer, 1913, new status Aspidosiphon schnehageni Fischer, 1 9 13:99- 100, figs. 4-6; 1914b:15.-Wesenberg- Lund, 1955:13. -Ditadi, 1975:200-202. Par aspidosiphon schnehageni. —Stephen & Edmonds, 1972:252. Material examined. — ZMUH, type spec- imen (V2 127). Discussion. —The type specimen is in very poor condition, missing, among other parts, the distal end of the introvert and the re- tractor muscles. Ditadi (1975) redescribed this taxon, but his material cannot be lo- cated at the Los Angeles County Museum or the Hancock Foundation where it had been deposited. The anal shield was de- scribed as furrowed by Fischer but as ran- domly arranged plates by Ditadi. This is another situation where the units may sometimes appear arranged in rows giving an impression of indistinct ridges and fur- rows. The longitudinal musculature is par- tially separated into 10-14 anastomosing bundles in the anterior part but is contin- uous elsewhere. In this regard, it is one of the borderline taxa not having distinct bun- dles, and, therefore, we have moved it into this subgenus. The decision to reduce this to a subspecies rather than a junior synonym was based on ill-defined differences from the nominate form; habitat (mollusc shells), hook shape (more triangular), nephridia (shorter, 33- 50%), trunk size and shape (length less than 8 times the width), and longitudinal muscle layer (splitting extends beyond the anal shield). These are not clear or distinct dif- ferences by themselves, but if one adds the geographical gap between the populations (most of the Pacific Ocean), this status may be appropriate pending additional material for better comparison. The west coast of Central and South America are very poorly represented in curated, accessible collec- tions. Distribution. —Chile and Pacific coast of Guatemala. Aspidosiphon misakiensis Ikeda, 1 904 Aspidosiphon misakiensis Ikeda, 1904:41- 43, text-figs. 9, 68-72. -Sato, 1939:428.- Stephen & Edmonds, 1972:229-231.- Cutler & Cutler, 198 1:78-79. -Cutler et al., 1984:305-306. 846 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Aspidosiphon hartmeyeri Fischer, 19 19:28 1— 282, text-figs. 1-3; 1926:204-205. -Ed- monds, 1956:306-307; 1 980:47. -Muri- na, 1967c: 1332. — Stephen & Edmonds, 1972:226-227. -Cutler, 1977a: 147-148. NotWesenberg-Lund, 1957a:7-8; 1959a: 197; 1959b:212. Aspidosiphon gerouldi ten Broeke, 1925:93, text-figs. 23-25. — Stephen & Edmonds, 1972:225. -Cutler & Cutler, 1979b:106- I07.-Migotto & Ditadi, 1988:251-253. Aspidosiphon speculator. — Saiz Salinas, 1986a:ll-14. Material examined. — USNM, paraneo- type; other material from the type locality (Cutler & Cutler 198 1); A. gerouldi, ZMUA, type (V. Si. 7), material from Azores (Cutler & Cutler 1987), and Brazil (Migotto & Di- tadi 1988); A. hartmeyeri, ZMUH (V8913- 14) and MNHU (6036), co-types; UZMK, Wesenberg-Lund's Niger material (=A. al- bus); A. speculator, three specimens iden- tified by Saiz Salinas. Discussion. — The anal shield is com- posed of closely packed, irregular, granular units, but it has borders that are not sharply defined, i.e., widely spaced, square blocks of shield material are around the anterior quarter of the trunk. The caudal shield is granular but does have vague radial grooves present. Bidentate hooks are present (25-40 jum) in distal rings, then proximally the hooks are scattered, unidentate com- pressed, 25-60 (xm tall (Fig. IB). The sec- ondary tooth has normal dimensions on distal hooks but becomes very small in proximal ones. The introvert is up to three times the trunk length (largest known worm is 25 mm). The longitudinal muscle layer exhibits some fractures/splits in some in- dividuals, and the gut coils are ill defined or only loosely wound. In 6-7 mm worms the intestine has a few loose folds, is an- chored at the posterior end, then has a straight tube to the anus. Sheets of connec- tive tissue link the ascending and descend- ing loops. The nephridia are 50-100% of the trunk and the retractors originate very close to the caudal shield. When Fischer (1919) described A. hart- meyeri from Australia, he made no refer- ence to Ikeda's A. misakiensis even though it is clear he knew of Ikeda's paper. Stephen & Edmonds' (1972) key uses the nephridia being mostly free to separate it from Ikeda's. Our examination of five type specimens showed considerable variation (up to 80% attached and length up to 1 00% of the trunk, not 0% and 50% as stated). The anal shield was described as having 5-6 flat grooves. These are not grooves but irregular units arranged in indistinct rows in a few worms (see also Edmonds 1956:306). Wesenberg- Lund's four worms from West Africa are clearly part of the A. albus population. She never recorded hooks, and her drawings of one, plus our examination of another, con- firm this. The decision to reduce the status of A. gerouldi came after examination of unpub- lished material from the Azores and com- parison to the Japanese worms. The original description was based on a single worm, but several hundred have been subsequently collected. The bidentate hooks in the Azores population are at the small end of the range (25-30 fim) as is the introvert (only up to twice the trunk in worms up to 20 mm long). Saiz Salinas (1986a) used the name A. speculator for 1 8 worms off Spain and the Canary Islands. This population is partic- ularly troublesome as indicated by his plac- ing it in A. {Par aspidosiphon). The longi- tudinal musculature is neither continuous nor divided into distinct bands. It varies from worm to worm in degree, but we in- terpret it to be continuous with fracturing in the anterior part, especially on the dorsal side. In other respects (hooks, shield, gut) his worms fit A. misakiensis better than any other taxon. Distribution. —In the Pacific from both sides of central Japan at 1-50 m depth, South VOLUME 102, NUMBER 4 847 and West Australia, and Kermadec Island. In the eastern Atlantic from the Azores, Cape Verde, and Canary Islands to the Gulf of Guinea, at depths down to 75 m, and the Spanish Mediterranean. In the western At- lantic from Brazil (14-16°N), Haiti, and Cuba. Aspidosiphon muelleri Diesing, 1851 Aspidosiphon muelleri Diesing, 1851:68.— Quatrefages, 1865:609-610.-Schmidt, 1865:56-66. -Baird, 1868: 101. -Selen- kaetal., 1883: 120-1 21. -Fischer, 1895: 18; 1914a:69-70; 19 14b: 13-1 4; 1922a:22- 23; 1925:25-26.-Sluiter, 1900:14; 1902: 18; 1912:19. -Herubel, 1904:564.- Southern, 19 12:3 1-34. -J. Fischer, 1914: 105-106. -Ikeda, 1924:38. -Stephen, 1934:173; 1941:257; 1958:133-134; 1960a:518-519; 1960b:22-23.-Steuer, 1936:5; 1939:2. -Sato, 1939:428.- Chapman, 1955:351.— Wesenberg-Lund, 1957a:4-5; 1957b:197-198; 1959a:194- 196; 1959c:68.-Longhurst, 1958:1.- Stephen & Edmonds, 1972:231-233.- Zavodnik&Murina, 1975:127; 1976:81- 82.-Cutler, 1977a:148.-Gibbs, 1977: 30-31. -Cutler & Cutler, 1979b:107; 1987a:73.-Ocharan, 1980:114-115.- Cutler et al., 1984:306-307. -Saiz Sali- nas, 1984:177-178; 1986a:9-ll. Sipunculus scutatus J. Muller, 1 844: 1 66-1 68 (not scutatum J. Muller, 1843).— Kefer- stein, 1867:52.-Selenkaetal., 1883:120. Phascolosoma scutatum. — Krohn, 1851: 371.-Selenkaetal., 1883:120. Aspidosiphon clavatus. —Diesing, 1851: 68.-Cuenot, 1922:12-13. -Herubel, 1924:1 11. -Leroy, 1 936:426. -Akesson, 1958:206.-Voss-Foucart et al., 1977: 135. Pseudaspidosiphon clavatum.— Baird, 1868: 103. Sipunculus cochlearius Valenciennes, 1854: 640. -Saiz Salinas, 1986b:554. Lesinia farcimen Schmidt, 1854:2. — Selen- kaetal., 1883:120. Aspidosiphon eremita Diesing, 1859:768 (not Phascolosoma eremita Sars, 1851). Phascolosoma radiata Alder, 1860:75.— Southern, 1913:32. Sipunculus heterocyathi McDonald, 1862: 78-81. -Saiz Salinas, 1986b:554. Aspidosiphon jukesii Baird, 1873:97. — Rice & Stephen, 1970:68-69. -Stephen & Ed- monds, 1972:228. -Cutler & Cutler, 1979a:969-970. -Edmonds, 1980:49.- Saiz Salinas, 1986b:551. Aspidosiphon mirabilis Theel, 1875:17; 1905:91-92.-Selenkaetal., 1883:121.- Southern, 1913:31-33. Aspidosiphon armatum Danielssen & Ko- ren, 1880:464; 1881:64.-Selenka et al., 1883:124.-Theel, 1905:93. -Southern, 1913:31-33. Aspidosiphon tortus Selenka et al., 1883:1 1 9— 120, pi. 14, figs. 196-201. -Herubel, 1904:564.-Fischer, 1923:21-22.-Ste- phen & Edmonds, 1972:236-237.- Gibbs, 1978:85. Aspidosiphon heteropsammiarum Bovier, 1894:98. -Saiz Salinas, 1986b:555-557. Aspidosiphon michelini Bovier, 1894:98.— Saiz Salinas, 1986b:557-559. Aspidosiphon corallicola Sluiter, 1902: 19-22. -Shipley, 1903: 169-1 71. -Ste- phen & Robertson, 1952:44 1-442. -Cu- tler, 1965:58. Aspidosiphon imbellis Sluiter, 1902:29, pi. 2, fig. 20.-Stephen & Edmonds, 1972: 227-228. Aspidosiphon inquilinus Sluiter, 1902:29-30, pi. 2, figs. 21-22. — Stephen & Edmonds, 1972:227. -Edmonds, 1980:47-49. Aspidosiphon exhaustum Sluiter, 1912:20— 21, pi. 1, fig. ll.-Murina, 1971:78. Aspidosiphon exhaustus. — Stephen & Ed- monds, 1972:224.-Murina, 1972:295- 296; 1 978: 120. -Cutler & Cutler, 1979a: 969; 1980a:4. -Edmonds, 1980:46.- Cutleretal., 1984:305. Aspidosiphon exhaustus mirus Murina, 1974:1715-1716, fig. 2. 848 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Aspidosiphon pygmaeus Fischer, 1921:45— 47, text-figs. 1-7.— Murina, 1967a:54; 1971:78. Par aspidosiphon pygmaeus. —Stephen & Edmonds, 1972:251-252. Aspidosiphon kovaleskii Murina, 1 964b: 5 1- 55, figs. 1-5; 1970:66. -Stephen & Ed- monds, 1 972:229. — Zavodnik & Murina, 1975:127. -Cutler & Cutler, 1979a:970- 971. Aspidosiphon hispitrofus LiGreci, 1 980: 1 23- 134, figs. 1-4. Material examined.— ZMU A, Skater's Indonesian worm (V. Si. 253); material from the Azores, 1-600 m depth collected by Zi- browius and CENTOB; Celtic Sea from P. Gibbs; French Mediterranean from Voss- Fouchart; Japan (Cutler, Cutler & Nishi- kawa 1984); A. armatus, ZMUB, type (1745); A. corallicola, ZMUA, types (V. Si. 1); A. exhaustus, MOMV, type; Cutlers' Brazilian, Japanese and Indian Ocean worms; A. imbellis, ZMUA, type (V. Si. 25/5); A. inquilinis, ZMUA, type (V. Si. 25/6); A.jukesii, BMNH, type (1965.25.3); Indian Ocean material (Cutler & Cutler 1979a); A. kovaleskii, ZIAS, type; Indian Ocean worms (Cutler & Cutler 1979a); A. mirabilis, ZMUB, type (15957); A. pyg- maeus, NHRS, type specimens (279). Discussion. — This species is the wide- spread, eurytopic, polymorphic member of this genus and in this way is comparable to Golfingia margaritacea, Phascolion strom- bus, Sipunculus nudus, etc. Each genus seems to have one such species with a long list of junior synonyms and a morphology difficult to define with precision. Stephen & Ed- monds (1972:232-233) discuss some of the early confusion about the proper name for this taxon. Most records are from the north Atlantic Ocean and Mediterranean Sea from abandoned mollusc shells. When biologists found similar worms from different parts of the world or from different habitats, new names were put forth. Our present concept of A. muelleri has an anal shield made up of very small units ar- ranged into variable sized plates, partially separated by longitudinal furrows dorsally, in the midsection by transverse furrows, and made up of raised wart-like or cone-shaped units ventrally (Fig. 5C). The possibility that two species exist, one with pointed, cone- shaped spines on the ventral part of the anal shield and the other with only flat wart-like units, was considered. It is our conclusion that this species has the capability to re- spond to some environmental stimulus (e.g., pressure, temperature, or host shell shape), or that random allelic frequency shifts can occur, to produce anal shields with ventral units varying in degree of cone develop- ment. The trunk may be straight or coiled de- pending on its habitat. The introvert is 1-3 times the trunk length, and there are 6-12 small nuchal tentacles. Hook morphology has been a long standing point of confusion (see Southern 1913, Stephen & Edmonds 1972:233). Our analysis of within-deme variation of compressed hooks clearly sug- gests that A. muelleri has the genetic poten- tial for producing only unidentate hooks or only bidentate hooks or some of both on a single worm. Proximal to the rings (covering about one-third of the introvert), the scat- tered hooks quickly change from com- pressed unidentate to pyramidal. The SEM photomicrographs reveal the existence of small comb-like structures at the posterior base of the compressed hooks (Fig. 1A). Internally the pair of introvert retractor muscles originates from the edge of the cau- dal shield. Under the anal shield the lon- gitudinal muscle layer divides into separate bands. The nephridia open at or just pos- terior to the anus, are from 25-100% of the trunk length, and the gut forms a regular helical coil. The rectal caecum and fixing muscle are only present in some individu- als. Aspidosiphon jukesii (Baird 1873) and its several junior synonyms must be moved into this species. When one considers each char- VOLUME 102, NUMBER 4 849 acter and looks not just at one author's per- ception but at the collected writings and the many specimens at hand, the presumed "bridgeless gap" ceases to exist. If a young worm occupies a gastropod shell in an area where solitary corals live, it is possible that this shell will be settled upon by a coral larva establishing a mutualistic relationship. This fact does not make the worm a different species. While it is true that most A. jukesii had anal shields with some number of cone- shaped spines, this is not universal in, or restricted to, any particular population. The type of A. tortus (Selenka et al. 1883) is not with their other type specimens in the Berlin museum. They had one 25 mm worm with a longitudinal muscle layer continuous except for the anterior dorsal area where it had some separation without being clearly separate. This worm also had just one ne- phridium, a condition Fischer (1923) as- serted was only an anomaly when he de- scribed his two worms. This has been presumed to be a separate species because it lacks unidentate hooks, now not a valid basis for separation, even if true. Their con- cept of spines (Stacheln) may well overlap with our idea of unidentate hooks (see their fig. 200, 201). Herubel (1904) recorded one worm from Djibouti with no comments. When Gibbs (1978) listed this species from the Great Barrier Reef he did indicate some doubt with the (cf.) notation. Our analysis of the literature (including figures) con- vinces us that A. tortus is a junior synonym of A. muelleri. When Sluiter (1902) erected A. imbellis he asserted that this single 13 mm worm with retracted introvert lacked hooks. Our reexamination showed this to be an error. Rings of 25-35 fim hooks exist, some with a small secondary point. The spindle muscle is attached posteriorly and it is clearly con- specific with A. muelleri. Aspidosiphon inquilinis was also based on a single specimen that Sluiter (1902) differ- entiated from A. muelleri on the basis of hook and skin body morphology plus a mis- understanding of retractor origins. His spec- imen had lived in a scaphopod shell and the anal shield has a peculiar slant. The intro- vert skin is folded back over the ventral edge of the anal shield. Edmonds (1980) pointed out that Sluiter had overlooked the small cone-shaped units along the shield margin just as Edmonds' four worms exhibited. These five worms have only compressed and pyramidal unidentate hooks. They are con- specific with A. muelleri. In 1912 Sluiter erected A. exhaustus for a single 1 7 mm worm taken from a scapho- pod shell in the east Atlantic on the slope off Morocco. No differential diagnosis was presented and examination of the type shows it to be like the other A. muelleri of this region with unidentate hooks. The name went unused for 59 years, after which Mu- rina used it four times followed by the Cut- lers who used it three times for cold water worms from diverse locations. This usage was predicated on the false assumption that the real A. muelleri must bear some biden- tate hooks. The name A. pygmaeus has been used by two authors. Fischer (1921) asserted that the longitudinal muscle layer consisted of bands anastomosing so frequently that he could not count them. He did say that these were most numerous at the anterior end and ran together at the posterior end. This statement caused Stephen & Edmonds (1972) to place the species in the subgenus A. {Paraspido- siphon). When we examined the type ma- terial (four of his seven worms in good con- dition with introverts out and dorsal tentacles showing), we saw a continuous muscle layer with some fracturing under the anal shield, a condition common to many worms in this taxon. The coast of Chile is not a common location for A. muelleri, but Fischer's worms do fit this construct as do Murina's (1967a, 1971). Her second record was a repeat of the first for two worms from 1 50 m in the Gulf of Aden. Aspidosiphon kovaleskii Murina, 1964, was presumed to differ from A. muelleri be- cause it lacked bidentate hooks. Several populations (see Fischer 1895, Southern 850 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. 7. A. spiralis: A-B, Bidentate and unidentate compressed hooks with broad bases (both 20 nm high); C, Anal shields with irregular squares in offset rows. Scale line in C = 2 mm. 1913, Gibbs 1977) of A. muelleri from the eastern Atlantic are reported to lack biden- tate hooks. By Murina's definition these should be A. kovaleskii. Recently Saiz Sa- linas (1984:177-180) provided a detailed account of these two taxa and proposed that in both species both kinds of compressed hooks exist. He suggested that A kovaleskii, if valid, might differ by having cone-shaped spines on the anal shield. He did express strong reservations about the validity of these taxa as separate and distinct biological entities; we conclude that they are not. Aspidosiphon hispitrofus LiGreci, 1980, was based on a series of worms from Sicily. None had their introverts all the way ex- tended, and he reported these to have no tentacles and only unidentate hooks. He did not mention the name A. muelleri in his paper but compared his material to A. cla- vatus. It is clear that his material is conspe- cific with A. muelleri and his unfamiliarity with the phylum led him to misinterpret the distal end of the introvert. Distribution. —Common in the northeast Atlantic from Norway through the British Isles, the Azores and Canary Islands and West Africa (to 10°N). It extends through the Mediterranean and Red Sea into the Gulf of Aden and along the coast of east Africa to Madagascar and South Africa. The rec- ords then skip to Ceylon and sparse reports up to Japan through Thailand, Vietnam, In- donesia, and down to Australia, New Guinea, and Kermadec Island. Most of the Pacific Ocean is unpopulated by this species. One record from Juan Fernandez Island off Chile (33°S) and one from southern Brazil VOLUME 102, NUMBER 4 851 (34°S) exist, but nothing else from the American hemisphere. This southern hemi- sphere distribution is intriguing. Through- out most of its range this species inhabits shelf depths (10-300 m), but there are sev- eral records down to 1000 m with the deep- est at 2930 m. Aspidosiphon spiralis Sluiter, 1902 Aspidosiphon spiralis Sluiter, 1902:25-26, pi. 2, figs. 9-13.-Shipley, 1903:171.- Stephen & Edmonds, 1972:236. Material examined. — ZMUA, type spec- imens (V. Si. 171-174). Discussion. —While Sluiter said this species has unidentate hooks, rings of small bidentate hooks are also present. These are about 20 nm high and the secondary point is not large (Fig. 7A, B). The unidentate hooks are about the same size and in rings. A gradual transition may exist during growth as there are some intermediate hooks. The anal shield does not have regular furrows, but neither is it made up of uniform gran- ules. It is divided into irregular squares overlain with horny protein (Fig. 7C). The retractor muscles originate from the pos- terior end of the trunk. This species does not have a firm foundation (four worms), but we propose no change in its status at this time. The three worms Shipley identified can- not be located, and we question whether he was looking at the same entity. His descrip- tion of the agglutinated sand packed around the opening of the gastropod shell is much more typical of Phascolion species. Also, he provided no morphological information (other than color) or station data. Distribution. —Indonesia, gastropod shells, 14-91 m. Subgenus Aspidosiphon {Par aspidosiphon) Diagnosis.— Introvert with compressed hooks in rings, longitudinal muscle layer di- vided into anastomosing bands. All bore in coral or rock. Aspidosiphon coyi Quatrefages, 1865 Aspidosiphon coyi Quatrefages, 1865:608- 609 (partim).-Baird, 1868:101. -Ste- phen & Edmonds, 1972:340.-Saiz Sali- nas, 1984:42-49. Phascolosoma truncatum Keferstein, 1867: 50-53, pi. 6, figs. 15-18. Aspidosiphon truncatus. — Selenka et al., 1883:1 18-1 19. -Selenka, 1885:20.- Sluiter, 1898:444; 1902: 17. -Shipley, 1899b:154; 1902:1 32. -Herubel, 1904: 564.-Ikeda, 1 904:38-39. -Lanchester, 1905a:34. -Fischer, 19 14b: 15. -Ben- ham, 1912:136. — Hammerstein, 1915: 2.-Sato, 1939:428. -Cutler & Cutler, 1979a:976.-Cutleretal., 1984:309-310. Par aspidosiphon truncatus. —Stephen & Edmonds, 1972:258. Material examined.— MNHN, type ma- terial (V20); A. truncatus, ZMUA, Sluiter's Indonesian worm (V. Si. 25.1); from Mo- zambique (Cutler & Cutler 1979a) and Ja- pan (Cutler et al. 1984); USNM, unpub- lished material from the Gulf of Panama (21477), and Gulf of California, Puerto Pen- asco, Mexico (26443). Discussion. —A. coyi is one of two species in this subgenus with a grooved anal shield. The primary distinction from A. laevis is the presence of distal rings of bidentate hooks 25-35 /mi tall. However, the second- ary tooth on these is very small and not consistently present. There can be uniden- tate compressed hooks as well as thin py- ramidal hooks, and the presence of both has led to some confusion in the literature. The skin papillae towards the two ends of the trunk are large and rugose. The longitudinal muscle layer can exhibit much anastomos- ing and is not always clearly banded. The retractors originate from the posterior end (95-100%), the spindle muscle may bifur- cate near the anus, and the wing muscle is well developed extending down to near the ventral nerve cord. The nephridia are 40- 95% of the trunk length and attached for most of their length. This has not been a 852 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON well defined species and the size of the data base is small. Striking similarities to A. (As- pidosiphon) muelleri exist (e.g., shield, re- tractors, hooks), and show another example of the less than clear boundary between the two subgenera. The careful redescription of this species by Saiz Salinas (1984) brought the name A. coyi back from its earlier dubious status (Stephen & Edmonds 1972) and these taxa are clearly conspecific. The location of Qua- trefages' specimens is not clear but is prob- ably somewhere in the Indian Ocean. Distribution. — Several locations in the western Indian Ocean and the western Pa- cific from Japan through Okinawa, Philip- pines, Indonesia, and Kermadec Islands. Two Latin American records of uncertain location exist: The type of A. truncatus from Panama (east or west coast?) and a single 3 mm worm from San Salvador (Bahamas or Galapagos?). Eastern Pacific locations for these two are supported by the USNM ma- terial from the Gulf of Panama and the Gulf of California. The former was collected in 1866 and may be part of Keferstein's type collection. Aspidosiphon fischeri ten Broeke, 1925 Aspidosiphon fischeri ten Broeke, 1925:92- 93, figs. 21-22.-Migotto&Ditadi, 1988: 250-251. Par aspidosiphon fischeri fischeri. — Stephen & Edmonds, 1972:244-245. -Amor, 1975:1 18-1 19. -Rice, 1975:38-44.- Rice & Macintyre, 1979:311-319. Aspidosiphon fischeri cubanus Murina, 1967b:39-42, figs. 5-7, 1967c: 1331. Par aspidosiphon fischeri cubanus. —Ste- phen & Edmonds, 1972:245. Material examined.— ZMU A, type ma- terial (V. Si. 5); USNM, Venezuelan spec- imens identified by M. Rice (100901); un- published specimens labeled A. truncatus (20726, 20725, 20727, 20836); our 1988 collections from the Caribbean. Discussion. —The body wall of these small worms (trunks 4-16 mm) is smooth, thin, and white, but the longitudinal muscles bands cannot be easily seen through it, mak- ing it easy to misplace these worms in A. (Aspidosiphon) during preliminary sorting, i.e., they are similar to A. A. misakiensis or A. A. gosnoldi. Few, thin, ill-defined muscle bands exist that anastomose frequently and seem to fuse in the posterior part of the trunk in some worms. The long introvert (one or two times the trunk) bears rings of compressed, 18-27 jum, bidentate hooks. The proximal rings may include hooks with a very small secondary point and unidentate hooks mixed together or a few rings of just unidentate hooks. Following these are scat- tered, pale, pyramidal hooks 15-50 ^m tall. The retractor muscles are thin and originate at or very near the posterior end of the trunk (95-100%). The nephridia are short; 33-50% of the trunk length. When ten Broeke described this species she provided no differential diagnosis, and the differences from other species, e.g., A. parvulus, or those mentioned above, are not always distinct. One might make a case for this being a subspecies or an incipient or sibling species, but we propose no change now. Migotto & Ditadi (1988) question the basis for the subspecies A. fischeri cubanus Murina, 1967, and we agree that it only rep- resents some of the variation within the species. Distribution.— Numerous southern Ca- ribbean locations from Cuba to Sao Paulo, Brazil, in shallow coral rock. Also from the Pacific coast of Panama, Ecuador, plus James and Hood Islands, Galapagos. Aspidosiphon laevis Quatrefages, 1865 Aspidosiphon laeve Quatrefages, 1865: 609.-Baird, 1868: 102. -Vaillant, 1871: 272-273; 1875: pi. 4, figs. Cl-4.-De- Rochebrune, 1881:233. — Saiz Salinas, 1984:55-62. Aspidosiphon laevis. —Stephen & Edmonds, 1972:340. Aspidosiphon cumingii Baird, 1868:102, pi. VOLUME 102, NUMBER 4 853 11, fig. 2.-Selenka et al., 1883:1 13— 115. -Fischer, 1892:85; 1922b:12.- Collin, 1892:177. -Sluiter, 1898:444; 1902:1 7. -Augener, 1903:321-322.- Herubel, 1904:564. -Monro, 1931:34.- Leroy, 1936:426.— Andrew & Andrew, 1953:1. -Rice & Stephen, 1970:67.- Cutler, 1973:179-180. Paraspidosiphon cumingii. —Stephen & Ed- monds, 1972:243-244.-Edmonds, 1980: 50. Aspidosiphon major Vaillant, 1871:270- 271; 1875: pi. 4, figs. Al-6.-De- Rochebrune, 1881:232. Aspidosiphon klunzingeri Selenka et al., 1883:115-116, pi. 13, figs. 187-189.- Fischer, 1896:338; 19 14a: 70. -Sluiter, 1898:444; 1912:20.-Shipley, 1898:471; 1899b:153.-Herubel, 1904:564.-Mon- ro, 1931:34. -Edmonds, 1956:308.- Wesenberg-Lund, 1957b:8-9; 1959a: 196; 1959b:211-212; 1963:138. -Cutler & Cutler, 1979a:974-975; 1979b:107. Paraspidosiphon klunzingeri. —Stephen & Edmonds, 1 972:247-249. -Rice & Mac- intyre, 1972:42. -Rice, 1975:40-41.- Haldar, 1976:7. -Gibbs, 1978:85. Aspidosiphon gigas Sluiter, 1884:39-57, pi. 2, figs. 1-11, pi. 2a, figs. 12-25; 1886:473; 1891:116; 1902:19. Paraspidosiphon gigas. — Stephen & Ed- monds, 1972:246. Aspidosiphon angulatus Ikeda, 1904:45-47, figs. 11, 78-80; 1924:37.-Sato, 1939: 428. -Cutler & Cutler, 1979a:974; 1981: 81.-Cutleretal., 1984:308. Paraspidosiphon angulatus. —Stephen & Edmonds, 1972:241. Aspidosiphon speciosus Gerould, 1 9 1 3:426- 427, text-fig. 16, pi. 62, fig. 22.— Fischer, 1922c:13.-Migotto&Ditadi, 1988:254- 257. Paraspidosiphon speciosus. —Stephen & Ed- monds, 1972:253. -Rice, 1975:38-45.- Rice & Macintyre, 1979:314. Aspidosiphon grandis Sato, 1939:414-419, pi. 21, fig. 21, text-figs. 46-50. -Cutler & Cutler, 1981:83-84. Paraspidosiphon grandis.— Stephen & Ed- monds, 1972:246-247. Aspidosiphon grandis obliquoscutatus Mu- rina, 1974:1713-1715, fig. 1. Aspidosiphon pachydermatus Wesenberg- Lund, 1937:9-16, text-figs. 4-9. Paraspidosiphon pachydermatus. — Stephen & Edmonds, 1972:250-251. Aspidosiphon brasiliensis Cordero & Mello- Leitao, 1952:277-282, 288-292, text-figs. 1-5. Paraspidosiphon brasiliensis.— Stephen & Edmonds, 1972:241-243. Asp idosiphonjohnstoni Edmonds, 1980:5 1- 53,figs.91, 100-102.-Lopezetal., 1984: 194-196. Aspidosiphon quatrefagesi Saiz Salinas, 1984:49-55, fig. 4. Material examined. — MNHN, type ma- terial (V20); our 1985 Hawaii and 1988 Venezuela material; A. cumingii, type no longer at BMNH; A. angulatus, Madagascar and Polynesia (Cutler & Cutler 1979a); A. brasiliensis, type cannot be located; A. gigas, ZMUA, holotype (V. Si. 8); A. grandis, ZITU, type specimen (24); A.johnstoni, two specimens from Edmonds' original mate- rial; A. klunzingeri, type cannot be located; ZMUA, Sluiter's Durban worm (V. Si. 9); UZMK, Wesenberg-Lund's Cape Verde material; South Africa and Cape Verde (Cutler & Cutler 1979a, b); USNM, 26437 from Saipan identified by W. K. Fisher; A. pachydermatus, the type cannot be located; USNM, two specimens identified by W. K. Fisher (from Saipan, 24645 and Philip- pines, 21480); A. speciosus, USNM, type material (16820, 16391, 4088); two speci- mens from Brazil (Migotto & Ditadi, 1988). Nomenclatural note. — Quatrefages' orig- inal spelling is incorrect (Steyskal, pers. comm.) and when the correct ending is ap- pended the name A. lexis Sluiter becomes a junior homonym according to ICZN Art. 58. Discussion. —A. laevis is a widespread but low density population of worms that has 854 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fig. 8. Unidentate, compressed, Type A hooks of several A. laevis from different populations to show differences in shape and size, the latter being roughly correlated with trunk size. Scale line = 40 urn. been given several names over the past cen- tury. One of the characters that has been weighted very differently by various authors is the nature of the caecum and its elabo- rations as discussed above. Setting this aside, a second issue has been the presence/ab- sence of introvert spines. It is clear that when a few authors made reference to spines, they were either looking at scattered unidentate compressed hooks, or had a different species in hand. These worms have a solid anal shield bearing 10-15 longitudinal grooves. They have unidentate, compressed hooks, sharply pointed or blunt, in many rings (Fig. 8). These are 20-80 nm tall, hook size being roughly correlated with trunk size. Also present are a few scattered compressed hooks (referred to as spines by some earlier au- thors). Up to 24 tentacles surround the nu- chal organ. Internally, the pair of retractor muscles are fused for most of their length, some- times giving the impression of a single broad muscle with the ventral nerve cord running through a notch in the base. These muscles usually have their origins from the body wall about 65-80% of the distance towards the posterior end of the trunk, i.e., well in front of the caudal shield. Another distinc- tive feature is the bifurcation towards the anterior end of the spindle muscle. One branch continues along the rectum into the connective tissue and wing muscle to join the body wall just anterior to the anus. The second, and often larger branch, leaves the posterior rectum going to the dorsal body wall well posterior to the anus. In many specimens the contractile vessel is not a smooth tube but has vesicular pouches or swellings along the part of its length united with the retractors (not unlike that seen in some Golfingia specimens, and sometimes confused with contractile vessel villi [see Cutler & Cutler 1987:750]). The rectum usu- ally bears a caecum that may be simple or complex (see part 8, Morphological Char- acters section). The 25-35 longitudinal muscle bands anastomose frequently, and the circular muscle layer also subdivides into anastomosing bundles of varying degree of development. The nephridia open at, or just posterior to, the anus, are attached to the body wall for about half to two-thirds then- length, and are usually more than half the trunk length. The unpublished USNM material (21480 and 24645) consists of two worms, 44 and 123 mm long. They both have complex multiple caeca, a grooved anal shield, uni- dentate hooks (in the larger one these are blunt triangular and up to 60 iim tall), and in all ways match A. laevis. This new arrangement reduces ten puta- tive species to the rank of junior synonyms of A. laevis, a decision reached only after extended analysis. We shall not detail each case, but in a few instances there were ob- servational errors made by the original au- thor (for example, A. gigas does have rows of unidentate hooks, a spindle muscle and a complex rectal caecum, all overlooked by Sluiter). However, in most cases the differ- ent conclusions are based on judgments VOLUME 102, NUMBER 4 855 about within-deme variation, or simply lack of either critical comparison or differential diagnoses (e.g., Gerould [1913:427] when describing A. speciosus simply says: "This species resembles A. klunzingeri from the Red Sea" and nothing more as to how it differed). In another case Saiz Salinas (1 984) erected A. quatrefagesi based on a suite of characters he assumed to be absent in older species while actually they were present but simply not mentioned by earlier authors. As discussed in the Morphological Char- acters section, the variability in rectal ap- pendages is great, therefore, to use such a variable character to differentiate species is unwise. A more difficult issue for us, partly because of statements by Edmonds (1980) and Rice (1975), centers on the hook mor- phology. We have examined many hooks from museum specimens and newly col- lected material, comparing hooks from small worms to those from large ones, and hooks from anterior to posterior rings in the same worm (Fig: 8). We conclude that while pop- ulations do exhibit among-deme differ- ences, much of the alleged between species variation in shape is present within demes. Therefore, hook shape alone cannot be used to separate these putative species. Distribution.— Widespread, but low den- sity, in warm water Indo-West Pacific Ocean (Durban to the Red Sea, Andaman Islands, Malaya to southern Japan, Indonesia, Great Barrier Reef and several islands out to Ha- waii). Also present in the Caribbean and west Atlantic (from 20°S to 31°N), then in the east Atlantic from the Canary and Cape Verde Islands to the Gulf of Guinea. In- habits shallow water coral rock. Aspidosiphon tenuis Sluiter, 1886 Aspidosiphon tenuis Sluiter, 1886:491^92, pi. 3, fig. 7; 1891:116; 1902:19. Par aspidosiphon tenuis.— Stephen & Ed- monds, 1972:257. Aspidosiphon levis Sluiter, 1886:493-494, pi. 3, fig. 8; 1891:116; 1902:18. Par aspidosiphon levis. —Stephen & Ed- monds, 1972:249-250. Aspidosiphon ambonensis Augener, 1903: 325-328, figs. 5-8. Aspidosiphon steenstrupii var. ambonensis Fischer, 1922a:24-26; 1923:21. Par aspidosiphon ambonensis.— Stephen & Edmonds, 1972:240-241. Aspidosiphon formosanus Sato, 1939:421- 424, pi. 21, fig. 23, text-figs. 55-57. -Cut- ler & Cutler, 1981:81-83. Par aspidosiphon formosanus. —Edmonds, 1971:144-146; 1980:50-51. -Stephen & Edmonds, 1972:245. Aspidosiphon havelockensis Haldar, 1978: 37-41, figs. 1-2. Aspidosiphon speculator. —Cutler & Cutler, 1979a:975-976 (partim). Material examined.— BMNH, (1889.6.15. 42/44) syntype; A. levis, ZMUA, type ma- terial (V. Si. 11); A. ambonensis, MNHU, type material (6956-6958); A. formosanus, type material cannot be located, but we have examined material from Guam and Austra- lia identified by S. J. Edmonds; A. speculator, Pacific Ocean (Cutler & Cutler 1979a). Discussion. —A. tenuis is here defined as having an anal shield made up of very fine dark units with a smooth overall appear- ance; a few very short grooves may appear around the margin. The distal rings of bi- dentate hooks (30-60 j^m tall) lack the dis- tinct tongue on the clear streak (Fig. 2C) and are followed by scattered, unidentate hooks (25-60 jiim tall) with an internal clear streak in the distal ones. More proximally, these scattered unidentate structures have lateral reinforcing ridges. Dark pyramidal or con- ical hooks are absent. Internally this species is very much like A. steenstrupii except that only two of the ten worms dissected has a rectal caecum and the nephridia are less than 50% of the trunk length. When Sluiter (1 886) described this species he overlooked the distal rings of bidentate hooks and posterior attachment of the spin- 856 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON die muscle. He overemphasized the few small grooves around the margin of the anal shield. The caudal shields are not all dis- tinctly grooved. Therefore, this is clearly conspecific with A. levis from the same lo- cation, and becomes the senior synonym by virtue of their position in the text. Fischer (1922a) reduced Aspidosiphon ambonensis to a variety of A. steenstrupii, but Stephen & Edmonds (1972) elected to elevate it back to species rank since the shape of the clear area in the hook "seems to be different." We assume they based their con- clusion on Augener's (1903) picture (their fig. 29K is Augener's fig. 6). Our examina- tion of the type material (19 worms) con- firms Augener's perceptions and supports our position that this population is conspe- cific with A. tenuis, but different than A. steenstrupii in accordance with Stephen & Edmonds (1972). Sato (1939) erected A. formosanus, but Cutler & Cutler (1981) reduced it to a junior synonym of A. steenstrupii. Given our cur- rent understanding we would move A. for- mosanus into the synonomy of A. tenuis as it shares its attributes. In 1978 Haldar described Aspidosiphon havelockensis. In his unpublished disserta- tion (pers. comm.) it is reduced to a junior synonym of A. steenstrupii ambonensis dif- ferentiated from the nominate form on the basis of hook, spine, and papillae structure. We agree with his conclusions as far as they go. However, we now consider both these names to be junior synonyms of A. tenuis. Upon reexamination, the specimens Cut- ler & Cutler (1979a) identified as A. spec- ulator from the Solomon Islands and Thai- land belong in this taxon. Distribution. —Andaman Islands to Thai- land, Formosa, and Guam, out to the east- ern Caroline Islands and down through the Solomon Islands to the Great Barrier Reef and Indonesia. Aspidosiphon parvulus Gerould, 1913 Aspidosiphon parvulus Gerould, 1913:425— 426, pi. 61, fig. 17, text-fig. 15. -Stephen & Edmonds, 1972:233-234. -Cutler, 1973:178-179. Aspidosiphon spinoso-scutatus Fischer, 1922c:13-14, text-figs. 2-3.-Murina, 1967b:42; 1967c:1332. Par aspidosiphon spinososcutatus. —Ste- phen & Edmonds, 1972:254. — Rice, 1975:38-45. Material examined. — USNM, type (15118); western north Atlantic (Cutler 1973); our 1988 Venezuelan worms; A. spinoso-scu- tatus, MNHU, type (6053). Discussion. — This species shares many attributes with A. fischeri, but the shield morphology seems consistent and distinc- tive. Centrally it is made up of larger flat plates; these are sometimes arranged in rows giving an impression of ridges and grooves. Ventrally and laterally the units become smaller, scattered, wart- or cone-shaped. The shields have a diffuse boundary where the units grade into coarse trunk papillae. At both ends of the trunk these darker papillae are located in rectangles that remind one of Sipunculus skin. The hooks (bidentate and unidentate in rings, scattered unidentate, and pyramidal, Fig. IE) are 25-35 nm tall. The 10-12 short tentacles may appear webbed together with connective tissue, and there are about 24 anastomosing longitudinal muscle bands. These bundles are generally quite distinct in worms over 5 mm long, more so towards the anterior end (see Cutler 1973:178 where Gerould's 3 and 4 mm specimens are discussed). The nephridia are 50-75% of the trunk length. Distribution.— Western Atlantic Ocean from Cape Hatteras through the Caribbean to Venezuela; often found together with Themiste alutacea and Nephasoma pellu- cidum in branching corals. Aspidosiphon planoscutatus Murina, 1968 Aspidosiphon planoscutatus Murina, 1968: 1722-1724, figs. 1-2; 1971:78. Material examined. — ZI AS, type mate- rial. VOLUME 102, NUMBER 4 857 Discussion. — This taxon is based on two specimens and is very similar to A. steen- strupii but has only unidentate compressed hooks on the introvert. Also, the shield units are smaller (fine grained like A. zinni) and the trunk is densely covered with more ob- vious skin bodies. The nephridia are 85% of the trunk length. The Red Sea is a marine habitat with un- usual abiotic conditions (e.g., high salinity, low oxygen, high temperature) that may re- strict gene flow and favor selection of dif- ferent allelic frequencies. The absence of bi- dentate hooks on the introvert may be a real difference, but we cannot verify this. De- spite our reservations we are leaving the name because of the habitat. It is hoped that more collecting in the area will produce ad- ditional material for analysis. Distribution. —Red Sea at 40 m. Aspidosiphon steenstrupii Diesing, 1859 Aspidosiphon steenstrupii Diesing, 1859:767, pi. 2, figs. l-6.-Quatrefages, 1865:610.- Selenka et al., 1883:1 16-11 8. -Sluiter, 1886:489-490; 1891:115; 1902:18.— Whitelegge, 1899:394.-Shipley, 1899b: 153-154; 1902:131-132; 1903:171.- Ikeda, 1904:40-41; 1 924:38. -Herubel, 1904:564. -Lanchester, 1905b:39.- Fischer, 1914a:70-71; 1914b:13; 1922a: 23; 1922c:13; 1923:21; 1926:108; 1931: 139. -ten Broeke, 1925:93-94. -Monro, 1931:34. -Sato, 1935:315-316; 1939: 424-426. -Leroy, 1936:426; 1942:36- 38. -Stephen, 1942:253. -Stephen & Robertson, 1952:441. -Edmonds, 1956: 307-308. -Wesenberg-Lund, 1959a: 197- 198; 1963:138. -Murina, 1967b:42; 1981:12-13. -Cutler, 1977a: 148. -Cut- ler & Cutler, 1979a:976; 1979b: 107- 108. -Cutler et al., 1984:308-309.- Migotto & Ditadi, 1988:259-260. Par aspidosiphon steenstrupii steenstrupii. — Stephen & Edmonds, 1972:254-255.- Rice, 1975:38-45. -Haldar, 1976:8.- Rice & Macintyre, 1972:42; 1979:311- 319. -Edmonds, 1980:51. Aspidosiphon steenstrupii var. faciatus Au- gener, 1903:322-325, figs. 1-4. Par aspidosiphon steenstrupii fasciatus. — Stephen & Edmonds, 1972:255-256. Aspidosiphon fuscus Sluiter, 1881:86-108; 1886:474; 1891:116; 1902:1 9. -Selenka etal., 1883:116. Aspidosiphon semperi ten Broeke, 1925:92, text-figs. 18-20.-Gibbs& Cutler, 1987: 56. Par aspidosiphon semperi. — Stephen & Ed- monds, 1972:252. Aspidosiphon speculator Selenka, 1885:19- 20, pi. 4, figs. 24-27. -Fischer, 1914b:71; 1920:413. -Wesenberg-Lund, 1959b: 213. -Cutler & Cutler, 1979a:975-976 (partim).— Not Saiz Salinas, 1 986a: 1 1-14. Par aspidosiphon speculator. —Stephen & Edmonds, 1972:253-254. Aspidosiphon makoensis Sato, 1939:419— 421, pi. 21, fig. 22, text-figs. 5 1-54. -Cut- ler & Cutler, 1981:82-83. Par aspidosiphon makoensis. —Stephen & Edmonds, 1972:250. Aspidosiphon trinidensis Cordero & Mello- Leitao, 1952:283-286, 292-294, figs. 6- 10.-Cutler& Cutler, 1979b:108; 1980c: 206. Par aspidosiphon trinidensis. —Stephen & Edmonds, 1972:257-258. Aspidosiphon exostomum Johnson, 1964: 331-332, pi. 7. Par aspidosiphon exostomus. —Stephen & Edmonds, 1972:244. Aspidosiphon ochrus Cutler & Cutler, 1979a: 976-979, figs. 15-1 7. -Edmonds, 1987: 204. Material examined. — ZMUA, Sluiter's 1902 specimens (V. Si. 21); our Pacific and Caribbean material; Brazil (Migotto & Di- tadi 1988); A. semperi, ZMUA, types (V. Si. 14); A. exostomum, RSME, type (1965.32.2); A. speculator, BMNH, 1885. 12.3.28, syntype; specimens from Canary Islands and Spain identified by J. Saiz Sa- linas; Madagascar (Cutler & Cutler 1979a). Discussion. —We came to the present un- derstanding of A. steenstrupii only after 858 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON looking at many worms from diverse lo- cations, and carefully reexamining the lit- erature. The color of the anal shield was a confusing element. It now seems clear that a range of colors is possible, from almost white to very dark brown, and that addi- tional calcareous material may be deposited externally, thus masking the underlying units. Geographic variation seems present with the Atlantic Ocean populations being dark, the mid Pacific Ocean populations being pale and the Indian Ocean popula- tions exhibiting a mixture (a higher fre- quency of dark shields in populations near continents, rare in island populations). The other major issue centers around the hooks. We are defining this species as hav- ing bidentate hooks in rings (30-60 ^m tall; up to 90 jitm in worms over 25 mm long), most with a tongue-like extension on the internal clear streak (Fig. 2B), and no uni- dentate compressed hooks. The proximal introvert does bear many scattered, dark, pyramidal hooks about 30-60 /*m tall. Internally the pair of retractor muscles originate about 70-85% of the distance to the posterior end of the trunk, not correlated with trunk size (see Table 2). The nephridia are commonly 50-80% of the trunk length and attached to the body wall for 50-75% of their length (over 90% in three worms). A simple rectal caecum was observed in 21 of the 30 worms dissected. The longitudinal muscle bands anastomose (14-22 anteriorly and 20-28 posteriorly), and the number is not correlated with trunk length. J. Silverstein (pers. comm.) determined the karyotype on a Japanese population. The 2N number is 20 with five pairs of meta/ submetacentric and five pairs of telo/sub- telocentric chromosomes. When Selenka (1885) erected A. specu- lator he made no reference to any other species (no differential diagnosis or key). In Stephen & Edmonds (1972) the key sepa- rates these two species based on the location of the retractor origins that were imprecisely stated in the original. In the syntype, the origins are at 75% of the distance to the posterior end of the trunk, well within the A. steenstrupii range. The internal structure of the compressed hooks and the nature of the pyramidal hooks, shield, and other char- acters all match this species. The part of Cutler & Cutler (1979a) collection that be- longs here are the Madagascar worms, the remainder are A tenuis. Saiz Salinas (1986a) used A. speculator for a collection that we consider to be A. misakiensis. Fischer (1922a) reduced Augener's 1903 A. steenstrupii fasciatus to a junior synonym of the nominate form where it remained until Stephen & Edmonds (1972) resurrect- ed its subspecies rank. They based their de- cision on the clear area in the hook and color differences on the shield and mid-trunk. Our examination of the type, within the context of this study, convinces us that Fischer's action was correct. When ten Broeke (1925) described A. semperi from Curacao she asserted that it had four retractor muscles, but Gibbs & Cutler (1987:56) determined that there are only two. Despite the pale colored anal shield A. semperi is clearly conspecific with A. steenstrupii, a conclusion confirmed by our recent collections in Curacao. Sato's two species, Aspidosiphon formo- sanus and A. makoensis were reduced to the status of junior synonym in Cutler & Cutler (1981). We reaffirm that action for the latter but not the former (see below). Aspidosiphon trinidensis was described from a single worm that cannot be located. The two subsequent reports were also based on single worms (in Cutler & Cutler 1980c, it should have read 25 mm trunk, not 125 mm). Reexamination of the two available worms revealed a few distal rings of biden- tate hooks that had been overlooked. Those structures reported as unidentate hooks are now interpreted as pyramidal hooks. Cor- dero & Mello-Leitao's worm (1952) had its introvert entirely withdrawn and the objects they described as unidentate hooks were not in rings. From their words and drawings we VOLUME 102, NUMBER 4 859 interpret these as pyramidal hooks. With this understanding (and the assumption that they too overlooked the bidentate hooks) nothing separates A. trinidensis from A. steenstrupii, thus we place it in synonomy. Aspidosiphon exostomus (Johnson 1964) was alleged to be different because of the dorsal crown of tentacles. This was clearly visible as the esophagus was protruding through the mouth. However, we now know that all members of the Aspidosiphonidae have dorsal tentacles and this taxon is clear- ly not unique. When Cutler & Cutler (1979a) described A. ochrus we were working within a different experiential framework. As a result of the present analysis, it clearly should be reduced to a junior synonym. In that same paper we identified a specimen from Madagascar as A. speculator that, upon reexamination, we now consider to be an A. steenstrupii. Ed- monds (1987) used the name A. ochrus after consulting with us, but these also are A. steenstrupii with pale anal shields. Distribution.— Throughout the western and northern Indian Ocean, Queensland through Indonesia and the South China Sea to southern Japan, out through the western Pacific islands to Hawaii. Also collected from numerous Caribbean locations, in the eastern Atlantic only from the Cape Verde Islands and the Gulf of Guinea. It lives in shallow water coral rocks. Zoogeographical Summary Ten of the 1 9 species live in the tropical/ subtropical western Atlantic Ocean and Ca- ribbean Sea, an area bounded by Cape Hat- teras on the north and the Amazon delta on the south (A. albus, A. exiguus, A. gosnoldi, A. parvulus, A. fischeri, A. mexicanus, A. elegans, A. laevis, A. steenstrupii, and A. mi- sakiensis). The first four of these are endem- ic to the region. The fifth extends into the eastern Pacific (Panama to Galapagos). The sixth extends its range in the other direction, to the eastern Atlantic (between Iberia and the Gulf of Guinea) but nowhere else. The next three species are circum-tropical while the last is found on both sides of the Atlantic Ocean and off Japan and Australia. Two species are found in the eastern At- lantic and elsewhere that do not live in the western part (A. venabulus from both sides of Africa, and A. muelleri, see below). Also in the north Atlantic (plus one record from the Mozambique Channel) is A. zinni, the one bathyal/abyssal member of this genus. Of special note is A. muelleri, since it has the most widespread distribution, almost cosmoplitan in temperate waters. Two ap- parent gaps occur; in the western Atlantic (except for one record off southern Brazil), and in the eastern Pacific there is only one record off Chile. This is the most eurytopic Aspidosiphon living in a much wider variety of temperatures and depths than other species. Six species are widely distributed within the Indo-West Pacific area. Aspidosiphon gracilis schnehageni and A. coyi extend into the eastern Pacific Ocean. Three are also found in the Caribbean (as above). The re- maining two do not get to Hawaii or the Atlantic (A. gracilis gracilis and A. tenuis). Two species (A. thomassini and A. spiralis) are more restricted within the Indian Ocean and A. planoscutatus is known only from a single collection in the Red Sea. Of particular interest is the number of endemic species in the warm water Atlantic/ eastern Pacific Ocean (six) when compared to the Indo-West Pacific area (five). Of the 1 9 species, 1 3 live somewhere in the Atlan- tic Ocean while 1 1 occupy some part of the Indo-West Pacific (six of these are in both areas). These data have interesting evolu- tionary implications suggesting that the tra- ditional "center of origin" hypothesis for marine invertebrates (Indo-West Pacific) may not fit Aspidosiphon. While it is true that common, widespread species bore in coral or soft rock, 1 1 species (58%) do not occupy this stereotypical as- pidosiphonid niche, i.e., they live in dis- 860 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON carded mollusc shells (8), arenaceous fora- miniferan tests (1), or interstitially (2). Acknowledgments The encouragement and exchange of ideas and specimens with S. J. Edmonds, Ade- laide, has contributed greatly to the com- pletion of this work. A working visit from J. Saiz Salinas, Bilbao, including an open exchange of perceptions, was very fruitful. The opportunity to read unpublished manuscripts and to examine some speci- mens of A. E. Migotto and A. S. F. Ditadi, Sao Paulo, and B. P. Haldar, Calcutta, has added significantly to our understanding. Our work in Hawaii was accomplished with the assistance of B. Burch, Bishop Museum and M. Hadfield, Kewalo Marine Labora- tory. The field work in Curacao was depen- dant on the cooperation of W. Bakhuis, Car- mabi Foundation. In Cumana J. Perez and associates at the Universidad de Oriente as- sisted us in many ways. The visit to the Fundacion Cientifica Los Roques was made possible by the cooperation of R. Laughlin and B. Alverez. Lynn Cutler, Mountain View, and J. Silverstein, Seattle, provided invaluable assistance in the field and lab. Correct spelling of species' names was kind- ly provided by G. Steyskal, Washington, and K. Bart, Clinton, assisted with the scanning electron microscope. Financial support was provided by the National Science Founda- tion (BSR 86-15315). The cooperation of the following persons and institutions in the loan of reference material and/or providing access to their collections was essential to the completion of this project and greatly appreciated: E. Easton (BMNH); J. Renaud-Mornant (MNHN); G. Hartwich (MNHU); C. Car- pine (MOMV); R. Olerod (NHRS); S. Chambers (RSME); M. Rice (USNM); J. Kirkegaard (UZMK); A. Ivanov (ZIAS); Zoological Institute Tohoku University, Sendai; A. Pierrot-Bults (ZMUA); E. Wil- lassen (ZMUB); M. Dzwillo (ZMUH). Literature Cited Akesson, B. 1958. A study of the nervous system of the Sipunculoideae, with some remarks on the development of the two species Phascolion strombi Montagu and Golfingia minuta Kefer- stein.— Undersokningar over Oresund 38: 1-249. Andrew, M., & N. V. Andrew. 1953. Some annelid and sipunculid worms of the Bimini region. — American Museum Novitates 1617:1-16. Amor, A. 1975. Notas sobre Sipuncula de la Argen- tina, Brasil y Peru.-Physis 34(88): 113-120. Augener, H. 1903. Beitrage zur Kenntnis der Ge- phyreen nach Untersuchung der im Gottinger zoologischen Museum befindlichen Sipunculi- den und Echiuriden.— Archiv fLir Naturgesch- ichte 69:297-371. Baird, W. 1868. Monograph on the species of worms belonging to the subclass Gephyreae.— Proceed- ings of the Zoological Society of London 1868: 77-114. . 1873. Descriptions of some new species of Annelida and Gephyrea in the collections of the British Museum.— Journal of the Linnean So- ciety of London, Zoology 11:94-97. Benham, W. B. 1912. Report on sundry invertebrates from the Kermadec Islands.— Transactions and Proceedings of the New Zealand Institute 44: 135-138. Blainville, H. M. D. de. 1827. Dictionnaire des sci- ences naturelles.— Vers 49:305-313. Bovier, E. L. 1894. Un nouveau cas de commensal- isme: association de vers du genre Aspidosiphon avec des Polypes madreporaires et un Mol- lusque bivalve. — Comptes Rendus Academie des Sciences, Paris 119:96-98. Chamisso, A. de, & C. G. Eysenhardt. 1821. De an- imalibus quisbusdam e classe vermium lin- neana.— Nova Acta Academia Caesarea Leo- poldino 10:343-374. Chapman, G. 1955. Aspects of the fauna and flora of the Azores. III. Gephyrea.— Annals and Mag- azine of Natural History 8(12):35 1-352. Collin, A. 1892. Gephyreen gesammelt von Herrn Stabsarzt Dr. Sander auf der Reise S.M.S. Prinz Adalbert.— Archiv fur Naturgeschichte 58(1): 177-182. Cordero, E. H., & A. Mello-Leitao. 1952. Duas navas especies do genero Aspidosiphon da Ihla Trin- idade.— Memorias do Instituto Oswaldo Cruz 50:277-295. Cuenot, L. 1922. Sipunculiens, Echiuriens, Priapu- liens.— Faune de France, Paris 4:1-30. Cutler, E. B. 1965. Sipunculids of Madagascar.— Ex- trait des cahiers ORSTOM-Oceanographie 3(4): 51-63. . 1969. New species of Sipuncula from the VOLUME 102, NUMBER 4 861 western North Atlantic. — Proceedings of the Bi- ological Society of Washington 82:209-218. -. 1973. Sipunculaofthe western North Atlan- tic— Bulletin of the American Museum of Nat- ural History 152(3): 105-204. -. 1 977a. The bathyal and abyssal Sipuncula.— Galathea Report 14:135-156. -. 1977b. Shallow water sipunculans collected by the Galathea Expedition (Sipuncula).— Steenstrupia 4:151-155. -. 1981. A new species of Aspidosiphon (Sipun- cula) from the western Atlantic Ocean. — Pro- ceedings of the Biological Society of Washington 94(2):445^49. -, & N. J. Cutler. 1979a. Madagascar and In- dian Ocean Sipuncula.— Bulletin du Museum National d'Histoire Naturelle. Paris 4e(l):941- 990. -, & . 1979b. Campagnes de la Calypso au large des cotes Atlantiques Africaines (1956 et 1959) et Sud-Americaines (1960-1961). 23. Sipuncula. — Campagnes de la Calypso XI: 103- 109. -, & . 1980a. Sipincula from southern Brazil.— Boletim do Institut Oceanografico, Sao Paulo 29(1): 1-8. -, & . 1 980b. Deep water Sipuncula from the Gulf of Gascogne.— Journal of the Marine Biological Association of the United Kingdom 60:449-459. -, & . 1980c Sipuncula collected by the R/V"Vema."- Journal of Zoology, London 190: 193-209. -, & . 1981. A reconsideration of Sipun- cula named by I. Ikeda & H. Sato. — Publications of the Seto Marine Biological Laboratory 26: 51-93. -, & . 1985a. A revision of the genera Sipunculus and Xenosiphon (Sipuncula).— Zoo- logical Journal of the Linnean Society 85:219- 246. -, & . 1985b. A revision of the genera Phascolion Theel and Onchnesoma Koren and Danielssen (Sipuncula).— Proceedings of the Bi- ological Society of Washington 98(4):809-850. -, & . 1987a. Deep-water Sipuncula from the eastern Atlantic Ocean. — Sarsia 72:71-89. -, & . 1987b. A revision of the genus Golfingia (Sipuncula: Golfingiidae). — Proceed- ings of the Biological Society of Washington 100(4):735-761. -, & . 1988. A revision of the genus The- miste (Sipuncula).— Proceedings of the Biolog- ical Society of Washington 10 1(4):74 1-766. -. . & P. E. Gibbs. 1983. A reconsider- ation of the Golfingia subgenera Golfingiella and Siphonoides (Sipuncula). — Proceedings of the Biological Society of Washington 96(4):669-694. . , & T. Nishikawa. 1984. The Sipun- cula of Japan: Their systematics and distribu- tion.— Publications of the Seto Marine Biolog- ical Laboratory 29(4/6):249-322. . & V. V. Murina. 1977. On the sipunculan genus Golfingia Lankester, 1885.— Zoological Journal of the Linnean Society 60:173-189. , & P. E. Gibbs. 1985. A phylogenetic analysis of higher taxa in the phylum Sipuncula. — Sys- tematic Zoology 34:162-173. Cutler, N. J., & E. B. Cutler. 1986. A revision of the genus Nephasoma (Sipuncula. Golfingiidae).— Proceedings of the Biological Society of Wash- ington 99(4):547-573. Danielssen. D. C, & J. Koren. 1880. New northern Gephyrea.— Annals and Magazine of Natural History (5)6:462-465. , & . 1881. Fra den norske Norhavsex- pedition (1 876-1 878)-Gephyreer. -Nut Ma- gasin for Naturvidenskapene 26:44-66. DeRochebrune, A. T. 1881. Materiaux pour la faune de l'Archipel du Cap Vert.— Nouveux Archivs Museum National d'Histoire Naturelle, Paris, 2eme ser. 4:215-340. Diesing, K. M. 1851. Systema helminthum.— Braumiller, Vindobonae, 1-558 pp. . 1859. Revision der Rhyngodeen. — Sitzungs- berichte de mathematisch-naturwissenschaft- liche Klasse, Akademie der Wissenschaften in Wien 37:719-782. Ditadi, A. S. 1975. Aspidosiphon schnehageni (Si- puncula) inhabiting Tornatina shells.— The Ve- liger 18(2):200-202. Edmonds, S. J. 1956. Australian Sipunculoidea II. The genera Phascolosoma, Dendrostomum, Golfingia, Aspidosiphon, and Cloeosiphon.— Australian Journal of Marine and Freshwater Research 7(2):28 1-3 15. . 1971. Some sipunculans and echiurans. chief- ly from Guam (Sipuncula and Echiura).— Mi- cronesica 7(1-2): 137-151. . 1974. A new species of Sipuncula {Aspido- siphon exiguus) belonging to the interstitial fau- na of marine beaches collected by Mr. Botosa- neanu during the second Cuban-Romanian Biospeleological Expedition to Cuba 1 9 73 . — In- ternational Journal of Speleology 6:187-192. . 1980. A revision of the systematics of Aus- tralian sipunculans (Sipuncula). — Records of the South Australian Museum (Adelaide) 18(1):1- 74. . 1987. Phyla Sipuncula and Echiura. Pp. 185- 212 in D. M. Devaney and L. E. Eldridge. eds.. Reef and shore fauna of Hawaii, sec. 2 and 3. Bishop Museum Press. Honolulu. 862 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON Fischer, J. 1914. Die Sipunculiden der Nord und Ostsee, unter Beruchsichtigung von Formen des nordatlantischen Gebietes. — Wissenschaftliche Meeresuntersuchungen Abteilung Kiel 16:85- 125. Fischer, W. 1892. Ubersicht der von Herrn Dr. F. Stuhlmann auf Sanzibar und an der gegeniiber- liegenden Festlands-kiiste gesammelten Gephy- reen.— Jahrbuch der Hamburgischen Wissen- schaftlichen Anstalten 9(2): 7 9-8 9. . 1895. Die Gephyreen des Naturhistorischen Museums zu Hamburg.— Abhandlungen aus dem Gebiet der Naturwissenschaften, Hamburg 13:1-24. . 1896. In Semon, Zoologische Forschungs- reisen in Australien und Malayischen Archi- pel. — Medizinisch-naturwissenschaftliche Ge- sellschaft zu Jena, Denkschriften 8:337-339. . 1913. Uber einige Sipunculiden des Natur- historischen Museums zu Hamburg. —Jahrbuch der Hamburgischen Wissenschaftlichen Anstal- ten 30(2):93-101. . 1 9 1 4a. Gephyrea. Beitrage zur Kenntnis der Meeresfauna Westafrikas. —Beitrage zur Kennt- nis der Meeresfauna Westafrikas 1:57-84. . 1914b. Weitere Meitteilungen iiber die Ge- phyreen des Naturhistorischen (Zoologischen) Museums zu Hamburg.— Jahrbuch der Ham- burgischen Wissenschaftlichen Anstalten 31:1- 28. . 1919. Gephyreen der Sud-westkiiste Aus- traliens.— Zoologischer Anzeiger 50:277-285. . 1 920. Gephyreen der antarktischen und sub- antarktischen Meere.— Deutsche Sudpolar Ex- pedition 16, Zoologie 8:409-430. . 1921. Aspidosiphon pygmaeus n. sp., eine neue Gephyree aus Juan Fernandez.— Natural History of Juan Fernandez and Easter Islands, Uppsala 3:45-47. . 1922a. Gephyreen des Reichsmuseums zu Stockholm. -Arkiv fur Zoologi 14(19):l-39. . 1922b. Gephyreen der deutschen Tiefsee-Ex- pedition, auf des Dampfer "Valdivia" 1 898— 1899. — Wissenschaftliche Ergebnisse der deutschen Tiefsee-Expedition, Valdivia 22(1): 1-26. . 1922c. Westindische Gephyreen.— Zoolo- gischer Anzeiger 55:10-18. . 1923. Gephyreen des Golfes von Siam.— Videnskabelige Meddelelser Dansk fra Natur- historisk Forening 76:21-27. . 1925. Echuiriden, Sipunculiden, und Pria- puliden. Pp. 1-55 in G. Grimpe and E. Wagler, eds., Die Tierwelt der Nord-und Ostsee. Aka- demische Verlagsgesellschaft, Leipzig. . 1926. Sipunculiden und Echiuriden der Hamburger Sudsee-Expedition 1908-1909.— Mitteilungen aus dem Zoologischen Staatsinsti- tut und Zoologischen Museum in Hamburg 42: 104-117. . 1931. Sipunculidae. — Mitteilungen aus dem Zoologischen Museum in Hamburg 44: 1 30-140. Gerould, J. H. 1913. The sipunculids of the eastern coast of North America. — Proceedings of the United States National Museum 44:373-437. Gibbs, P. E. 1977. British sipunculans. — Synopses of the British Fauna No. 12. The Linnean Society of London 12:1-35. . 1978. Macrofauna of the intertidal sand flats on low wooded islands, northern Great Barrier Reef.— Philosophical Transactions of the Royal Society of London 284:81-97. , & E. B. Cutler. 1987. A classification of the phylum Sipuncula.— Bulletin of the British Mu- seum Natural History (Zoology) 52(l):43-58. , E. B. Cutler, & N. J. Cutler. 1983. A review of the genus Thysanocardia Fisher (Sipuncu- la). -Zoologica Scripta 12(4):295-304. Graveley, F. H. 1927. The littoral fauna of Krusadai Islands in the Gulf of Manaar. Gephyrea and Phoronis.— Bulletin of the Madras Government Museum 1:87-88. Griibe, E. 1868. Beschreibungen einiger von Georg Ritter von Frauenfeld gesammelter Anneliden und Gephyreen des rothen Meeres.— Zoolo- gisch-Botanische Gesellschaft in Wien. Ver- handlungen der kaiserlich-khoniglichen Zoolo- gisch-Botanischen Gesellschaft in Wien 18:629— 650. Haldar, B. P. 1976. Sipuncula from the Andaman and Nicobar Islands.— Records of the Zoolog- ical Survey of India 70:109. . 1978. Aspidosiphon {Par aspidosiphon) hav- elockensis, a new Sipuncula from the Anda- mans, India.— Bulletin of the Zoological Survey of India 1(1):37-41. Hammerstein, O. 1915. Geophyreen von Madagas- car gesammelt van W. Kaudern 191 1—1912. — Arkiv fur Zoologi 9(10): 1-3. Herubel, M. A. 1904. Liste des Sipunclides et des Echiurides rapportes par M. Ch. Gravier du Golfe de Tadjourah (mer Rouge).— Bulletin du Museum National d'Histoire Naturelle, Paris 10: 562-565. . 1924. Quelques echiurides et sipunclides des Cotes du Maroc et du Mauritane.— Bulletin du Societe des Sciences Naturelles et Physiques du Maroc 4:108-1 12. Horst, R. 1899. Aspidosiphon cylindricus, n. sp.— Notes from the Leyden Museum (Rijks Museem van natuurlijke historie) 20:195-198. Ikeda, I. 1904. The Gephyrea of Japan.— Journal of the College of Science, Imperial University of Tokyo 20(4): 1-87. VOLUME 102, NUMBER 4 863 . 1924. Further notes on the Gephyrea of Ja- pan, with descriptions of some new species from the Marshall, Caroline, and Palau Islands.— Japanese Journal of Zoology 1:23-44. Johnson, P. 1964. Two new species of Aspidosiphon (Sipunculoidea). —Annals and Magazine of Nat- ural History 7(1 3):33 1-335. Keferstein, W. 1867. Untersuchungen iiber einige amerikanische Sipunculiden.— Zeitschrift fur Wissenschaftliche Zoologie 17:44-55. Krohn, A. 1851. Uber die Larve des Sipunculus nu- dus. — Archiv fur Anatomie und Physiologie 1851:368-379. Lanchester, W. F. 1905a. Gephyrea in the marine fauna of Zanzibar and British East Africa from the collections made by C. Crossland in 1901— 1902.— Proceedings of the Zoological Society of London 1:28-35. . 1905b. On the Sipunculids and Echiurids col- lected during the "Skeat" Expedition to the Ma- lay Peninsula.— Proceedings of the Zoological Society of London 1 : 3 5-4 1 . Leroy, P. 1936. Les sipunculiens du Museum d'His- toire Naturelle de Paris.— Bulletin du Museum National d'Histoire Naturelle, Paris, Serie 2 8: 423-426. . 1942. Sipunculiens dTndochine. Notes sta- tion maritime de Cauda.— Notes of the Institute Oceanography Nhatrang 40:1-51. LiGreci, F. 1980. Sul rinvenimento di un Sipuncu- lidae del genere Aspidosiphon, Diesing 1 85 1, sui fondali batiali della Sicilia occidentale. — Riv- ista di Parassitologia 4 1(1): 123-1 34. Longhurst, A. R. 1958. A ecological survey of the West African marine benthos.— Colonial Office ' of Fisheries Publications London 11:1-102. Lopez, R. J. S., J. I. Saiz-Salinas, & M. C. Diaz-Vilela. 1984. Sobre la fauna de sipunculidos (Sipun- cula) del archipelago canario. Pp. 189-196 in V. Monteiro-Marques, ed., Actas do 4th Simpo- sio Iberico de Estudos do Benthos Marinho, As- sociacao des Estudantes da Faculdade Cien-Cias des Lisboa. Migotto, A. E., & A. S. F. Ditadi. 1988. The Aspi- dosiphonidae (Sipuncula) from the northern and northeastern coasts of Brazil. — Revista Brasi- leira de Biologia 48(2):245-263. Monro, C. A. 1931. Polychaeta, Oligochaeta, Echiu- roidea and Sipunculoidea. — Scientific Report of the Barrier Reef Expedition 4:1037. Miiller, J. 1844. Uber einem neuen Wurm Sipunculus (Phascolosoma) scutatus.— Archiv fur Natur- geschichte 10:166-168. Murina, V. V. 1 964a. Report on the sipunculid worms from the coast of South Chinese Sea.— Trudy Instituta Okeanologii Academii Nauk SSSR 69: 254-270. . 1964b. Sipunculid fauna of the Mediterra- nean Sea.— Trudy Sevastapol Biologicheskoi Statsii 17:51-76. . 1967a. New facts about the distribution of marine sipunculid worms in the Red Sea and the Gulf of Aden. Pp. 52-56 in Some results of Expedition III to the Red Sea Inbuma. —Express Information, Academy of Sciences Ukrainian Soviet Socialist Republic. . 1967b. On the sipunculid fauna of the littoral of Cuba.— Zoologicheskii Zhurnal 46:35-47. . 1 967c. Report on the sipunculid worms from the sublittoral zone of Cuba and the Gulf of Mexico.— Zoologicheskii Zhurnal 54(9): 1329- 1339. . 1968. New species of Sipuncula from the Red Sea.— Zoologicheskii Zhurnal 47:1722-1725. . 1970. Contributions to the fauna of sipun- culid worms from the Aden Gulf.— Vestnik Zoologii 2:65-71. . 1971. On the species composition and ecol- ogy of sipunculids of the Red Sea. Pp. 76-88 in A. O. Kovalevski, ed., Shelf benthos of the Red Sea. Academii Nauk SSSR, Moscow. . 1972. Contribution of the sipunculid fauna of the Southern Hemisphere.— Zoologii Insti- tuta Akademii Nauk SSSR 1 1(19):294-314. . 1974. New subspecies of sipunculid worms.— Zoologicheskii Zhurnal 53:1713-1717. . 1978. On the sipunculid and priapulid fau- na of the southern Pacific— Trudy Instituta Okeanologii Academii Nauk SSSR 1 1 3: 1 20-1 3 1 . . 1981. First report of the sea worms Sipuncula and Echiura from littoral Somali.— Vestnik Zoologii 1:11-18. Ocharan, F. J. 1980. Primeros datos sobre los sipun- ulides del litoral N.Y. N.W. de Espana (Santan- der, Asturias y Galicia).— Boletin de Ciencias de la Naturaleza I 26:1 1 1-1 19. Quatrefages, A. 1865. Gephyrea inermia. — Histoire naturelle des Anneles marins et d'eau douce, Paris 2:599-632. Rice, M. E. 1970. Asexual reproduction in a sipun- culan worm.— Science 167:1618-1620. . 1975. Survey of the Sipuncula of the coral and beach rock communities of the Caribbean Sea. Pp. 35-49 in M. E. Rice & M. Todorovic, eds., Proceedings of the International Sympo- sium of Sipuncula and Echiura, Naucno Delo Press, Belgrade. . 1976. Larval development and metamor- phosis in Sipuncula.— American Zoologist 16: 563-571. , & I. G. Macintyre. 1972. A preliminary study of sipunculan burrows in rock thin sections.— Caribbean Journal of Science 12(1— 2):41— 44. , & . 1979. Distribution of Sipuncula 864 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON in the coral reef community, Carrie Bow Cay, Belize.— Smithsonian Contributions to Marine Sciences 12:311-320. , & A. C. Stephen. 1970. The type specimens of Sipuncula and Echiura described by J. E. Gray and W. Baird in the collections of the British Museum of Natural History.— Bulletin of the British Museum of Natural History 20(2):49- 72. Saiz Salinas, J. I. 1984. Gusanos sipunculidos (Si- puncula) de variaslocalidades de la costa de Por- tugal. Pp. 177-188 in V. Monteiro-Marques, ed., Actas do 4th Simposio Iberico de Estudos do Benthos Marinho, Associacao des Estudantes da Faculdade Cien-Cias des Lisboa. . 1986a. Los gusanos sipunculidos (Sipuncula) de los fondos litorales y circalitorales de las cos- tas de la peninsula Iberica, Islas Baleares, Cana- rias y mares adyacentes. — Monografias Instituto Espanol de Oceanografia, Madrid 1:1-84. . 1986b. The overlooked species of sipuncu- lans commensal of solitary corals (Sipuncula).— Bulletin du Museum National d'Histoire Na- turelle, Paris 4e ser 8, sec. A 3:551-565. Sato, H. 1935. Sipunculoidea and Echiuroidea of the West Caroline Islands.— Tohoku Daigaku, Sen- dai, Japan, Science Reports series 4, 10:299- 329. . 1939. Studies on the Echiuridea, sipunculids, and priapulids of Japan.— Science Reports of Tohoku Imperial University 14(4): 3 3 9-45 9. Schmidt, O. 1865. Uber den Bau un die systema- tische Stellung von Aspidosiphon muelleri Dies- ing (Lesinia farcimen Schmidt).— Mitteilungen des Naturwissenschaftlichen Vereines fur Steiermark 3:56-66. Selenka, E. 1885. Report on the Gephyrea collected by H.M.S. Challenger during 1873-76. -Report of scientific results of the Voyage of Challenger Zoology 13(36): 1-25. , J. G. de Man, & C. Biilow. 1883. Die Sipun- culiden, eine systematische Monographic — Semper Reisen in Archipel Phillippinen 2 4: 1- 131. Shipley, A. E. 1898. Report on the Gephyrean worms collected by Mr. J. Stanley Gardiner at Rotuma and Funafuti.— Journal of Zoology 1898:468- 473. . 1899a. On Gephyrean worms from Christ- mas Island.— Proceedings of the Zoological So- ciety of London 22:54-57. . 1899b. A report on the Sipunculoidea col- lected by Dr. Willey at the Loyalty Islands and in New Britain. Pp. 151-160 in A. Willey, ed., Zoological results based on material from New Britain, New Guinea, Loyalty Islands and else- where collected 1985-1897 by A. Willey. Uni- versity Press, Cambridge. . 1902. Sipunculoidea, with an account of a new genus Lithacrosiphon. Pp. 131-140 in J. S. Gardiner, ed., Fauna and geography of the Mal- dive and Laccadive Archipelagoes. University Press, Cambridge. . 1903. Report on the Gephyrea collected by Prof. Herdman at Ceylon in 1 902. —Ceylon Pearl Oyster Fisheries Supplementary Reports 3:169— 176. Sluiter, C. P. 1881. Beitrage zu der Kenntnis der Gephyreen aus dem Malayischen Archipel.— Natuurkundig Tijdschrift voor Nederlandisch- Indie 41:84-108. . 1884. Beitrage der Gephyreen aus dem Ma- layischen Archipel.— Natuurkunding Tijdsch- rift voor Nederlandisch-Indie 43:26-88. . 1886. Beitrage zu der Kenntnis der Gephy- reen aus dem Malayischen Archipel.— Natu- urkundig Tijdschrift voor Nederlandisch-Indie 45:472-517. . 1891. Die Evertebraten aus der Sammlung des Koniglichen naturwissenschaftlichen Ver- eins in Nederlandisch-Indien in Bata via.— Na- tuurkunding Tijdschrift voor Nederlandisch-In- die 50:102-125. . 1898. Gephyreen von Sud Africa.— Zoolo- gische Jahrbucher Abteilung fur Systematik 1 1 : 442-450. . 1900. Gephyriens (Sipunculides et Echiurides) provenant des campagnes de l'Hirondelle et de la Princesse Alice, 1886-1897.— Resultats des Campagnes Scientifique accomplies sur son yacht sur Albert I, Prince Souverain de Monaco 15: 1-29. . 1902. Die Sipunculiden und Echiuriden der "Siboga" Expedition, nebst Zusammenstellung der Uberdies aus den indischen Archipel be- kannten Arten. — Siboga-Expeditie, Monogra- phic 25:1-53. . 1912. Gephyriens (Sipunculides et Echiu- rides) provenant des campagnes de la Princesse Alice, 1898-1910.— Resultats des Campagnes Scientifique accomplies sur son yacht sur Albert I, Prince Souverain de Monaco 36:1-36. Southern, R. 1913. Gephyrea of the coasts of Ire- land.— Scientific Investigations of the Fisheries Branch, Department of Agriculture for Ireland 3:1-46. Stephen, A. C. 1934. The Echiuridae, Sipunculidae, and Priapulidae of Scottish and adjacent waters.— Proceedings of the Royal Society of Edinburgh 22:159-186. . 1941. The Echiuridae, Sipunculidae, and Priapulidae collected by the ships of the Dis- covery Committee during the years 1926- 1937. -Discovery Reports 21:235-260. . 1942. The South African intertidal zone and VOLUME 102, NUMBER 4 865 its relation to ocean currents. Notes on the in- tertidal sipunculids of Cape Province and Na- tal.— Annals of the Natal Museum 10(2):245- 256. . 1958. The sipunculids of Haifa Bay and neighborhood.— Bulletin of the Research Coun- cil of Israel 7b(3^1): 129-1 36. . 1960a. Echiuroidea and Sipunculoidea from Senegal, West Africa.— Bulletin de lTnstitut Francais d'Afrique Noire 22a(2):5 12-520. . 1960b. British echiurids (Echiuroidea), si- punculids (Sipunculoidea), and priapulids (Pri- apuloidea). Synopses of the British Fauna No. 12. The Linnean Society London, 27 pp. . 1964. A revision of the classification of the phylum Sipuncula.— Annals and Magazine of Natural History 7(13):457-462. , & S. J. Edmonds. 1 972. The phyla Sipuncula and Echiura. Trustees British Mus. (Nat. Hist.), London, 528 pp. , & J. D. Robertson. 1952. A preliminary re- port on the Echiuridae and Sipunculidae of Zan- zibar.— Proceedings of the Royal Society of Edinburgh 64(4):426-444. Steuer, A. 1936. Sipunculidea, Phoronidea, Brachi- opoda, Enteropneusta und Acrania von Alex- andrien in Agypten.— Note dell'Instituto Bio- logia marina di Rovigno 23:1-18. . 1939. The fishery grounds near Alexandria. XVIII. Sipunculoidea, Phoronidea, Brachiop- oda, Enteropneusta and Acrania.— Fouad I. In- stitute of Hydrobiology and Fisheries Notes and Memoirs 30:1-3. ten Broeke, A. 1925. Westindische Sipunculiden und Echiuriden.— Bijdragen tot de Dierkunde 24:8 1- 96. Theel, H. 1875. Etudes sur les gephyriens inermes des mers de la Scandinavie, du Spitzberg et du Groenland.— Bihang till Kungliga Svenska Ve- tenskaps-Akademiens Handlingar 3(6): 1-30. . 1905. Northern and Arctic invertebrates in the collection of the Swedish State Museum I. Sipunculids.— Kungliga Svenska Vetenskaps- Akademiens Handlingar 39(1): 1-1 30. Vaillant, L. 1871. Note sur quelques Gephyriens de Saint- Vincent du Cap Vert. Pp. 268-276 in Les fonds de la mer. Etude internationale sur les particularites nouvelles des regions sous-ma- rines (Vol. 1), L. De Folin & L. Perier (dirs.), Paris, Savy. . 1875. (Planche IV, Gephyriens). In Les fonds de la mer. Etude internationale sur les particu- larites nouvelles des regions sous-marines (Vol. 2), L. De Folin & L. Perier (dirs.), Paris, Savy, pi. IV. Voss-Foucart, M. F., S. Barzin, C. Jeuniaux, & J. C. Bussers. 1977. Etude comparee de la com- position chimique des regions souples et durcies de la cuticule de quatre especes de sipuncu- liens.— Cahiers de Biologie Marine 1 8: 1 35-1 45. Wesenberg-Lund, E. 1937. Gephyrea. — Bulletin Mu- see royal d'Histoire naturelle de Belgique 13(36): 1-23. . 1954. Priapuloidea, Sipunculoidea and Echiuroidea. — Bulletin Institut royal des Sci- ences naturelles de Belgique 30(1 6): 1-1 8. . 1955. Reports of Lund University Chile Ex- pedition, 1948-49. 19. Gephyrea from Chile.— Lunds Universitets Arsskrift 5 1(10): 1-23. . 1957a. Sipunculoidea and Echiuroidea from West Africa together with a bibliography on Ge- phyrea after 1920.— Bulletin Institut royal des Sciences naturelles de Belgique 33(42): 1-24. . 1957b. Sipunculoidea from the coast of Is- rael.— Bulletin of the Research Council of Israel 6b(3-4): 193-200. . 1957c. Sipunculoidea and Echiuroidea from the Red Sea.— Contributions to the knowledge of the Red Sea No. 3, The Sea Fisheries Re- search Station, Haifa Bulletin 14:1-15. . 1959a. Sipunculoidea and Echiuroidea from tropical West Africa.— Atlantide Report 5:177— 210. . 1959b. Sipunculoidea and Echiuroidea.— Resultats Scientifiques des Campagnes de la "Calypso" 4:207-217. . 1959c. Sipunculoidea and Echiuroidea from Mauritius. — Videnskabelige Meddelelser Danske fra Naturhistorisk Forening i Kjobenhavn 121: 53-73. . 1963. South African sipunculids and echiu- roids from coastal waters. — Videnskabelige Meddelelser Danske fra Naturhistorisk Foren- ing i Kjobenhavn 125:101-146. Whitelegge, T. 1899. The Hydrozoa, Scyphozoa, Ac- tinozoa and Vermes of Funafuti.— Memoirs of the Australian Museum, Vermes 3:371-394. Zavodnik, D., & V. G. Murina. 1975. Contribution to Sipuncula of North Adriatic insular region.— Rapports et proces-verbaux des reunions com- mission internationale pour l'Exploration Scientifique de la Mer Mediterranee 23(2): 127- 128. , & . 1976. Sipuncula of the region of Rovinj (North Adriatic Sea). — Biosistematika 2(l):79-89. (EBC) Biology Department, Utica College of Syracuse University, Utica, New York 13502; (NJC) Biology Department, Ham- ilton College, Clinton, New York 13323.