Pibrary of the Museum OF COMPARATIVE ZOOLOGY, AT HARVARD COLLEGE, CAMBRIDGE, MASS. Founded hy private subscription, in 1861. Deposited by ALEX. AGASSIZ. THE CHONIOSTOMATIDA: THE CHONTOSTOMATIDAL AY FAMILY OF COPEPODA, PARASITES ON CRUSTACEA MALACOSTRACA BY De. HJ. HANSEN WITH THIRTEEN COPPER PLATES AT THE EXPENSE OF THE CARLSBERG FUND We want facts, not inferences, observations, not theories, for a long time to come. »Natural Science«, 1896. COPENHAGEN ANDR. FRED. HOST & SON, PUBLISHERS “1897 VAR Gig Vi O10 GK S109. UL : ANCA af WAD 1697, jee oy es PREFACE, his work contains an account of forty-three species of Copepoda, all parasitic on mala- a. Crustacea, and all belonging to the same family. When in 1890 I began my study of this group, there were published descriptions of only three species, and mention had been made of a fourth. Two more have been since described and a seventh named, but not described; so that until now (July 1897) only five species have been really made known. In the present work I increase this number about nine times, and yet, most likely, my discoveries only extend to one fifth or one sixth or perhaps a much smaller part of the species extant. I have been brought to this conclusion by the consideration that no less than thirty-three of my species have been found exclusively on Crustacea in the Zoological Museum of the Copenhagen University. What multitudes of these animals are likely to be discovered, when some day the large foreign- museums acquire rich collections of non-decapod Malacostraca, and when this material is submitted to a thorough research! On the whole, my studies of late years have given me the impression that of nearly all the Crustacea living on the bottom of the sea — the Decapods excepted — we only know from about half down to a very small percentage of existing species. Especially to the parasitic forms does this apply, and I think one of the most important results of the present work is to show the wealth of a group, which hitherto has occupied only a very diminutive place. It may be added that, in the course of the last two years, I have found on the material brought home from the sea near Iceland and Greenland by the »Ingolf« expedition several new forms which cannot be included in the present treatise, but which will be subjected to future examination. A chance led me to this study. In dissecting a female of Idothea marina (L.) I discovered in its marsupium an unknown parasite belonging to the Epicaridea, and further researches led to the discovery of a number of specimens of this species and of a form nearly akin to it on Edotia nodulosa (Kr.). Both parasites were afterwards described from my material by Giard and Bonnier (the genus Clypeoniscus G. and B.). Those authors had just previously described an Epicarid living as a parasite on Ampelisca diadema Costa. What I had found on Idotheidse tempted me to go on looking for Epicaridea, so I examined our Ampeliscidz and found — not these forms, — but several species of Choniostomatide as well as another most remarkable parasite, which I described in 1892 under the name of Rhizorhina Ampelisce H.J.H. Professor Sars has told me (1886) that he had found some species of Spheronella on Amphipoda. Now, as my own discoveries had called forth my interest, I began in the Copenhagen Museum an examination of the material of Amphipoda and later on of the other orders of Malacostraca. Professor G. O. Sars lent me all his material of this family for my researches, and he further provided me with newly discovered forms — seven in all, — of which four are particularly interesting; two of the most remarkable genera, the parasites on Myside, are owing entirely to him — for all of which I have great pleasure in offermg the eminent naturalist my best thanks. — The Rev. Canon A. M. Norman, F. R. 8., lent me the types of Aspidoecia Normani Giard and Bonnier, and the Rev. Th. R. R. Stebbing, F. R. S., determined for me some Amphipoda from the Mediterranean, the West-Indies, the Cape and Hong-Kong, for which I beg these gentlemen to accept my thanks. Last, not least, I wish to express my warm gratitude to the managing Committee of the Carlsberg Fund for having allowed me a considerable sum to defray the expenses of the present work. The English translation from the Danish manuscript is the work of Miss Louise von Cossel. CONTENTS. MPeMEROCUCTORYLREMARKS << )o:o-crcrarorev keratin ste eine ote Spel eeIe ey ASrsVID CCCIRSIENS ave, Ae eave TBtate a reveal ayaa iebel ss @eoucatere sain ote eis Generale Historical aView s.<- © ae & i. Geographical and Bathymetrical Distribution of the Family ...................-.0.e0eeeeee j. Geographical Distribution of the particular Species relatively to that of their Hosts ......... k. Frequency of the Parasites in Proportion to their Fertility...................0.00ceeeeeee FA DOULA LOSESIUCALEOM eter Seer Pie OT MTT ACY ree Thayer east stche, eee ac aT aINVSy ta Teele ose SE dee imitation gands Ghareclers molt heas pecles meer anne aeeiil< sere iacisirckseiatereieisicts iersleke ae eae baimitationy andy Characters: of thes Generage eye ievisercic-teciae eles stein cieieisiele sete tere niet ieteteiete chara ctersuo tthe wh alnilyra dm crecestaisce cine ertisitice cae rel iste ves so er esisioice Slee © Slee See meee noe dee lacerorathesMamilyveins thes oySsberuectcrsye tyes eters craiceteerastevora a2 yrte ayatevoreretayas a ayeye lela loreue ener eles Rio, o. 6 09 | Ss ® = 5 = =] oO =] ix] oO ° Lear} es S oO Las} i) tl p n a. i9") n ° 5 = a 3 i= an 5) n a i?7] IWemDesoriptiontof (Genera and: SPOCiesia..c .jeicraca.cie crecave ors cteravarars, arsyo x, sxars ales raynia Sisters s cleeeioters cee ae eee ie SLCHOLAOCHENES ang COWS. aon Nav Tae coteie Ne cieibelcle tale aie cine eicis ciea ee a stele a a er EE eee UST OTAOCOSCELES ITS ETI ect Micyepata te ois. Sy ciao tates ee sv 0 aa STaTaISTa ey Sava] wy er au arapvere disse) senate alee ster ae II. ISD er OMELL ES SAL EMSK- Vii iar) oars at cvapete ois eu ctorei erie teteia elie oie iahd Se RISES aT ISe Ewe oA Oe ast LALASILESH OLMEA TN DHT OCA opera) 5.26 ce shea soye aie cvs ayale. 515) 2154) hate eVeyas Sha pie vayelevevere ova naveyeytee oi syeteler ae Hye ACASICS ROLE CUMACE Atego rarer 2 wh .t er Palas jh craters Meare ose atheros Miner te tela biw Heeroma ee eae CoE ATASILES BON ISO PO Ualears oy chacay =e cromreictaiedtetessiefesefectae obaeas old. ocossi vince wi uate ater OC aret a LaTaS aha Oe oe ee LVS NC Romsostonn ce UM Mew ele ya acto osha 2is\o/6s 4 osjave xe vse, save sy sleve) wvavele: eve) aye\ ace, ave. ai erate (ico w dan cverslele Alsfavale ves ern Vee My stdion Tin B etleca astern nate i rn aseinrs si ctses als! 'a)0/ ators. clatete sueievS si aieid ae atauelecen tae nate. Masri Cee Wate) Aspidoccia | GlardwandeBonmierpryetetsctstiarsiasi einer < oeteras: afhasdvtra sia obvaccrsi aw cae east deen Sepoe eee Explanations Of they Plates icrsryrscave seis areata e aves veces se hares yensteccraelaybi sino tavese. erage svar tagale eaten ie ace ear aes Errata ‘ a! 7 7 1 al ‘ = a dl los j . » Ve > 7 _ ~~ i ons =—5 ’ & 7 - oy 7 _ . ae - i, 7 . >: _ 7 ; = P 7 _ : a eg a (oUa o% — - oe — - mS o> ak 2 I. INTRODUCTORY REMARKS, he majority of the species here described I found in examining systematically for this purpose the collections of the Zoological Museun: in Copenhagen. Throughout a number of years the two directors of the entomological department, the late Professor J.C. Schiedte and his successor, Inspector Dr. F. Meinert, have taken care not only to acquire as many species as possible, but — of the smaller forms — also as many individuals as could be procured, so that of a good many northern Amphipoda, and of a great number of Danish Amphipoda, Cumacea, etc., the museum possesses hundreds of specimens. This has been of the greatest use to me in my researches, for while a few of the parasites — at least of those found on our own material — are met with rather frequently, the greater number are very rare, and a considerable part so scarce, that only one or two specimens are found on each hundred of the animals examined. As a matter of course, I have examined numerous species without finding a single parasite. Of the following forty-three species only one lives on the outside of its host (Mysidz vere), two occur in the branchial cavity of Cumacea, two in the branchial cavity of Hippolyte ; all the remaining species are only found in the marsupium of the female of Amphipoda Gammaridea, Isopoda, Cumacea and Myside vere (or sometimes in young individuals of Amphipoda on the ventral side of the thorax between the gills). Im the Isopoda, the Myside, and sometimes in the Amphipoda, parasites can be seen by looking through the plates of the marsupium. In most Amphipoda and in Cumacea the marsupium has to be submitted to a closer examination; some of the plates have to be lifted up and examined through a lens; in the small forms even the adult parasites can only be discovered by help of a simple microscope. Where a closer search of an infested marsupium is required, it is usually necessary to place the host in a hollow ground glass-plate under water, and to examine it very carefully twenty or thirty times magnified under a simple microscope, in order to be able to discover the male animals which are generally '/i:—*/s mm. in length, as well as the free larve and the pupe, and to find out the way in which these minute animals are hinged. 2 Parasites are not at all easy to deal with; when taken out, everything — except perhaps tolerably large females and ovisacs — must be kept in glycerine on an object- glass, for if males, pup, etc. are put in spirit, they are generally difficult to find and to get out of the tubes. For this use the glycerine must always be strongly diluted with water, otherwise the animals shrink very much, and the females especially are very apt to lose their shape. The water is made to evaporate by standing exposed to the air. Neither the females nor the males nor the animals in any stage of development, can bear the pressure of a glass-cover, or of part of it, without losing their natural shape. In order to make drawings of the entire animals or of parts of the females, the following method was employed: I took very small covers (frequently a middle-sized cover was cut into four parts), and placed a very thin wooden wedge under the middle of the back edge of the cover, so that by very carefully pulling the wedge a little, I made the glass touch the animal, or part of it, just sufficiently to keep it in a certain position; by means of a hair, which was introduced through the opening, its attitude could easily be changed. Im this way I was able after some practice to manipulate a male of the length of '/smm., so that I could make accurate illustrations of one specimen seen from below and sideways without damaging it at all. After use the animal and all the parts that had been examined were placed under a large elass-cover in the way described above, and the opening was closed with varnish. The female was always dissected, in order to submit the head and the genital area to a careful examination. The latter part was treated in the following way: with a sharp and very small knife I cut through the animal a little above the genital region, after which this part was placed under a simple microscope which magnified it a hundredfold; the inside of it was cleaned with a knife, so as to leave only the muscles of the genital apertures and one or both of the receptacula seminis. I specify this proceeding, which I learned by degrees through rather troublesome experiences, partly that the reader may judge of the accuracy of my illustrations, partly to enable future students, who may not possess such ample material, to conquer the difficulties with comparative ease. As far as possible, I have everywhere given figures of an adult female, a male and an ovisac (sometimes adding one or two pupz) magnified to the same scale, in order to show the relative size of the two sexes, the ovisacs and eggs. The size of the male compared with the female and the ova varies very much in the different species. For the convenience of the student, and in accordance with earlier statements (1890), I have always figured the males vertically from the ventral side and laterally from the left-hand side. While in symmetrically shaped Arthropoda, in dorsal or ventral view, I generally arrange the position of the legs and figure those on one side to correspond with those on the other, with the parasites described in the present work I have not ventured to do this.’ The animals were often a little crooked on account of a slight pressure, to which they had been exposed in the marsupium; maxillipeds and legs were found straddling in different directions, and as a rule were too small to allow of much alteration in their attitude, without great risk of 3 damage ensuing. I only ventured slight attempts at construction, not being able to calculate _ how the details — e. g. legs of males and larve — would appear, if drawn in a position differing from the one in which they were found. As a rule I: have figured the animals with all the irregularities they presented, and the limbs in the position they happened to occupy at the time of drawing. Where I had several specimens at my disposal, of course I chose the one which was most suitable for illustrat on. I must briefly mention one point in my nomenclature. In 1893 I stated (in » Zool. Anzeiger«) that the two pairs of limbs which had been formerly named the first and second pairs of maxillipeds, ought to be regarded as the second pair of maxille and a pair of maxillipeds. Shortly afterwards Dr. W. Giesbrecht gave very detailed proofs of the same fact (Mitth. Zool. Stat. Neapel, 11. B.). I also proposed to introduce the names »maxillulee« and »maxille« (in analogy with the commonly used names »antennule« and »antenne«) for the two pairs of jaws, and I shall here avail myself of these short,’ convenient and very intelligible names. In conclusion a few remarks may be offered about the plan of the present work. For several reasons I have contented myself with representing the external structure of the adult animals and their post-embryonic development, and I have spent an exceedingly long time, partly in finding females and eggs, males, larve and pupx, partly in studying the material I had discovered. The result is that at present scarcely any moderately large family of genuinely parasitic Copepoda is so well known as the Choniostomatide. I have found the males of thirty-two of the forty-three species, the larve of twenty-three, the pup or other stages of the post-larval development of a pretty considerable number of species. At the same time I must call attention to the great and numerous gaps in the knowledge of the metamorphosis of these animals, which vary remarkably according to the different species. On their embryology I do not enter at all, and their anatomy is almost totally omitted; I could not have given information of any value unless I had stayed long enough at the seaside to enable me to collect a large supply of living animals of several species, but this would have considerably delayed and increased the work, which is rather voluminous as it is; so, not being able to present an exhaustive study of these topics, I have — contrary to the habit of numerous authors — only treated what was indispensable to classification (the genital region and receptacula seminis). Besides, I should advise students not to enter upon the anatomy of forms so small, difficult and for the most part rare, before having acquired a thorough autoptical knowledge of representatives of various other families among parasitic Copepoda. 1" Il. GENERAL HISTORICAL VIEW. nfortunately I am obliged to go much into detail in this chapter, not only in order to U give a summary of our previous knowledge and its defects, but also and particularly in order to throw light on a number of very objectionable postulates, reflections and theories put forward by Mssrs. A. Giard and J. Bonnier in their two (four) papers. Very short contri- butions (by G.O. Sars and J. Sparre-Schneider) are mentioned in the special part. W.Satensky: Spheronella Leuckarti, ein neuer Schmarotzerkrebs (Archiv fiir Natur geschichte, 34ter Jahrgang, 1868, p. 301—322. Taf. X). The author has given a very extensive account of this new genus and species, the first form which was discovered of this family. He has found females, males, eggs, larve and pup, in fact all stages, and on the whole his descriptions are good, but unfortunately the illustrations are rather rude, which is indeed a pity, as the species happens to belong to the most difficult group of the large genus. I do not think it necessary to point out some slight differences between the author’s account and my own, e. g. his incorrect statement of the number of joints in the antennule of the larve etc., but it must be mentioned that he has overlooked the rudimentary antenne (2nd pair) in the male and the female, that his very detailed description of the rostrum is not correct, as he has taken the hairs outside the membranous border of the mouth for »Radiarfalten« in the membrane itself (p. 303), and that his long description of the more solid chitine lists of the rostrum is too diagrammatic. This is connected with his quite wrong idea on the maxillule, about which he writes: »Es sind nimlich zwei solcher Kiefern vorhanden, welche eingliedrig sind und an ihrem Ende eine Borste tragen« (comp. my description below). On the other hand it must be acknowledged that he has found and described correctly the legs and the caudal stylets of the female, but in the male he mis- interprets the stylets, taking them for a third pair of legs; he has found spermatophores ete. Furthermore, his representation of the genital area is defective, and he has overlooked receptacula seminis, but he is right in stating that the female has no anus. He also gives a somewhat detailed account of the embryology of these parasites, making out their stages of development till they appear as full-grown larve, but this part of the development I have 5 scarcely studied at all. Finally he describes three stages of the pup, mentioning their want of internal structure during the first stage and their considerable growth, but he has failed to understand their mouth, nor does he mention the possibility of a very different development of the two sexes. He concludes with some reflections on the place which the new form ought to occupy in the system, thinking — with good reason — that it »in keine der bis jetzt aufgestellten Familien vollkommen hineinpasst« (p. 320), but that it is nearest akin to the Lerneide on account of similarity in the structure of the mouth, an opinion which I cannot share (s. below). Salensky took his species at Naples on an Amphipod which was many years after determined by Della Valle as Microdeutopus gryllotalpa Costa. About its occurrence on males as well as on females he has a statement (p. 302) which will be mentioned later on in the part headed »Habitation, biology and distribution. Max Weser: Die Isopoden gesammelt wihrend der Fahrten des Willem Barents in das nordliche Eismeer in den Jahren 1880 und 1881 (Bijdr. tot de Dierkunde, 1884). The author informs us (p. 35) that in a vesicular swelling on the carapace of a specimen of Hippolyte Gaimardii M. Edw. he found four globular bodies which contained either eggs or larve, and he thought they were »Bopyriden-Larven im ersten Larven-Stadium« and that the eggs »werden wohl schubweise abgesetzt vom Weibchen und von einer gemeinsamen Hiille umgeben«. His suggestion of Bopyrid-larve is a great mistake; what he found were the ovisacs of a Choniostoma. The statement is only of interest in so far as it indicates a locality of the genus; the fact that this otherwise excellent author happens to be the first who found such ovisacs appears more than valueless to me, considering how he explains the matter, and I only mention it here, because it relates to my remarks in the criticism of Giard and Bonnier. H. J. Hansen: Oversigt over de paa Dijmphna- Togtet indsamlede Krebsdyr (Dijmphna- Togtets zool.-bot. Udbytte, 1887). In this paper (p. 271—278, Tab. XXIV, fig. 7—7h), I gave a detailed description of the female, of ova and larve of a species found on Hippolyte Gaimardii M. Edw. and Hipp. polaris (Sab.) in the Kara Sea, and I gave it the name of Choniostoma mirabile. Furthermore, on this torm I established a new family, Choniostomatide; I did not know Salensky’s paper at the time, but when Prof. G. O. Sars had called my attention to it, I mentioned it in the French résumé worked out later on (p. 511); however, I maintained my new genus. In the female I found antennule, antenne and a mouth with supposed mandibule, the anterior branch of the maxillule and the maxille. The description of the mouth is not quite correct, as I did not mention the membranous mouth-border, but I found the hairs which I thought proceeded from the margin of the mouth; I also over- looked the rudimentary maxillipeds, nor did I find the genital apertures. The description of the larva is pretty correct on the whole, but I have with some hesitation mentioned four joints instead of three in the antennule, nor have I understood its olfactory seta as such. In 1889 Giard and Bonnier were of the opinion that the specimen found by me on Hippolyte polaris belonged to another ‘species which they called Choniostoma Hanseni; this opinion 6 was based on the fact that it was much larger and lived on another species. The animal did in fact prove to differ from Choniostoma mirabile; however, the two reasons alleged by the authors proved to be wrong, for a female with eleven ovisacs found on Hippolyte Gai- mardii and proving to be identical with the species on Hipp. polaris, was even somewhat smaller than the largest Choniostoma mirabile. Consequently Chon. Hansenii is found on two species of Hippolyte, whereas Chon. mirabile has as yet only been noticed on one. A. Giarp and J. Bonnier: Sur un Epicaride parasite d’un Amphipode et sur un Copépode parasite d’un Epicaride (Comptes-rendus de l’Acad. des Sciences, 29 avril 1889). This preliminary note is only mentioned here for the sake of completeness, as its contents are largely worked out in the following publication. A. Garp and J. Bonnier: Note sur UV Aspidoecia Normani et sur la famille des Choniostomatide (Bull. scientifique de la France et de la Belgique, T. XX. 1889, p. 341—72, Pl. X—XI).. In this paper the authors have partly described and figured the Aspidoecia Normani, the new species and genus established in their preliminary note, partly given a very detailed critique of all that has been written on the subject. Each of these parts de- serves a special mention. Of their new species the authors have examined a female with five ovisacs and two males attached to it, sitting on the back of the carapace of Hrythrops microphthalma G.O. Sars (belonging to Myside vers) under an obliquely placed EKpicarid, Aspidophryxus Sarsi Giard and Bonnier. Accidental circumstances led them to adopt the following conclusion as the most plausible: »qu il existe un rapport soit de parasitisme soit de mutualisme« (p. 353) between Aspidoecia and Aspidophryxus (which is a mistake; 9: below); they say that the female Copepod »était reliée a /Aspidophryxus par un appareil fixateur« (p. 344), though such an object does not exist, and they declare that it »adhérait certainement a la Mysis par une ventouse« (p. 344), which is not the case either, as it is attached by what later on I shall call »the adhesive plate«, a congealed substance forming a plate-like cover on the forehead in front of the mouth, and which is secreted by the »glandes cémen- taires« mentioned by the authors (p. 349). In their description of the female (p. 347—50) they mention »les deux points chitineux« (entrances to the receptacula seminis), and they give a correct description of the genital apertures, except that the small opening which they call »pore de fécondation«, and of which they say that it serves »évidemment a l’entrée des spermatozoides«, does not serve this purpose at all. They have found »la ventouse« on the head, but they cannot make out whether the mouth is situated at the bottom of it (which it does), or whether it is found »& la partie supérieure de la ventouse, celle-ci servant uniquement a la fixation du parasite«. Finally, they have overlooked the antennule, the maxillule and the maxille. However, it must be borne in mind that having had only one individual which they were not allowed to dissect, it would be unfair to expect them to be able to study the organs of this small and extremely difficult animal much better than they have done. With regard to the male the case is different; it is much easier to examine, besides they had two specimens. After having studied my own material of the same species, 7 T came to results which differed very much from the figures and descriptions of the authors. Though feeling convinced that I had studied animals belonging to their species, I wanted to make quite sure of it and asked the Rev. Canon A. M. Norman to lend me the animals which had served as types to the French authors, and I received a male and a female. The male was kept in a preparation made by Mssrs. Giard and Bonnier, but it was considerably flattened in an oblique direction, these animals —as stated above — not being able to with- stand the pressure of a glass-cover; its position was about the same as that shown on pl. XI in their paper. The spot where the animal was found was encircled by a red ring on the glass-cover, and there could be no doubt that it lay just as it had been placed by the authors. I did not open the preparation, as all I wished to see was clear enough. I found what I expected: perfect similarity between this specimen and my own males —, and the statements of the authors proved to be incorrect in the following important points: 1) »Les pattes nageoires font complétement défaut, ou sont réduites a des appendices difficilement visibles (pt.)<. The first part of this sentence is right, but to judge from the specimen in hand, the two dots marked pt. are spots possessing a slight deviation in the refraction of light, and situated beneath the inner side of the skin; according to my expe- rience with other animals, they are accidental. 2) »La partie postérieure du corps est divisée en deux renflements arrondis renfermant chacun une sphére a contour trés net dont le contenu est formé de quatre sphéres appliquées les unes contre les autres et déformées par pression réciproque comme les blastomeres d'un oeuf au stade quatre de segmentation. Les deux sphéroides sont des spermathéques« (p. 346—47). In the following pages I also call the two globules spermatothece, though I am not abso- lutely certain that they are not testicles; so far we agree, but no further. In the male of their preparation there was no vestige of a fold in the middle of the body. The spermato- thece showed inward folds which were not nearly so regularly arranged as it would appear from their description and figuring of the contents, nay they seemed to be empty. A careful and exact adjustment of the microscope showed that the granular substance usually contained ‘in the animal was outside the spermatothece, though a less accurate adjustment might give the impression that it also was inside; filled spermatothece have a very different look. The folds are easiiy explained by the flattening of the animal through the pressure to which it had been exposed. 3) About the antennule they write: »elles sont formées d'une saillie basilaire sur laquelle est inséré un article unique en batonnet terminé par une pointe courte«. However, this »saillie basilaire« in their preparation is considerably longer and somewhat different in shape from their figure of it; it is in fact the antennule itself (comp. my figure pl. XIT, fig. 3k.). What they call »un article« is the olfactory seta; nor is its extremity so slender and pointed as they represent it. 4) They say about the mouth (p. 346): »La membrane de la ventouse est soutenue par de fins rayons chitineux constituant les génératrices du tronc-cdne. Ces rayons ont été 8 vus par Satensky et par Hansen dans la ventouse de Spheronella et de Choniostoma. Mais le premier de ces observateurs les a consideres comme de simples replis de la mem- brane; le second n’a pas vu la membrane et a pris les rayons pour des cils chitineux. Un examen trés attentif peut seul permettre d’éviter cette double erreur«. In spite of this well worded phrase, I must observe that they have not arrived at any better result than the predecessors they criticise. The membrane exists without folds and without »rayons chitineux«, for these »rayons« are free hairs, »cils chitineux«, which originate at the base of the mem- brane, leaning freely against it on the outside, and in their own preparation these hairs, as usual, stand clearly out beyond the edge of the membrane. 5) »La premiere patte machoire (mxpi) est réduite a un long stylet droit aigu, beaucoup plus simple que l’organe correspondant du male de Spheronella« (p. 346). What they describe and figure here is only the terminal joint of the mavilla (according to my definition of this pair of limbs); it is not straight, but slightly curved, in their own type specimen, as well as in my drawing (pl. XII, fig. 3 k.). They have also overlooked the very large, long and broad basal joint, which appears distinct enough in their own type; if they had seen it, they would have found the missing resemblance with Spheronella, and it seems difficult to understand this gap in their observation. 6) However, the climax of the incomprehensible is reached in their description of the maxillipeds. In their text they mention three joints, of which »le troisiéme se prolonge en une dent crochue«, yet this »dent« is drawn as a claw-like joint, which is well separated by an articulation and can be folded up towards the joint above it. But in examining their type specimen, I found that it agreed perfectly with my figure on pl. XII; what they describe and draw as the three first stout joints, indeed is only one single joint without a vestige of the two articulations they mention and figure. The »dent crochue« is really jointed on, as they figure it, but furthermore, in their own preparation it consists of two distinct joints, and I cannot have misunderstood their text, for their statement about the claw »a laquelle fait face un petit tubercule pointu« is fairly correct. So, seeing that their own type specimen agrees exactly with my illustrations, I leave it to the reader to compare their description, and especially their figure, with mine, and to find out how they can possibly have been so much mistaken; as for me, I am at a loss to understand it. I have two reasons for yiving this detailed demonstration of the mistakes committed by the authors in their description and figure of this male specimen. In the first place I wish to verify in detail the identity of their species with my own, secondly I wanted to be able to refer to this substantiation in the following pages, where I shall have to point out that in a later paper the same authors have made considerable mistakes in their description of two other forms, of which I have not seen their type specimen. The authors (p. 356) state their opinion that the family Choniostomatide is nearest akin to Chondracanthide, Lernzopodide and Ascomyzontide. I agree with them as to 9 Lernxopodide; Chondracanthide seem to me to differ much more, and Ascomyzontide do not show any real relationship. The authors quote and criticise at great length all that has been written about this family, but in their eagerness to exhaust the matter, they seem to go a little too far. They give a long quotation from H. Kroyer: »Monografisk Fremstilling af Slegien Hippolyte’s nordiske Arter (Kgl. Danske Vidensk. Selsk. Skrifter, Nat. Math. Afh. LX, 1842, p. 263—64)<« in order to prove that this excellent investigator was the first to discover an animal of this family, and that his specimen belonged to the genus Choniosfoma. They quote the passage in Danish (p. 368—69) and in a French translation; the latter is correct, except in three points, of which one may be called a very free translation, whereas the others are indeed important mistakes and will be mentioned presently. Kréyer states that he has found a specimen of Hippolyte gibba (from Spitzbergen), whose carapace was much swollen on both sides; however, he found no Bopyrid in it, but about a score of sub-globular, yellowish white bodies of different size (from ?/5‘ to nearly 11/2‘’ in diameter), which were lying free and unconnected side by side. He supposes them to be eggs of an unknown parasite and adds: »the smaller ones I found filled with a yolk-like, granulous substance« {»de mindre af dem har jeg fundet opfyldte af en xggeblommeagtig, grynet Masse«|, which Giard and Bonnier translate as follows: »Les plus petits étaient remplis dune masse grenue ressemblant a des oeufs«, but this gives a very different meaning from the word »yolk-like«, and may quite well be understood, as if the globules were ovisacs containing the eggs of a Choniostoma, though Kréyer’s expression does not imply such an idea at all. Kroyer continues: »In the larger globules, which were probably very near maturity, I have noticed a rather long (6—7‘“), thin, vermiform body. It may be, that some leech-like animal develops itself out of these eggs« [>i de storste, som rimeligviis vare nerved Modenhed, har jeg iagttaget et temmelig langt (6—7‘“) tyndt, ormedannet Legeme. Maaske udvikler der sig altsaa af disse Aig et igleagtigt Dyr«|. Judging from the two sizes indicated by Kroyer, we might suppose that the larger globules were females, the smaller ones ovisacs of a Choniosloma, but it seems to me very improbable, that a naturalist like Kroyer should not have seen that the small globules in reality contained eggs or larvee, instead of supposing their contents to be a yolk-like, granulous substance, and his statement that he found a vermiform body about 13—15 millim. in length in the large globules, must in my opinion do away with any idea that it could be the female of a Choniostoma (comp. my description of this genus later on). But then, how shall we explain that Giard and Bonnier could adyance such an opinion? Well, in their translation of Kréyer’s description of the contents of the large globules, they translate the first words: »7 de storste« {»im the larger ones«| by: »prés des plus gros« which gives quite a different meaning, allowing this remarkable, vermiform body to be taken for a free animal belonging to another class. hus two faults in their trans- lation of Kréyer lead them to find a similarity which does not really exist between a Cho- niostoma with its ovisacs and Kréyer’s description. [| am unable to tell what the objects 2 10 examined by Kroyer could be, but the suggestion that the large globules which, according to his statement, contained a long, vermiform body of about half an inch or a little more in length, should be females of a Choniostoma, indeed seems overbold to me, even in our golden age of loose conjectures, and if we could really suppose Kroyer to have made such extraordinary mistakes in his statements, we should indeed consider them worse than worthless and deserving of everlasting oblivion. When in 1889 I read this passage by Giard and Bonnier, I remembered, that while working at my previous investigation of Choniostoma, I had perused the short paragraph in Kréyer’s excellent monograph: »Et Par Bemerkninger om Snyltedyr paa Hippolyter« |»Some remarks about parasites on Hippolyte«| (p. 262—65) without finding anything at all applying to the parasite I was going to describe. On p. 371 the two authors write further: >I est singulier que Hansen ait laissé passer inapercue l'observation de Wrser, et surtout le passage beaucoup plus important de son compatriote Kroyer«. I shall presently make a few remarks about Weber, and as far as regards my overlooking Kroyer, I will only observe that it would certainly have been wiser of Mssrs. Giard and Bonnier, whose success in finding a pretty good proof in favour of their assertion was entirely owing to two rather unfortunate faults in translation, to consider whether they themselves had not read Kroyer wrongly, before accusing me of having done so, especially as this countryman of Kréyer's has repeatedly expressed his appreciation of him, precisely in the report on the results of the Dijmphna-expedition, and who about twenty pages earlier (p. 258) has pointed out Kroyer’s description of small, but interesting, joints in the antennz and in the mandible-palp in another Copepod. Concerning the censure of my ignoring Max Weber, I will make a few remarks. In my dissertation: Fabrica oris Dipterorwm, 1883 (Naturh. Tidsskr. 3 R. B. XIV), in order to avoid unnecessary length, I did not mention all authors and their opinions, but confined myself to the statement (p. 8) that I had made a rule of leaving out writers whom I did not consider as having added new elements of importance to the existing knowledge of its [the mouth’s} structure, or its use for classification, or whose incorrect views had proved to be of no importance. I have followed the same principle in later works, but it seems that, in order to avoid the accusation of ignorance, I shall have to use the same precaution as in my dissertation, where, immediately after the quoted passage, I enumerate the authors who are not mentioned, because they are unimportant with regard to the subject in hand, though they may be excellent in their treatment of other branches. I do not think that I had noticed the above-mentioned erroneous observation by Max Weber before publishing my essay (of which separate copies were distributed in July 1886), and I cannot tell now if [ should have quoted it, had I known it then, but, as a matter of fact, I had read and understood it before I wrote the French résumé (in which, as mentioned above, I corrected my omission with respect to Salensky’s (to me) important work) and I purposely forbore mentioning Werner, considering his observations irrelevant, though four or five lines would have been sufficient to reproduce their essence. The interest attached to his statements 11 consists in his indicating new a locality for Choniostoma on the other side of Nova Zemblia opposite to mine (the Kara Sea), and that an otherwise very deserving author has committed a most peculiar mistake. That is all; whether I ought to have mentioned the subject is a matter of opinion; at the time I thought it might as well be left out. I shall pass over several other remarks which might call for censure, and take up some hypotheses set forth rather hesitatingly by the authors, p. 852—-53. After having declared themselves at a loss to understand that a Choniostoma with its ovisacs can cause a swelling in the carapace of a Hippolyte entirely resembling that which is produced by Gyge Hippolytes, they write: »I] nous parait beaucoup plus vraisemblable d’admettre que le Copépode a infesté les Hippolytes déja parasités par les Gyge, et qu il supplante les Epicarides ou tout au moins profite pour se loger de la déformation produite par ces derniers«. To this conclusion they add a doubt which I think rather irrelevant, and say further: »Néau- moins en rapprochant l’éthologie d’Aspidoecia de celle de Choniostoma, il nous semble bien probable qu'il existe un rapport, soit de parasitisme, soit de mutualisme, entre ces parasites et les Epicarides des genres Aspidophryxus et Gyge:. However, they go still further. They have found a genus of Epicaridea, Podascon G. and B., on a species of the genus Ampelisca, and Salensky has found numerous examples of a Spheronella in all stages on an Amphipod of an allogether different family. Here we should think it would be rather difficult to establish a connection between the Epicaridea (Podascon) and the Choniostomatide (Spheronella), which live »exactement dans les mémes conditions«; nevertheless they continue: >on peut se demander s'il n’a pas existé autrefois entre ces deux groupes de parasites des rapports analogues a ceux que nous avons cherché 4 démontrer entre les autres Choniosto- matidés (Aspidoecia et Choniostoma) et certains Epicarides«. With the word »autrefois« the authors resort to the past, but it will be impossible in a case like the present one to gain any perfect or imperfect knowledge concerning the former state of things. We confess that this invention would be ingenious if — as sometimes happens where an excellent thing is carried to an extreme — it had not overstepped the limit and become ridiculous. My experience, which is based on very extensive researches, enables me to declare that, as far as the present time is concerned, these hypotheses, which the authors repeat with additional remarks in two later papers, are entirely destitute of foundation. Of infested Isopoda this work mentions four examples of three species with three species of Spheronella; of Cumacea with parasites in the marsupium twenty-four examples belonging to six species (the parasites belong to five species), and of these six species I have examined several hundred specimens, in order to find those that were infested. Of two species of Cumacea seventy-three instances were found with (two species of) Homocoscelis under the carapace; finally, one hundred and forty examples of Amphipoda (belonging to twenty-eight species) were found and proved to be infested with twenty-eight species of Spheronella and Stenotocheres. Of these twenty eight species of Amphipoda I have examined several thousand specimens. So the result is, that of all three orders together I have seen about two hundred OF 12 and forty specimens belonging to thirty-eight species infested with Choniostomatide, but neither on these, nor on any other of the thousands of individuals belonging to these thirty-eight species, have I found one single Epicarid. So we have done with thirty-eight of my species of Choniostomatidee, and of the five remaining species two may be passed over, viz. the species of the genus Mysidion, for neither I nor any other author have found any Hpicarid in the marsupium of the hosts of Mysidion, viz. the genera Hrythrops and Parerythrops. Only on the outside of the body of the species belonging to the genus Hrythrops, and in the branchial cavity of two species of Hippolyte, others as well as myself have found altogether three species of Choniostomatide, and at the same time species of Epicaridea. As a rule the animals of each order were found on separate specimens; in one case observed by myself, and in one case mentioned by Giard and Bonnier, animals of both orders were found on the same specimen. Still it can be proved that these two quite different types of parasites, though perhaps in very rare cases they may be in each others way, stand at least in no other mutual relation. As for Choniostoma Hansenti, I can prove that the animal itself produces the swelling on the carapace (comp. my special description of this animal), and in the only case where Choniostoma and Gyge were found on the same side under the carapace, a male and a still smaller female of the latter genus had lodged themselves in a large swelling, which was inhabited by an adult female Choniostoma with eleven ovisacs. As for the last of my species — Aspidoecia Normani — I have found it on twenty-one specimens of all five species of the genus Hrythrops, but I found no Epicarid on any of these animals. Moreover, the occurrence of Aspidoccia, not only on the shield, but also on the exterior side of the thorax and on the six abdominal segments, as well as on the eyes, proves sufficiently that it stands in no connection whatever with Aspidophryxus, which parasite lives only on the carapace. Immediately after the paragraph criticised above the authors write: »Toutes ces considérations sont sans doute fort hypothétiques, mais elles peuvent inspirer de nouvelles recherches et indiquer la voie aux investigateurs. Elles ont de plus lavantage de rattacher par un lien éthologique commun les types de Copépodes si étranges qui constituent la famille des Choniostomatide«. This »lien éthologique« is quite broken now and will scarcely ever be restored. As for the first part of the quotation, I regret to say that it has indicated no path to me, and that, far from having been inspired by their »considérations«, I have been obliged to waste time and space upon proving the untenability of some unwarranted hypotheses. To suggest such hypotheses indeed is not very difficult, and most zoologists have imagination enough to invent scores of them. If productions of this kind had any real value, it would be easy to promote the progress of science. But I confess that, though I honour everybody who is capable of suggesting a theory which proves to be well founded and fertile in results, I have always felt and, as time goes on, feel more and more distaste for superficial conjectures. 13 A. DetiaA VALLE: »Gammarini del Golfo di Napoli« (Fauna und Flora des Golfes von Neapel, 20. Monographie, 1893, 4to). In the chapter » Parassiti dei Gammarini« (p. 289—90) the author informs us of some observations he has made, and suggests some hypotheses about Spheronella. The species on which Salensky found his Speronella Leuckartii is said to be Microdeutopus gryllotalpa, and the author has found it in the locality indicated by the discoverer of the species. He further states that he has found the same Spheronella on Ampelisca diadema Costa, where it lives under the same conditions as Podascon Della Vallei G. and B. And he proposes three hypotheses, viz. that Spheronella changes colour according to its residence, in order to look like the eggs of the two different species of hosts; that it does not live at the expence of the host itself, but by consuming its progeny, and that for some time after having left the egg, the young Spheronella is entoparasitic, not ecto- parasitic, developing itself in the oviduct and consuming the eggs successively as they appear. In support of this last conjecture he states that he has found on an Ampelisca a Spheronella with its multitude of ovisacs, which host at the same time »racchiudeva in uno dei suoi ovidutti, verso l’estremo esterno, uno piccolissima Spheronella, in cui nondimeno erano gia ben visibili le uova quasi mature« (p. 290), but in spite of this rather peculiar observation, his conjecture seems unduly hasardous, as an attentive perusal of Salensky’s excellent treatise with the description of the pupa stage, which follows the larval stage, would have shown its absurdity. Besides, Giard and Bonnier have refuted all these hypotheses in a later paper; they justly maintain that there is a physiological reason for this castration (» castration parasitaire«) effected by the parasite on its host, and they consider the form found on Am- pelisca as a different species from Sph. Leuckartii, in which no doubt they are right. So I think I need not throw further light on these questions. — About Rhizorhina Ampelisce H.J.H. the author in his Bibliographia, p. 897, only writes: »Questo nuovo Copepodo rassomiglia molto alla Spheronella Leuckarti, Salensky. The quality of this resemblance is treated in the following pages. A. GiArp et J. Bonnier: » Ser deux types nouveaux de Choniostomatide des cotes de France: Spheronella microcephala, G. et B. et Salenskia tuberosa, G. et B. (Comptes-rendus de l’Acad. d. Sc., 25 sept. 1893). The contents of this preliminary note appear in a later essay, much enlarged and — in one point — altered. A. Grarp et J. Bonnier: » Contributions @ Vétude des Epicarides (Bull. Scientif. de la France et de la Belgique T. XX V, 1895 — the part headed: » Les Spheronellide«, p. 462 —85, Pl. XIJ—XII1). This part calls for a detailed comment. The authors describe and figure the female and eges of Spheronella microcephala G. et B., a species found on four specimens of Ampelisca tenuicornis Lilljeborg from Croisic. Doubtless the frame of the head is incorrect, for a list like the one represented in the illustration (Pl. XII, fig. 43) as going from the outermost posterior angle towards the median line behind the base of the maxillz, does not exist. If there is a connection between the frame and the sub-median skeleton -- which by the by they have not seen — but which is 14 never wanting in any Spheronella, there must also be a list behind the maxillipeds. How- ever, the whole frame seems to me most problematical, nor have I found it in specimens which, as far as I can judge, belong to the same species. JI should not have dared to suppose so great a fault in this illustration, if I had not seen their type specimen of the male of Aspidoecia, which enabled me to ascertain their astonishing mistakes in the repre- sentation of several organs, especially in the maxille and the maxillipeds (comp. above p. 7—8). Moreover, they have decidedly overlooked the maxillule, which I have never found wanting in any female of this family. About the maxille (»les maxillipédes internes«) they say that they are »formés de quatre articles« (p. 464), but this is wrong, for these limbs in all females, males and larve of this family contain at most three joints, and the two last joints are even frequently so completely fused that we only find two distinct joints, as shown in my illustration (pl. VIII, fig. 2d) of the head of this species. Neither do I doubt that their representation of the maxillipeds with their strange flexion and the second joint thick and quite as long as the first, is entirely wrong. Their description and figure of the genital region (p. 465, pl. XII, fig. 44) is not successful either. By the words of the text: »un are de cercle chitineux (c) qui, postérieurement, se termine par deux branches...« and by the illustration, it is seen that they have turned the whole part wpside down, as in reality both branches turn forward towards the head of the animal, seen from the ventral side (comp. my fig. 2a on pl. VIII). The chitinous arch with its branches is pretty correct. Their representation of the genital apertures and their muscles is perfectly correct, while the apertures marked @ and designed as being »les ouvertures d'une paire de grosses glandes ... les glandes collétériques« — are the orifices of the receptacula seminis (comp. my de- scription below and my figures of several other species of the genus). In fig.2f on pl. VIII, as in several other instances, I have not represented these orifices, but after a renewed examination of the same species, I can state that the orifices, leading to the receptacula seminis in my Sph. microcephala G. and B., ave found precisely in this place, and from these openings each of the middle-sized receptacula — forming an oblong sac — curves gently backward and somewhat inward towards the centre. I am at a loss to understand anything about these glands illustrated by the authors. They also represent a pair of very large »receptacula seminis« as opening into the genital apertures; though unable to explain what they are, I am positive that they are not what the authors suppose them to be. Finally, what they describe as follows: »Au centre méme de laire génitale i] existe un espace cordiforme clair (ec), avec trois petites vésicules granuleuses aux trois sommets, la supérieure étant la plus grande et la plus nette; toute cette partie est située profondément, sous le tégument« is certainly no organ or organs, but accidental formations produced by coagulation or the like. The authors have taken their species on Ampelisca tenuicornis Lilljeborg from Croisic (south coast of Brittany), and their determination of the host has been confirmed by the eminent Carcinologist, Prof. G. O. Sars. The specimens described later on in this work, which I have considered as belonging to the same species, were taken on Ampelisca typica 15 Sp. Bate. During the interval between the appearance of the first publication and that of the principal essay I corresponded with the authors about these questions, and as they quote some of my written statements, [ must make a few remarks. It is not only the fact that Sph. microcephala had been found in Denmark on Ampelisca typica and in France on Amp. fenuicornis, which I may have thought »trés curieux«, but in examining a large quantity of Danish material of Amp. tenuicornis, not only had I found no specimen of Sph. microcephala whatever, but I had found several specimens of a very different species (Sph. longipes n. sp.), so it struck me as »very curious« that Amp. tenuicornis from the Danish coast had a parasite which it had not near the French coast, while in the latter locality it had a parasite belonging to the same genus, and which was not found on the Danish Amp. tenwicornis, though this very parasite lives in Denmark, but had passed on to Amp. typica. However, I will add that future researches may prove both species of parasites to live on both species of hosts in either locality. In this case we shall wonder no longer, but until further notice we have reason to find the circumstance curious. Subsequently the authors enter upon a critique of Della Valle’s observations and hypotheses. To the species found by Della Valle on Amp. diadema Costa, they give the name of Sph. diadema G. and B., which consequently is put down without description. However, as I have briefly stated the principal points of Della Valle’s observations on a former page, I may pass them over here; I will only add that I am not prepared to judge of the value of the reflections set forth by Giard and Bonnier about the colour of the eggs of parasites — though I can say for certain that Della Valle’s opinion is wrong. On p. 462—63 the authors repeat the above criticised suggestion of a connection between Cho- niostomatidee and Hpicaridea: »Les Choniostomatides sont-ils des parasites des Hpicarides dont ils prendraient la place en les faisant périr, ou les Epicarides facilitent ils seulement Ventrée des Choniostomatides en produisant sur les Malacostraca des déformations et une castration parasitaire plus ou moins complete? C'est cette derniére hypothése qui nous parait actuellement la plus vraisemblable«. That Della Valle had found a species of Spheronella on two specimens of Ampelisca diadema, and a species of the genus Podascon (an Epicarid) on two other specimens of the same Amphipod indeed was the only fact of interest which had occurred since their previous work in 1889, but this fact only proves that a fourth species of Choniostomatide has been added to the three, of which it has been stated above that they live on species infested with Epicaridea, and this is of the slightest importance compared with the statistics I give on p. 11—12, and the conclusions drawn from these statistics and from my observations. We now arrive at the most unfortunate idea advanced by these authors, their grouping of Choniostomatide H. J. H. and of Herpyllobiide H. J. H. as sub-families (with the suffix ine) of the family Spheronellide G. and B. In order to refute this combination — one of the most inappropriate I have ever met with in Carcinology — and some ‘hypotheses connected with it, I shall also have to mention the family Herpyllobiide. 16 In 1892 I published an essay: »Rhizorhina Ampelisce n. gen. n. sp. En ny til Herpyllobiide n. fam. horende Copepod, snyltende paa Ampelisca levigata Lilljeb. (Entomol. Meddelelser, 3. B. 5. Hefte p.207—34, Tab. III), which in the first place contains a detailed description of the above-mentioned new and very curious form, in the second place makes an important contribution to the knowledge about Herpyllobius Stp. and Ltk.; finally the new family Herpyllobiide is established, and the genera — seven in all — which can with more or less certainty be referred to it, are grouped together. Two of these genera, Trophoniphila M’ Intosh and Oestrella M’ Intosh, are described so defectively that we prefer not to consider them in this place. The female of the other five genera has a globular or oblong body without any vestige of mouth or limbs; posteriorly are two genital apertures, each with its ovisac. The front part of the female of Rhizorhina forms a short, slender stalk, which pierces the skin of the gill of its host; the inside of this stalk consists of two tubes. Just beneath the skin of its host the stalk expands very much, the tubes are consi- derably dilated, they separate and ramify irregularly throughout the gill, even entering somewhat into the body of the host. In the genera Herpyllobius Stp. and Ltk. and Hury- silenium M. Sars, the stalk, which consists of a single tube, is found on the ventral side of the body, pierces the skin of its host and expands inside it like a collar, but this collar is surrounded by the root of a large, oblong, foliaceous or irregularly sausage-shaped body, which is decidedly homologous with the tubes of the Rhizorhina, and, like these, has the function of drawing nourishment from the host to the external, limbless body, whose business it is to develop the eggs. In Saccopsis Ley. and Bradophila Ley. Levinsen has indeed found the stalk, but no body at the expanded end of it in the body of the host. However, he had but slight material of both forms to work with, so I will now state as my personal opinion, that a body, or one or two tubes, may have proceeded from the stalk into the body of the host; otherwise it wouid be impossible to understand how the parasites could get their food. Moreover, L may mention that, when (in Noy. 1896) I spoke to the author, Inspector G. M. R. Levinsen, about the matter, he felt inclined to share my opinion. Giard and Bonnier (in their above-mentioned paper) describe a new parasite, Salenskya tuberosa, of which a single specimen was found on Ampelisca spinipes Boeck from Croisic. They confess (p. 474) that it »présente certainement une trés grande ressemblance avec Rhizorhina ampelisce.... et nous avons longtemps hésité a maintenir le genre Salenskya, crée par nous {in the preli- minary note] quelques mois apres la publication du travail de Hansen«. Still they think they are justified in maintaining it, »au moins provisoirement«, on the following basis: »Au lieu d’étre fixé a son héte par des racines rappelant un peu celles de Sacculina, ou par un renflement comparable a celui des Herpyllobius, la femelle de Salenskya possede un appareil chitineux spécial, qu'on pourrait rapprocher plutot de celui de Saccopsis terebellidis figuré par LEVINSEN...... «(p. 475). I have just spoken of Saccopsis, and I will now express my opinion that if a specimen of Salenskya is found again on Amp. spinipes, and the part of the host occupied by the parasite is cut off, this part will contain internal tubes caactly 17 like those I have described in Rhizorhina; the two apertures mentioned and described by the authors are the roots of these tubes. Separate copies of my essay about Rhizorhina were distributed in July 1892 (one of them was sent to the authors). Their preliminary note, in which they establish Salenskya, mentioning its »appareil fixateur en forme d’amphi- disque ou de bouton de manchette«, is dated Sept. 25th 1893, but it is quite evident that, at the time their manuscript was sent to the press, they had not read my essay. So, having but one specimen of the animal to work upon, they committed the same mistake which I had made with my first specimen of Rhizorhina: without having any idea of the tubular system inside the host, I detached the visible part of the parasite, thus breaking the stalk which united it to the hidden part. After what I have just said about their investigation of the male Aspidoecia, I am quite justified in not trusting their statements in a question so difficult as that concerning Salenskya, where their judgment rests on the examination of but one individual. The result is that the genus Salenskya G. and B. must be cancelled, being established only on this one single character. Whether their species differs from Rhizorhina Ampelisce will have to be proved by ascertaining if the slight differences between our repre- sentations of the males agree with facts. Though this on the whole may possibly be the case, I doubt that they are right in stating that the larva of the parasite they describe has two orifices for the ducts of the genital organs; I have only found one hole surrounded by a somewhat thickened ring. The authors quote from their preliminary publication (p. 475—76) a long passage, in which they suggest »progénése« and »dissogonie« in the male of Salenskya. They now give up these theories, saying: »Les recherches de Hansen prouvent que chez Rhizorhina la métamorphose régressive existe bien chez les males de ce genre d’Herpyllobiine et quelle est tout aussi accentuée que chez les Choniostomatine.« However, the last sentence which is meant to establish a relationship between the two groups to each other, is very misleading, as the male of Rhizorhina (and Herpyllobius) is a body entirely without limbs, mouth or any other external organ or internal muscles, with nothing in fact but genital organs, the male of any Choniostomatid whatever is a highly developed animal with anten- nule, a very complex mouth with mandibles, besides maxillule, maxille and maxillipeds with some joints, internal muscles etc. So in saying: »Ce charactére différentie] {»pro- génese« in Salenskya and other Herpyllobiidee] entre les deux sous-groupes ne peut done étre maintenu«, they are perfectly right, but such a negative feature does not imply any kinship. However, the principal points are contained in the following paragraph, and in order to criticise it I am obliged to quote the last half of p. 476 and a little of p. 477 in their paper; I will, however, divide the quotation into three parts. They write: »Le reste de l’organisation concorde d'une facon remarquable, non seulement chez la femelle ot, en raison de la dégradation, toute comparaison peut sembler dépourvue de valeur, mais aussi chez les males et les embryons: méme tendance a la disparition de la deuxiéme paire 3 18 d’antennes, méme structure de l'appareil buccal avec la ventouse si spéciale et les appendices transformés en stylets, méme disposition des membres thoraciques, etc. Les jeunes individus surtout présentent une ressemblance extraordinaire et indiquent nettement la parenté des deux groupes. Mais il est un caractére du male sur lequel nous désirons particuliérement attirer lattention, parce qu'il est trés exceptionnel et qu'on ne Je retrouve dans aucune autre famille de Copépodes, en dehors des Choniostomatine et des Herpyllobiine. Les canaux excréteurs des glandes génitales males débouchent dans la partie cépha- lique de Vanimal et dans le voisinage de la bouches. Let us examine this a little more closely. Though the authors think that the larvee in particular show »une ressemblance extraordinaire«, we find that these larve, which indeed may be said to be in the first Cyclops-stage, resemble each other less than the larve of a Choniostomatid and of an Achtheres respectively, according to the illustration given by Claus (Zeitschrift wissensch. Zoologie B. XI, Taf. XXIII, fig. 5). At any rate, the likeness between the mouths of the larve of a Rhizorhina and of that of a Choni- ostomatid is not so great as the authors seem to think, and it is certainly much smaller than that between the mouth of a larve of the last-mentioned group and e. g. of a larva of Pennella. The maxille of the two groups deviate much from each other in shape and position etc. Several great differences between the males of Choniostomatide and of Her- pyllobiida have been pointed out above, and we shall soon mention more. The differences between the adult females also seem to be so great that we are struck by the astonishing boldness of the assertion that: »en raison de la dégradation, toute comparaison peut sembler dépourvue de yaleur«. In the former type, the Choniostomatidze, the female possesses at least the antennule, a well-developed mouth with mandibles, maxillule and maxille; in the latter, the Herpyllobiide, the body has no vestige of these organs or of any limbs, and in the three genera which are examined so thoroughly, that our knowledge about their nutrition is perfectly reliable, we know that it takes place through a large mysterious body (in Herpyllobius and Sileniwm) or through an equally mysterious tubular system (hizo- rhina) which is found in the body of the host, and which has a most curious, hitherto unexplained development (comp. my essay about Rhizorhina). Indeed, I can find no other likeness between the females of these families than the small size of their bodies, their sub- globular or oval form, and their two genital apertures, and as this last character seems to be common to all parasitic Copepoda, we might as well pass it over. But still more objectionable is the statement printed in italics, that in the males of both families the genital aperture is found on the head near the mouth. I shall begin by speaking of Herpyllobiide. The authors substantiate their opinion in these words: »Ce caractére, tellement extraordinaire que nous ne l’avions signalé qu’avec réserve dans notre étude sur Aspidoecia et dans nos recherches plus récentes sur Salenskya, Hansen l'a mis 19 completement hors de doute dans son beau travail sur Rhizorhina ...« This requires a comment. I have proved the following facts. The males of Rhizorhina and of Herpyllobius are not the larve. The larva fastens itself to the female by a gluey substance, after which all its muscles etc. are dissolved; the limbs are emptied of their contents and the whole plasma of the larva contracts and surrounds itself with a new skin, thus forming a male without limbs, mouth or other external organs, and without visible internal organs except testicles and their efferent ducts which gradually develop themselves. In the Rhizorhina this male remains inside the skin of the larva, pushing its remarkable spermatic ducts out through the hole in front of the mouth of this dead case. In the Herpyllobius the skin af the larva bursts, the male fastens itself with its front, and the spermatic ducts proceed (behind the attached end) through the split produced by the bursting of the larval skin. So in both cases the male is transformed to such a degree as to render a morphological orientation rather uncertain; at all events, we can no longer speak of »le voisinage de la bouche», as there is no mouth at all. This description of the male of Herpyllobiide will also give a sufficient idea of the immense difference between this animal and the males of Choniosto- matidee which, moreover, fix their spermatophores on the females in the usual way. The authors continue: »Chez tous les Spheronellide, les canaux génitaux males servent aussi a l’excrétion d'une substance cémentaire avec laquelle le male se fixe sur la femelle d’une facgon plus ou moins durable. Ce rédle nouveau et ces connexions singuliéres des canaux génitaux constituent a coup str le trait le plus saillant de la morphologie de la famille des Spheronellide, telle que nous la comprenons«, namely Choniostomatidz and Her- pyllobiidee together. The authors are bold indeed; they do not hesitate to suggest one hypothesis after another, the second more erroneous than the first. Now, to begin with Herpyllobiide, who has said anything that could justify the statement that the genital organs of the male secrete the viscous substance by which the animal attaches itself? The authors have seen nothing themselves, and they cannot base their statement on my essay about Rhizorhina, as I maintain that the larva of this animal attaches itself by a gluey matter proceeding from the mouth before the male is developed and before there is any indication of genital organs. The male keeps inside the skin of the larva, which remains attached to the female, and no further fixation takes place'). How then must we qualify the sentence the authors pronounce as if it were proved? To put it mildly, we can only call it a product of imagination. — We shall now turn to the second division of their »Spheronellide«: the Choniostomatidz, and here again we shall have an opportunity of considering their above quoted lines in italics: »Les canaux excréteurs des glandes génitales males débouchent dans la partie céphalique de l’animal et dans le voisinage de la bouche«. 1) In the Herpyllobius the male attaches itself a second time by its front end, but the genital aperture is found at some distance behind this fixation (Entom. Meddel. 1. c. p. 230). 3* 20 The authors have proved (1889) that in Aspidoecia the male is hinged by a thread which proceeds from a hole on the ventral side of the front part of the head: »ce filament est secrété par deux grosses glandes cémentaires probablement homologues de celles qui servent a la fixation chez les Cirripédes«. No doubt it is this comparison on which they base their opinion that the genital aperture is found on the head, and also that the spermatic glands secrete the viscous substance which forms the thread, as these organs are believed to perform this double function in the Cirripeds'). A slight basis indeed for such remarkable statements! The observation about the hingement of the male is correct, but then, has the thread to disappear in order to allow the spermatophores to come out of the hole, or is the order of the two processes to be inverted, or does the male possess another genital aperture on its front near the base of the thread? Unfortunately we get no answer to all these legitimate questions — though indeed we can scarcely imagine any possibility besides these three. No, the doctrines about the genital aperture on the head and the double function of the sexual organs in the Choniostomatida are postulates without any foundation. Within the family mentioned it is an ordinary phenomenon to find the male attached by a thread; this prevents it from being washed away and allows it to creep as far as the thread can reach, giving it frequent opportunities to fix its spermatophores on the entrances to the receptacula seminis. Besides, the genital aperture is not found on the head; in Spheronella paradoxa 1 have been able to prove the existence of two genital apertures at a short distance from each other on the ventral side of the trunk: from each spermatotheca proceeds an efferent duct forward and obliquely towards the median line, and these canals open on the posterior side of the depression between the first pair of trunk-legs, or at least somewhat behind the basis of the maxillipeds. But then, what remains of the hypotheses advanced as facts by the two authors, that the genital aperture of the male in the Choniostomatide is found on the head, and that the »canaux génitaux« secrete the viscous substance by which the animal attaches itself? Nothing, absolutely nothing! And what remains of their best proof - based on these organs —, that Choniostomatidee and Herpyllobiidze ought to be grouped in one family? Equally: nothing! except a rather surprising impression of the loose method of the authors: to establish unreliable conjectures as facts in order to prove an absurdity. Though I suppose that most readers have now formed a pretty clear idea of the creat differences between the two families, I will give a summary. The likeness between the two families is limited to the following features: both are parasitic Copepoda, in which the males are several or many times smaller than the females; in both sexes the body is small, sub-globular or oblong; the last larval stage of Herpyllobiide is the first Cyclops stage, it resembles to a certain degree the larva just coming out of the egg in the Cho- 1) T will not here enter upon criticisms which have appeared elsewhere about Darwin’s unfortunate statements upon this subject, nor on Giard’s later suggestions concerning Rhizocephala. 2] niostomatide. The differences between the families will be shown most clearly by giving a short description of each. In the Choniostomatidz both sexes possess at least antennule, a mouth with mandibles, maxillule and maxillz, and the males have always maxillipeds, and they fix their spermatophores on the females in a normal way. The female deposits its eggs in one or two free lumps or, in most cases, in ovisacs, of which at least four or five and sometimes more than twenty are found; the larve attach themselves by an adhesive plate on the forehead and — whether passing through the pupa stage or not — develop them- selves into animals of either sex. In Herpyllobiidz both sexes lack antenne, mouth and appendages; the females project a mysterious body or two ramified tubes into the host and draw nourishment through these organs. The males project from the anterior part of their body in advance of the mouth of the larval skin two long spermatic ducts, which are formed by a secretion in the genital organs, and through these canals nearly the whole sub- stance of the body, having been transformed in the service of propagation, is transferred into the female. The female has two ovisacs; the larva attaches itself by a gluey sub- stance proceeding from the mouth, and is transformed into a limbless male or female. In the latter case the animal forms a stalk which pierces the skin of the host, inside which it dilates and develops into the above-mentioned organ of nutrition. — Whereas the Cho- niostomatidz, on the whole, fit in well among the other families of parasitic Copepoda, the Herpyliobiidze remove themselves from the others by a series of very peculiar features, occupying a more isolated position than any other of the families. This, I hope, will be sufficient to prove that the juxtaposition by the authors of the two families in question as sub-divisions of one family, is contrary to all sound classi- fication. JI think also that sufficient light is thrown on the characters and hypotheses of the authors. The present work being a kind of monograph, I found it necessary to write this rather detailed critique of their publication. However, this task hhas not been at all pleasant to me, because in another branch, the Epicaridea, they have published works which must be considered the principal sources of our knowledge about important groups belonging to this large and difficult family. In the interest of the authors and of carci- nology, as well as for my own sake, I wish they had not published their four, at least not the two last of their contributions (the preliminary note in 1893 and their final essay 1895) about Choniostomatide. It would indeed have been very natural to postpone the publication of their two last papers, as their material of these animals (whose manipulation presents considerable technical difficulties) was rather scanty, and as, even-as early as 1891, they know that I was preparing a work based on very abundant material. (I need scarcely add that the fact of their publishing a report about one species previously to myself affects me very little; indeed I might easily have secured this priority by some »preliminary note«). Tf, nevertheless, they were intent upon describing their few animals, their researches might and ought to have been much better, and they ought to have abstained from filling up real or imaginary gaps by a number of unproved assertions and unwarranted hypotheses regarding structure, biology and classification. Nowadays many authors have a remarkable weakness for publishing innumerable immature notes, for building zoological card-houses, drawing up genealogical trees and inventing theories and hypotheses, especially where they know very little. Where they have acquired considerable knowledge based on thorough study of a large material, as a rule, they abstain more from hazardous conjectures. One result indeed has been obtained: Zoo- logy has been encumbered with endless preliminary notes, with papers abounding in faulty and defective representations and unaccountable postulates and reflections, so as to render the study of it troublesome to an almost unsurmountable degree. JuLes Bonner: Résultats scientifiques de la Campagne du »Caudan« dans le Golfe de Gascogne, Aout-Septembre 1895. Edriophthalmes. (Ann. de l'Université de Lyon, 1896)'). In an appendix to this valuable work the author describes and figures a new species, Spheronella sedentaria Bonn., which he has discovered in the branchial cavity of Cyclaspis longicaudata G. O. Sars of the order Cumacea, in a depth of 960 metres, lat. 44° 5’N., long. 4° 45' EH. He found an adult female, four ovisacs and a small specimen, which he considers to be a young female, but which is no doubt a male. The species belongs to my new genus Homoeoscelis, and comes very close to my H. minuta. He begins by describing the small specimen, and his description of its body, the borders of its head, its antennule, maxillipeds, trunk-legs and caudal stylets is essentially correct. He also corrects Salensky’s erroneous conception of the caudal stylets as a third pair of legs, but he has certainly overlooked the maxillule (comp. my drawings of the males of my species: pl. II, fig. 1i—1k and pl. XIII, fig. 1f—1g), which are never wanting in any species of the whole family — unless the outer part of the mandibles possibly may be the larger part of the maxillule, which might indeed be supposed from the drawing. The hairs surrounding the membranous border of the mouth are overlooked, and the basal joint of the maxillas which he mentions (his »maxillipede interne«) does not exist; what he takes for this joint is no doubt a part of the sub-median skeleton. As will appear from my subsequent description, the only feature by which the male and a young female of the same size of the genus Homoeoscelis can in all cases be distinguished from each other, is the distinctness of the genital apertures in the female. The author has found no such apertures, and this circum- stance, as well as the occurrence of the animal together with an adult female, indicates that it must have been a male. The author’s comparison of the female with the small specimen is correct; only his description of the genital area calls for a few remarks. He ') A special copy of this paper, kindly sent me by the author, arrived on Febr. 11th 1897, so that the present remarks had to be written and inserted in my work when a large part of the fair copy of it was already written. 23 is of the opinion that each of the genital apertures is provided with a separate frame, of which he presents a drawing (fig. 5e), but the anterior part of these frames is scarcely correct in drawing, as it is not likely to reach up to the »pore de fécondation« (orifice of the receptaculum seminis) which — we see — here for the first time is proved to exist in an animal of this family. Neither has he seen the median part of the firm chitine of the genital area which unites the two »frames«, but it must be pointed out, that without a special dissection — in which, moreover, a certain amount of practice is desirable — these details are difficult to discover. The whole description of the two small specimens is considerably better than the above-mentioned joint work on this family by the same author and Prof. A. Giard. / II. GENERAL REPRESENTATION OF THE FAMILY. ie order to facilitate the use of this large section it is divided into three chapters, the first of which contains a general view of the structure and development of the animals, the second of their habitation, biology and distribution, the third some general observations about the classification. Hach of these chapters contains several sub-divisions. A. Structure and Development. a. The Female. This sex is known in all species. The body is nearly always a little flattened — seldom more than a little; if seen from below or from above it is ovate or globular. The young specimens are generally much longer than they are broad; the adults are now a little longer than they are broad, now the reverse; sometimes their broadest dimension is a little in front of, sometimes a little behind the middle. Specimens which are going to lay, or have commenced laying eggs, are always somewhat — probably as a rule much — larger than old ones which are emptied of eges (pl. 6, fig. 3a shows such a female which is going to lay eggs, fig. 3¢ a female (with a male) which is emptied of eggs, both enlarged to the same scale). In consequence of this evacuation the animals frequently shrink and become vaguer of outline. We often happen to see specimens which have become crooked and irregular from pressure, otherwise all the animals are naturally symmetrical. The size of the adults varies considerably; in most species the diameter_of the animal seen from below is ?/s—1*/2mm.; it can even decrease to about °3 mm. (Homoeoscelis mediter- ranea), and Choniostoma Hansenii G. and B. can obtain a lenght of 5:3 and a breadth of 55mm. As a rule there is a certain proportion between the size of the parasite and that of its host; however, it must be borne in mind that the parasites themselves differ in size according to their habitation in the marsupium or under the carapace; in the latter place they are comparatively smaller. As a matter of course, small Amphipoda cannot - support large parasites, whereas large species like Calliopius leviusculus Ky., Munnopsis typica M. Sars and Hippolyte, are inhabited by large animals. In most species the regular, rounded, ovate or globular shape of the body is inter- rupted in front or a little behind the anterior margin on the ventral side by a small! pro- truding head, which as a rule is tolerably well defined at the basis. In the adults it is most frequently very small, compared with the trunk; in small, and_ particularly in recently hatched specimens (pl. VII, fig.2e, and especially pl. III, fig. 2c) it is of a very con- siderable size. The reason of this difference is that the head and ils organs do not grow or at most grow very little, whereas the trunk greatly increases in size, in order to give room for the mighty production of eggs. In some forms there is no separate head at all, so that its (very small) organs: antennule, antenne, mouth, maxillule, maxille and maxillipeds, are situated near each other anteriorly on the ventral side of the vaulted body. In Stenothocheres (pl. I) the thorax has two rather small pairs of limbs; in the other genera these limbs are quite minute or wanting altogether. In Stenothocheres we find a distinctly marked, prominent abdomen. — In no species the body shows any vestige of segmentation. This will give a general idea of the females. In giving a closer description of their structure I think the best plan is to begin with Spheronella and kindred forms, as the genus Stenothocheres, though in two important points — the size and development of the trunk-legs and the existence of an abdomen — more closely related to less transformed Cope- poda, in other respects is less qualified for serving as base of the description. I. Homoeoscelis, Spheronella and Choniostoma. Many species have a prominent, well defined head: the back, front part and sides are evenly vaulted and pretty well chiti- nised, and the chitinous border to the front and on the sides stands out a little beyond the ventral side, which is partly covered by a soft membrane, and has a somewhat concave surface. Seen from below, the sides of the head are arched posteriorly, for the above- mentioned protruding lateral borders are somewhat removed from the outline of the head; they are generally ciliated, whereas the margin of the frontal border is mostly hairless. As a rule, a narrow, arched, transversal list, or two narrow, parallel lists, proceeding from the posterior ends of the lateral margins and passing behind the basis of the maxillipeds, forms or form the posterior limit of the head. Sometimes this list is interrupted at the median line (pl. ITI, fig. 2c), sometimes it does not reach the lateral margins (pl. VIII, fig. 14d). Choniostoma (pl. XI) at first sight seems to have no distinct head at all, however, the above-described borders in front, posteriorly and on each side remain, forming a frame round the soft area, in the middle of which the mouth and its appendages are situated. The front part of this frame in Chon. Hansenii (pl. XI, fig. 2d) rises a little beyond its sur- roundings, thus representing the only remaining part of the anterior and upper surface of the head. Spher. Acanthozonis (pl. VII, fig. 5a and 5b) presents a fine intermediate form between Choniostoma Hansenii and the species that have a well defined, prominent head 4 26 (illustr. on pl. IL to pl. VIL etc.). In all species with well-developed head, or at least with the frame left, we see behind the frontal margin and inside the lateral parts of the frame a broad band of thin, soft skin. Somewhat behind the middle of the frontal margin is the rostrum (proboscis) with antennze and maxillule, and from this part backward towards, or quite up to the list behind the basis of the maxillipeds, we see a system of plates or lists. This system, which I shall call the swb-median skeleton, is partly or all the way divided into two halves by softer skin along the median line; its structure differs in nearly every species; as a rule it expands considerably in the middle of its lateral margins. The inner margin of the basal joint of the maxille touches the outer margin of the front part of the expansion, whereas the maxillipeds are articulated behind the expansion touching the outer margin of the narrower posterior part of the skeleton. In several species of the genus Spheronella, namely Spher. microcephala, S. dispar, S. insignis, S. Munnopsidis and S. marginata (pl. VIII, fig. 2d; pl. IX, fig. 3f and fig. 4¢; pl. X, fig. 4b, and pl. XIII, fig. 6d), there is no separate head and no harder chitinous borders (only in S. marginata and in S. microcephala there is a low border or a transverse list in front of the mouth (pl. XIII, fig. 6d, pl. VIII, fig. 2e)), whereas the sub-median skeleton exists, strongly developed as a solid plate in S. Munnopsidis (pl. X, fig. 4b), much reduced in S. dispar and S. insignis (pl. LX), and particularly so in S. marginata. The Antennule. In all species, except the five without separate head and without frame, the antennulz are well developed, and in these they are articulated to the solid frame, each at one of the angular points where the lateral margin merges into the frontal margin (comp. e.g. pL II, fig. 1h and fig. 3a, pl. XI, fig. 1a). Each antennula is usually composed of three joints, of which the second is generally the shortest, the third the longest. The front angle of the first joint is mostly provided with two or three shorter or longer sete; the terminal joint is rather well provided with bristles of different length, among which an olfactory seta (b) can be frequently pointed out. In the genus Homoeoscelis the antennulze become 2-jointed by the fusion of the second and third joints (pl. XII, fig. 1d). In Spher. decorata (pl. VIII, fig. 3e) the first and second joints are coalescent. In Spher. marginata the antennule (pl. XIII, fig. 6d) are constructed as in Homoeoscelis. In the other four species of Spheronella, which are devoid of separate head and of frame, the antennulz are situated at the same points, but fastened to the thin membrane, besides being shorter and reduced so as to show only indistinctly separated joints or no division at all. The Antenne. These organs I have been unable to discover in the species of the genus Homoeoscelis, and in Spher. modesta, S. dispar, S. insignis, S. marginata, S. Munnopsidis and S. microcephala, whereas they exist in the other species of Spheronella and in Chonio- stoma. They are always placed on the side of the rostrum itself near the margin of its expanded basal part, and they are always short, slender, generally 3-jointed (e. g. pl. V, fig. 2d), without hairs and terminating in one shorter or longer seta. In a few species, e.g. Spher. decorata, the number of joints is reduced to two, in Spher. antillensis (pl. III, fig. 2c) they are rudimentary, 1-jointed and have a very short seta, and so they are in the seyen remaining species belonging to the group of Spher. Leuckartii Sal. (comp. the systematic part). The Rostrum. It is always of good size and bluntly conical, or like a cylinder with dilated base. Its structure is very complex, and we will begin by studying its distal part, for the representation of which Choniostoma Hansenii (pl. X, fig.6a and fig. 6b) will serve as type. In fig. 6a the cylinder is seen sideways and without the expanded part at its base. At the margin of the terminal face of the cylinder originates a membrane which has the shape of a kind of border or very short inverted cone. In looking at it from the distal end (fig. 6b) we see that the membrane covers the whole terminal face, having the shape of a cup or perhaps rather of a flat funnel, as it leaves an oblong aperture at the bottom in the centre; this is the entrance of the mouth, beyond the margin of which the points of the mandibles are seen to proceed. In front of the mouth the membrane is divided in the middle by a deep incision; the opening thus produced is filled by an odd median plate, on each side of which is another plate which is partly covered by the membrane. In the illustration these parts are marked d. The membrane is downy at its edge (fig. 6b), and the whole inner surface of the funnel is covered with peculiar dots, which are smaller near the edge than towards the centre, and which probably represent tiny knots. Outside the membrane are seen a number of cylindrical hairs which are sometimes furcate at the apex (b). They are articulated to the distal edge of the cylinder at the base of the membranous border and, being longer than its height, proceed somewhat beyond its free margin. When — as in the present instance — the rostrum is cut off, it is easy enough to see that these hairs do not exist within the membrane, but only lean against it. In some species, e. g. in Spheronella curtipes (pl. X, fig. 2d), the membrane (viz. the free part of it) is considerably broader, in others narrower, than in Choniostoma Hansenii. The hairs in some species are much more numerous and much thicker than in others, and they often converge or diverge very irregularly, according to the position they happen to occupy; in a few species I was not able to discern them. My figures as a rule are too small to allow of drawing the membrane, but these hairs are drawn as well as it could be done. It must be observed that the shape of the mouth varies considerably in specimens of the same species; I have found it more or less funnel- or cup-shaped, in accordance with the angle formed by the membranous border and the surrounding hairs against the terminal face. In the systematic part of the present work the free part of the membrane together with the hairs is called the mouth-border. The outer surface of the rostrum shows seyeral harder chitinous lists, and when the rostrum is examined from its distal end, some harder parts are seen through the semi-dia- phanous membrane as circles, which are interrupted in front at the median line. In my opinion the distal part of the rostrum must be explained as being a highly modified /abiwm, or rather hypopharynx, which forms a kind of sheath round the mandibles and stretches so far towards the front that its edges approach very near to each other, and that the above- mentioned median part marked d. must be considered as the /abrwm. However, I am not 4t 28 able to account more fully for the structure of these parts and the attachment of the mandibles. When looking at the rostrum from its distal end (fig. 6b), we see them through the inter- mediate substance, like narrow lists in appearance, the free distal points of which are visible in the mouth-aperture and are somewhat different in shape on the right and the left mandible. Departing from the points, they turn outward, at the same time running down the rostrum, their basal end lying inside its walls rather far from the aperture of the mouth. It is only the distal part of the rostrum which can be considered as formed by the hypopharynx and the labrum, the proximal part must be chitine belonging to the ventral side of the head itself, which here has become cone-shaped or forms the foot and the proximal part of the cylinder. I draw this conclusion from the fact that the antenne, where they are found, proceed from the basal part of the cone or from the foot (Choniostoma, pl. X1, fig. 2d), and that the maxillule are situated on its lateral surfaces (see e. g. pl. X. fig. 6a, ¢, and many illustrations of heads of females seen from below). But these last-mentioned mouth-organs must be treated separately. The Maxillule ave found in all species. Each maxillula consists of a somewhat oblong plate which almost throughout its whole length is coalescent with the middle and the more proximal part of the rostrum, and in the latter place this coalescence is so complete that it becomes impossible to distinguish tke outline of the proximal part of the maxillula (fig. 6a), whereas its distal part (c) detaches itself from the lateral surface of the cylinder. Here it divides itself into two branches, the anterior of which forms, now a shorter or fairly long, now, and mostly, a very long process, which looks somewhat like a proximally very thick and distally more slender seta. The posterior branch has a quite similar structure. These two more or less setiform processes I consider as the principal branches of the maxillula; , they are never wanting, and as a rule they are somewhat curved (in the specially examined specimen of Choniostoma their terminal half was sinuous), and on examining the head from below, the anterior branch of the maxillula is mostly seen to proceed beyond the foremost part of the lateral margin of the mouth-border, the posterior branch behind the posterior — part of the same lateral margin, whereas the distal part of its plate and the base of the two branches are covered by the lateral part of the mouth-border, through which they can be seen (e. g. pl. V, fig. 2d and especially fig. 3d). Besides, in most species the maxilla possesses as an additional branch a process shaped like a stout and usually long seta, articulated to that part of the maxilla which is coalescent with the rostrum, and often so proximally that, in looking at the head from below, we get the impression that it is situated outside the basis of the maxilla. The basal part of this additional branch is frequently set off by an articulation. This branch is wanting only in Homoeoscelis and in the three species ot Spheronella which are parasites on Cumacea, and which have no separate head. The whole rostrum is movable, so that its distal part with the mouth is turned more or less forward or backward, now protruding, now receding considerably, which differences are seen most distinctly by observing the head sideways, and comparing the 29 position of the rostrum with the lateral margins of the head (comp. the rostrum in the numerous figures of males seen from left side). The Masxille ave always (except in one single species mentioned below) well deve- loped, often very powerful. They are situated far from each other, somewhat behind the base of the rostrum, on the outer margin of the sub-median skeleton, the expansion of which reaches their inner margin and frequently extends behind their posterior margin. Typically they have three joints, of which the first one is very thick, often not much longer than broad; the second and third joints together are usually shaped like a slender, distally somewhat curved cone, which can be folded up like a claw against the oblique terminal margin of the basal joint, and as a rule these two joints are coalescent, though sometimes we find them very distinctly separated (e. g. in Spheronella insignis, pl. IX, fig. 4¢). The basal joint is often provided with one or two protruding knots or taps, and its terminal margin at the articular membrane is frequently furnished with hairs, or, as in Spher. Munnopsidis, with some peculiar cylindrical bristles or fine processes (pl. X, fig. 4b); in Spher. decorata (pl. VIII, fig. 3e) and in S. modesta (pl. [X, fig. 2d) a part of the articular membrane between the first and the second joint is decorated with rather numerous small chitinous taps. The terminal joint usually ends in a point; in Spher. dispar (pl. IX, fig. 3f) the apex is blunt, but has several fine, setiform points. In Spher. marginata (pl. XII, fig. 6d) the maxille are quite rudimentary. The Mazillipeds are well developed in Spheronella and in Homoeoscelis. They are articulated on the posterior part of the sub-median skeleton and are usually somewhat closer to each other than the two maxilla. They consist typically of four joints, of which the basal one is thick, very long and always distinctly longer, often much so, than the others together; these can be folded up against it in a very acute angle. The basal joint is often decorated with processes, spines, rather long hairs, shorter or very short hairs, or very fine, conical taps; the hairs and taps are arranged in spots, stripes or rings. The second and third joints are slender, distinctly or indistinctly articulated or quite fused together without the slightest distinction. The third joint has generally on the inner side of its distal end a spine, which in those species of Spheronella which live on Cumacea, is provided with fine points, besides being sometimes broad and flat (pl. XIII, fig. 6d). The last joint is more slender than the others, somewhat curved and often ending in a point with one or two spines on the inner side behind the point; in most of the Spheronelle living on Cumacea the joint expands a little towards its somewhat flattened and rounded extremity, along the margin of which we see numerous fine and short, setiform processes. A somewhat similar structure is noticed in Spheronella Munnopsidis (pl. X, fig. 4b). In Spher. microcephala (pl. VILL, fig. 2d) the maxillipeds are weak and comparatively rather small, second and third joints coalescent and very short, the last joint very small and stunted. In the genus Choniostoma the maxillipeds are quite rudimentary (pl. XI, fig. 1a, g¢ and fig. 2d), and reduced to two very small or quite diminutive joints. 30 Before leaving the head I will mention some peculiar formations, which I am at a loss to understand. In Spher. frontalis we notice at the middle of the frontal margin (pl. VIII, fig. 1d) a strange cup-shaped, rather large expansion in which I have been unable to find any hole which might be the outlet from some gland. In Spher. modesta, on the ventral side of the protruding frontal border, inside its margin we see a square of considerable size (pl. [X, fig. 2d, x) with rounded corners, which seems to be pierced with rather numerous holes. The Trunk. The body — except the head — of course corresponds to thorax and abdomen, but in the three genera treated here, the latter never appears as a separate part; we must consider it as being represented by the genital area and its surroundings, which, however, are not marked by distinct outlines. I beg to give notice that, as a separate abdomen only appears in the genus Stenothocheres (s. below) and in no other genus of the whole family, I shall — for practical reasons —in mentioning and describing all the genera, except Stenothocheres, always both here and in the systematic part use the word »trunk« for the whole body, except the head. The shape of the trunk is mentioned above, for as the head, at least in adult spe- cimens of most species, is very small, I can refer to my description of the body (p. 24—25). The skin — except on the genital area —is very thin, often quite naked, sometimes covered with hairs behind the head, being naked everywhere else, sometimes hairy all over. In several species the trunk is more hairy during the early, not half-developed stages, than when the animals have grown to their full size, so e. g. in Spher. danica, whose young ones are covered all over with peculiar thin, flat hairs, whereas the older specimens are either quite naked or have only a hairy part behind the head. In Spher. Calliopii (pl. IU, fig. 3d and fig. 3g) the trunk has a rather close coat of very peculiar three-branched hairs growing out from tiny knots, the middle hair being longer than the two others. In Spher. irregularis (pl. XIII, fig. 5¢ and fig. 5d) somewhat similar two- and three-branched hairs are seen. In most species the trunk has two pairs of entirely uniform legs, but in a good number of species (as in Spher. microcephala, and in all eight species of the Spheronella living on Cumacea and Isopoda) legs are entirely wanting. The legs are placed on the ventral side, now at some distance within the outline, now at the lateral margin, and as a rule there is no considerable difference in the distances between the first pair and the head, between the first and second pairs and between the second pair and the leg-like caudal stylets. In Homoeoscelis the legs, though small, are comparatively conspicuous, each apparently consisting of a diminutive, short and rather thick basal part, from which proceeds a much longer, very narrow, conical, almost setiform branch and a pair of very short bristles or a short tap as an indication of a second branch. In Spheronella and Choniostoma the legs are nearly always exceedingly small; in recently hatched or young specimens they are as a rule easy to find, but as they do not grow, they are often very difficult to point out in adult animals. Hach leg consists of a small cylindrical joint ending in two short sete. In Spher. longipes (pl. VII, fig. 2a and fig. 2e) only, the legs are somewhat larger, particularly because one of the sete 31 is long. In Spher. Acanthozonis (pl. VII, fig. 5a and fig. 5b) they are reduced to rounded eminences without sete. A genital area is found in all species of these three genera, and it is in some cases much smaller, in others somewhat larger than the head. In its most developed form it is a more or less thickly chitinised plate, which is sometimes nearly circular (Spheronella curtipes, pl. X, fig. 2e), mostly considerably broader than it is long, and not unfrequently with a more or less concave anterior or posterior margin. In this plate we find the genital apertures more or less close to each other, so that the distance between them is nearly always shorter than the length of each; they are usually placed near the posterior margin, seldom in the middle or even nearer the anterior margin. Sometimes the central part of the plate or two rather lateral parts of it are thin-skinned (pl. II, fig. 3b), and in this last case the plate is really reduced to an oval ring with a median longitudinal band. In Spher. Munnopsidis (pl. X, fig. 4c) the plate is more than twice as broad as it is long, and a large inner part of the same shape as the outline is more thin-skinned; the genital apertures are placed transversely and somewhat further from each other than the length of each. In other species the plate is reduced to about two thirds-of a more or less oval, transverse ring, the posterior margin of which is close to the genital apertures, whereas the sides are further removed from them. A further reduction is noticed e. g. in Spher. frontalis (pl. VII, fig. 6i), where the more conspicuous parts consist only of a chitinous arch behind and outside each genital aperture, the two arches yet being connected in the median line. In Spher. micro- cephala (pl. VIII, fig. 2f) the genital area is much longer than it is broad, and the chitinised part of it forms a semi-circle which opens towards the front, its two extremities running forward and forming two rather long, nearly parallel and partly dilated lists. The genital apertures are — as stated above — nearly always closer to each other than the length of each, besides they are curved and placed in an oblique direction, so that their convex sides turn towards each other, and their anterior ends are much closer together than the posterior ones; e.g. in Spher. microcephala, and especially in Spher. Munnopsidis, these apertures are turned so as to be almost or quite transverse; and in Spher. Munnopsidis the distance between them is greater. Hach genital aperture is provided with two chitinous lists, the lips, of which the hindmost one is nearest to the median line and covers the front part of the other lip, when the genital aperture is closed. From the outer lip proceeds a strong muscle outward and obliquely forward, its proximal end being attached to the inner side of the plate or to the ring mentioned above. The contraction of this muscle pulls outward the outer lip, thus opening the genital aperture (pl. XI, fig. 4d). For this purpose the skin close outside the outer lip is always thin (in many figures kept in a grey tint) though the sur- rounding parts may be a pretty hard chitinous plate. In front of each genital aperture, at a shorter or longer distance from it, though always within the genital area, is a very diminutive orifice which forms the entrance to an oval or somewhat elongate vesicle, the receptaculum seminis (pl. I, fig. 3a,r). These two 32 orifices are shown only in some of the illustrations of the genital region, and they are often very difficult to find, if one or each of them has not a spermatophore attached to it; this, however, is rather frequently the case; sometimes we find even two spermatophores, or at least their stalks, on each orifice (pl. XIII, fig. 1e). Such a spermatophore is a globular or ovate vesicle with a stalk twice or three times as long as itself; this stalk — a thin tube — is attached to the skin closing on the above-mentioned orifice, or sometimes — by mistake — outside it (pl. LV, fig. 2c, where we find one spermatophore on each orifice and the stalk of a third one outside it). The two receptacula, when filled, have a strong refraction of light, which as a rule makes them easy to find. Their outlines are traced with dotted lines in some of the illustrations. — In Spher. Munnopsidis I have found in the an- terior part of the plate two holes (pl. X, fig. 4c, k) corresponding to those in Mysidion abyssorum and Aspidoecia (see p. 34—35). In the species which have trunk-legs there is always a pair of caudal stylets shaped somewhat like the legs. Jn Homoeoscelis (pl. II and pl. XIII) they are a little thicker and longer than the legs; in Spheronella and Choniostoma they consist either of a cylindrical, a rounded or a triangular joint terminating in two or three sete (which rather frequently fall off during the preparation); they are sometimes longer, sometimes shorter than the sete of the legs (as e.g. pl. VII, fig. 2e). Im Spheronella Acanthozonis each caudal stylet has one single rather long seta (pl. VII, fig. 5d). In Spher. modesta, which has no trunk-legs, each stylet consists of a rather short, thick joint, from the inner posterior angle of which proceeds an acute »joint« twice as long but scarcely half as thick, which must be considered as a trans- formed seta, and outside it are seen one or two simple sete (pl. [X, fig. 2e). Nearly all the other species which lack trunk-legs are devoid of caudal stylets as well. The place of these stylets varies much; in most species they are situated close together, either on the plate or the ring, a little behind the genital apertures, or close behind the posterior margin of the ring or plate, but in the species belonging to the group of Spheronella Leuckartii they are situated pretty far or very far from each other, and also more or less far behind the genital area (pl. II, fig. 2e and fig. 3b). ‘The remarkable fixation of Spheronella paradoxa will be described in the systematic part; here it may be sufficient to draw attention to it. II. Stenothocheres (pl. 1). This genus, comprising two species, deviates considerably from the three recently mentioned genera, and in at least two important features: — larger trunk-legs with two branches and a separate abdomen — it comes nearer to the less transformed Copepoda. The body is sub-ovate or nearly globular; its abdomen is comparatively rather small and prominent posteriorly on the ventral side or on the hind margin itself. It has no separate head, not even the vestige of a frame (like the one in Choniostoma). The sub-median skeleton is reduced to a plate in front of each maxilliped (pl. I, fig. 1e,h and fig. 2f), and this plate may extend forward like a list between the maxilla and the outside of the rostrum. An- tennule, antenne, rostrum, maxilla and maxillipeds occupy a larger space on the ventral 33 side of the body than in the other genera. The antennule (fig. 1a, a and fig. 2g) are of medium length, without distinct articulation, they have a few rather short sete, among them one olfactory (fig. le, b). The antennz (comp. fig. le and fig. 2f) are placed somewhat obliquely outside and in front of the rostrum, but not on its basal part; they are of medium length, in Sé. egregius (fig. le, c) probably 3-jointed, with a couple of short terminal seta of unequal length; in St. Sarsii (fig. 2f, c) they are weak, with indistinct articulation. The rostrum is on the whole like that in Spheronella, though it must be observed that the mouth-border is very narrow. The maxillule (fig. 1e, e) are on the whole like those of Spheronella, the principal branches rather short or of medium length, the additional branch wanting. The maxillz (fig. la, f; fig. le and fig. 2f) are powerful and do not show any important differences from those species of Spheronella which are parasitic on Amphipoda. The same remark can be applied to the maxillipeds with regard to their structure, but these limbs, compared with the maxilla, are shorter and slenderer than in most species of Spheronella, and we may add that the second and third joints are always fused into one single comparatively short joint, which at most is a little longer than the pointed terminal joint and lacks the spine at the distal inner angle, as the terminal joint lacks a spine inside its apex. The trunk is naked all over (so is the whole body with all its appendages). The trunk-legs are placed differently from those of the preceding genera; both pairs being situated on the ventral side at a good distance within the lateral margin, the first pair (fig. 1a, m) somewhat behind the middle of the body, and the second pair (fig. 1a, n) close in front of the basis of the abdomen. Both pairs, though rather small, are very large compared with those of the preceding genera. Hach leg consists of a peduncle with two branches not distinctly set off by articulation, and as a rule the outer branch is the longest. In the first pair the outer branch terminates in two strong sete of unequal length, in the second one (fig. 1g, u and fig. 2i) each branch apparently consists of two joints, of which the terminal one is somewhat spine-like, but it must be preferred to consider each branch as being composed of one joint with a long and very thick terminal spine. A comparison between the figures 2a and 2d shows that in the same species the abdomen may be found more or less distant from the posterior margin on the ventral side of the trunk, according as the animal is more or less swelled with eggs. The abdomen is not set off from the trunk by an articulation; it consists of a broad, rather stout basal part with arched lateral margin (fig. 1g and fig. 2i), and a narrower terminal part with a more or less deeply incised extremity, which forms two very short and clumsy, badly defined caudal stylets (fig. 1g, t), each with four thick sete. The abdomen seen from below (fig. 1g and fig. 21), presents near the outer margins of the basal part two very long genital apertures (g) in their whole or a considerable part of their length; in the abdomen seen sideways - (fig. 1h), the genital aperture (g) shows its longest extent, and the muscle which opens it (m) is directed towards the dorsal side of the abdomen. Fig. 1h also shows a receptaculum seminis (1) as a large oblong vesicle, placed a little above the abdomen. I have repeatedly 5 34 seen two such receptacula, but I cannot indicate their external orifices, as, strangely enough, I have never found spermatophores on any of the rather numerous specimens I have examined. ILL. Mysidion (pl. XI—X1I). The head is pretty well defined from the trunk, but so feebly chitinised in front and at the sides that the frontal and lateral borders are wanting. The antennulze are much reduced and either 2-jointed or 1-jointed (pl. XII, fig. 2a, and fig. 1a, a). Antenne seem to be wanting. Mouth and maxillule as in Spheronella. The basal joint of the maxilla has at the inner edge one or two processes, and the appendage is a powerful prehensile organ. The basal joint of the maxillipeds has irregular outlines, Trunk-legs and caudal stylets are wanting. There is no genital area: the genital apertures are situated very far from each other (pl. XI, fig. 3b and fig. 3e); each of them has — besides the lips — its own skeleton, consisting of a list which is semi-circular or forms the larger part of a defective oval, the longest diameter of which runs parallel with the median line of the animal, and the opening of which is turned towards this line. The genital aperture is situated close to the posterior part of the list, and the muscles radiate towards its foremost part. The receptaculum seminis — odd, as far as I can see — is situated in the median line, far in front of the genital apertures (pl. XI, fig. 3e). The skin covering it is closely set with many — as many as twenty-six — spermatophores (s), and between them are seen stalks of other vanished spermatophores, some of these sticking together in bulks which cover the skin so completely that, in spite of several attempts, I have been unable to find the entrance or entrances to the receptaculum seminis. In fig. 8e the letter + marks recep- taculum seminis, which on each side opens into an obliquely backward running duct, which I have been able to follow towards the genital aperture (comp. the following genus). In Mysidion abyssorum 1 have found in the semi-circle surrounding the genital aperture a hole (or perhaps rather a spot, covered with a thin membrane pierced with small holes (pl. XII, fig. 2b, k) forming the outlet from a gland which I have found, though I have not been able to examine it more closely, and whose function is incomprehensible to me. — Several parts of the head of this animal are frequently covered with a viscous substance, by which it fastens itself to the marsupium of the host. This substance, in the females as well as in the males and the larve, is probably secreted by glands placed in front of the mouth (comp. the female of the following genus). IV. Aspidoecia (pl. X11). This genus (one species) approaches very near to Mysidion. In this place only its most important characters will be mentioned, an exhaustive description being given in the special part of this work. The body is considerably broader than it is long, the head is distinctly defined from the trunk and pretty well chitinised, with rounded forehead and sides. The front is covered by a large adhesive plate (fig. 3d, s) by which the animal is attached, this plate at the same time covering the 1-jointed antennule (fig. 3h, a). Antenne are wanting; the maxillule are very small, without additional branch; the mouth is normal, but I have found no hairs along the mouth-border; the maxillz are like those in Spheronella; maxillipeds are wanting. Trunk-legs and caudal stylets are wanting. Hach 3D of the genital apertures — as in Mysidion — has its own list, which in this animal forms a ring (fig. 3i); in the front part of this ring we see a rather large hole (k) which serves as opening to a gland (comp. Mysid. abyssoruwm). The genital aperture (g) lies up to the part of the ring which is turned towards the median plane of the animal. Contrary to Mysidion, the distance between the rings varies between being a little greater and very much smaller than the diameter of each. A long way in front of the genital apertures we see two knots a little apart from each other (fig. 3g, r‘, fig. 3e, 0) which show as it were irregular cracks in the thick chitine; no doubt they form the entrances to the recep- taculum seminis (fig. 3e,r) which is odd, much broader than it is long, and at each side bends backward, thus continuing as two almost parallel ducts, which are wide, at the middle sowewhat narrowed, and run to the genital apertures. No spermatophores have been found. b. The Male. Out of my forty-three species I know the males of thirty-two, viz. of all species of the genera Stenothocheres, Homocoscelis, Mysidion and Aspidoecia, as well as of twenty-five out of the thirty-four species of Spheronella. So in nine species of Spheronella and in the two species of Choniostoma the males are still unknown. With respect to the latter genus particularly the gap is keenly felt. The male is always much smaller than the adult female — as a rule quite dispro- portionally so. In three species only: Spher. frontalis, S. decorata and S. curtipes, its length exceeds 4/2 mm. The largest male I know belongs to S. curtipes and measures ‘92 mm. in length and °/s of this size in breadth. In most species the length is about +/4—*/smm., and the length somewhat _exceeds the breadth. The smallest normal males I have found in Aspidoecia Normani, two specimens of which were respectively 147 and 138 mm. long. In Mysidion abyssorum the normal male seems to be about *164 mm. long, though I have found two perfectly dwarfish specimens, of which one was ‘099 mm. long — and this is one of the smallest of adult Copepods hitherto discovered, however, it may be that these two specimens were recently hatched and had not grown to their full size (s. below under »post- larval development«). The genus Homoeoscelis shows least difference between the male and the female, especially H. mediterranea, of which the largest female was °31 mm. long, 32mm. broad and rather flat, whereas the male was 174mm. long, ‘096 mm. broad and equally thick. The greatest difference between the sexes I have found in Spher. microcephala, in which the largest female was 1°-44mm. long, of the same breadth and almost perfectly globular, whereas the male without counting the rostrum was 18mm. long, ‘15 mm. broad and ab. ‘11mm. thick, which gives a volume of between 800—1000 times smaller than the female. The body, as a rule, is somewhat longer than broad, and seen from below, varying from sub-globular to an elongate oval, seen sideways, the back is strongly vaulted, the 5 36 ventral side almost flat, and the animal is nearly always somewhat broader than it is thick. In the genus Homoeoscelis (pl. If and pl. XIII) as in Spher. curtipes (pl. X, fig. 2f and fig. 2¢) the body is much longer than it is broad, the length varying from 17/3 to a little more than twice the breadth; besides, in S. cwrtipes it is curved, so that, seen from the side, it presents a moderately concave ventral outline. The length of the head varies between a little more and a little less than half of the total length, in Spher. frontalis (pl. VII, fig. 6a—6b) it does not take up a third part. The body is usually broadest somewhat behind or almost on a line with the base of the maxillipeds. In Spher. modesta (pl. LX, fig. 2f) the greatest breadth lies before the middle of the maxillee; in several species it is rather far behind the head, and in this case the trunk is somewhat or much larger than the head, viz. in Spher. Bonniert (pl. VIL, fig. 1a), in S. frontalis (pl. VI, fig. 6a) and in Mysidion abyssorum (pl. XII, fig. 2e). A distinct abdomen is found only in the genus Sfenothocheres. Antennule, antenne, rostrum, maxillule, maxilla and maxillipeds are much like those of the female, still we find a number of minor differences which must be mentioned. Trunk-legs and caudal stylets are well developed in all the species whose females possess these organs, though as a rule they differ very much in the two sexes, and they are also found in a few species, as Spher. microcephala and S. modesta, whose females lack both trunk-legs and caudal stylets, or only trunk-legs. In Spher. dispar, S. insignis, S. marginata, as well as in the genera Mysidion and Aspidoecia, the males have not the slightest rudiment either of trunk-legs or of caudal stylets. In the genus Stenothocheres the antennulz, the antenne, the rostrum and the mouth- appendages are situated on the foremost rather flatly vaulted part of the ventral side of the body. In all the other genera the surrounding of the rostrum and the mouth-appendages in front and at the sides lie more or less deep and are limited anteriorly by an outstanding border, which is frequently rather high or forms a slanting plate, at the sides by very conspicuous lateral borders, which usually run nearly parallel from the front towards the base of the maxilla, whence they curve more or less outward towards the lateral margins of the animal and vanish somewhat behind the base of the maxillipeds. The shape of these lateral borders is rather variable and difficult to describe, but the numerous illustrations will show two outstanding rounded plates bending like a cape towards the base of the maxilla, and outside these protruding borders we can always see something of the slanting lateral surfaces of the head, when looking at the animal from the ventral side. The frontal border is sometimes distinctly separated from the lateral ones, but it usually forms a direct or nearly direct continuation of them; in most forms it is evenly curved; in Spher. elegantula (pl. II, fig. 2f) and in kindred species it has a deep incision on each side, by which the frontal plate is divided into a large, median, almost square part and two much smaller rounded lateral lobes. In Spher. Calliopii (pl. III, fig. sh) the frontal margin is divided into six lobes, in S. decorata (pl. VIII, fig. 3f) the frontal plate is much elongated, with the anterior end cut off transversely and with several incisions, one in the middle and two on each side, by which it is divided into four square and two low triangular lobes, all of which. 37 bear small spiniform processes on their outer margin. In Spher. dispar (pl. UX, fig. 3h) and in Spher. insignis (pl. X, fig. 1b and fig. 1c) the frontal plate expands in a most peculiar way. Seen from below, the head tapers very much towards the front, whereupon it dilates to an almost circular plate or a transverse oval with acute lateral angles; the sides and the front margin of the circular plate and the front margin of the oval project into a series of closely situated spiniform little processes; on the ventral side of the plate we see a chitinous ring, from the inner edge of which four processes run towards the centre; in S. dispar these processes do not meet, but form the surroundings of a cross-shaped space, whereas in S. insig- nis they meet in the centre, thus forming a cross, and dividing the space into four parts, (which perhaps are pierced with small holes). The purpose of this peculiar ring is un- known to me. The part surrounded by the protruding frontal and lateral borders frequently lies very deep, and where the lateral borders are high, sometimes, as in Spher. curtipes (pl .X, fig. 2f and fig.2¢), they hide the rostrum and the maxilla, if the animal is seen from the side; as a rule a lateral view shows these organs in almost their whole, or in half of their length. Several males of the same species may show great individual difference in this point, whereas on the other hand, there are species, as e. g. Spher. curtipes, in which the rostrum and the maxille are always partly invisible because of the height of the lateral borders, in other species again, as in Spher. microcephala (pl. VIII, fig. 2 g and fig. 2h), the borders are so low, that the above-mentioned organs are always visible in nearly their whole length. The sub-median skeleton, which is found in all females and has been described as far as this sex is concerned, is also seen in all males, and in most respects shows a similar structure, but in most species of the genus Spheronella it is produced into free processes. Three pairs of such processes may be found. ‘Those of the first pair are usually rather short and broad, sometimes rounded, in S. microcephala (pl. VIII, fig. 2 g and fig. 2 h, i) pretty long, slender and pointed, being situated behind or below the basis of the maxille; some- times, as in S. elegantula (pl. II, fig. 2f and fig. 2g), they extend backward over the basal part of the maxillipeds. The second pair of processes are found most frequently, and may become much longer than any of the other two pairs; they proceed at a shorter or longer distance from each other between the maxillipeds, and are sometimes parallel, sometimes or mostly diverging. The third pair appears in very few species only, as in S. paradoxa (pl. III, fig. 4h and fig. 41), in S. Metope and in S. Holbolli (both on pl. V.); they proceed between and a little behind the second pair and are much shorter than these ones. In the systematic description of the species I use the terms: first, second and third pair, in speaking of these processes. The Antennule are found in all species; in Stenothocheres they are situated in front of and outside the antenne, in all other forms they are found on the lateral margins of the head, where these merge into the frontal margin. They are constructed much like those of 38 the females, e.g. where the antennule of the females are well developed and 3-jointed, those of the males are equally so, and where they are reduced in size or in number of joints, we generally — though not always (Spher. microcephala, Mysidion commune) — find a similar reduction in the male. In Spher. microcephala (pl. VIL) the antennulz of the female are very short and 1-jointed, those of the male long and likewise without distinct articulation. In Mysidion commune (pl. XI and pl. XII) those of the female are very short and 1-jointed, those of the male short and 2-jointed. Their bristles are much the same as in the female, and they have frequently a rather conspicuous olfactory seta. The greatest reduction is noticed in Aspidoecia (pl. XII, fig, 3k and fig. 31), the antennule of which consists of a very short, naked joint terminating in an olfactory seta which is several times longer than the antennula. The Antenne are altogether so like those of the female that a special description of them is superfluous; where they are wanting or reduced in the female, they are equally so in the male. The Rostrum is very like that of the female, but seems now and then to be longer and more slender; in Spher. microcephala it is much longer than that of the female (pl. VIII). In several species, e. g. in Spher. modesta (pl. LX), the mouth-border is much broader, its hairs at least are considerably longer in the male than in the female. The Maxillule are always found and are constructed like those of the female. In dissecting the head of a male of Spher. frontalis I found that, in addition to the two long — though unequally long — principal branches and the long additional branch, the maxillula of this animal possesses a fourth shorter one (pl. VII, fig. 6d), which proceeds within the base of the anterior principal branch. As this supplementary branch is not likely to be discovered without undertaking a dissection expressly for this purpose, I cannot tell whether it is found in the female as well; without dissecting the rostrum I have looked for it in vain in the female as well as in both sexes of other species of Spheronella. Fig. 41 in pl. IL. will give a good idea of the rostrum with antenne and maxillule in a male which possesses all these organs in their typical form. The Mazxille are well developed in nearly all species and are in the main constructed like those of the females; sometimes they are somewhat smaller, sometimes rather larger, in Aspidoecia (pl. XII) much larger. Occasionally we find differences of detail in the two sexes; in Spher. capensis the basal joint of the male has at its distal end, where the inner and the posterior side meet, a rather high, prominent plate, the margin of which runs out into spiniform processes (pl. VI, fig. 1 ¢ and fig. 1 d); this excrescence is wanting in the female. In the male of Spher. insignis the posterior side of the basal joint is provided with a conside- rable number of peculiar processes (pl. X, fig. 1c) which the female lacks. In Spher. margi- nata (pl. XIII) whose female has rudimentary maxille, these organs in the male have about the same general shape and size as in kindred species, but all three joints are fused together. The Mazillipeds ave well developed in all species and essentially like those of the females, though in the male the basal joint is frequently provided with hairs, and also some- 39 times with processes, whereas in the female there are no processes and fewer or no hairs etc. In Mysidion commune the male possesses on the outer side of the basal joint a knot- like excrescence and a process of very considerable size (pl. XI, fig. 3g and fig. 3h), and its second joint has a conspicuous process, all of which are wanting in the female. In Aspi- doecia (pl. XII) the female lacks maxillipeds; in the male, though somewhat smaller than usual, they are well developed: the basal joint is much as in Stenothocheres, but the second joint, which as in several other forms consists of two completely fused joints, is exceedingly short, somewhat shorter than the terminal joint. Trunk-legs and caudal stylets. The occurrence of these organs is mentioned above on p. 36. In Homoeoscelis (pl. II and pl. XIII) there is but a slight difference between the two similarly shaped pairs of trunk-legs and the caudal stylets, and as both are like those of the female, they do not require further mention; we shall only add that they are sometimes rather larger than, sometimes of the same size as those in the other sex. In Stenotho- cheres (pl. I) the trunk-legs are very similar to those of the female and of almost equal size, however, as the trunk of the female is large, that of the male small, of course the legs of the latter are much more conspicuous and appear larger. In the male the basal part of the legs stands more out from the body, and the longer robust terminal spine on the outer branch is longer than in the female; the other differences are insignificant. Thus, in Stenothocheres Sarsii the spine on the inner branch of the second pair of legs is curved like a hook, in Sten. egregius it is less curved, and in the illustrated specimen the right and the left spine curve differently. Behind and above the basis of the second pair of legs both species show two considerable spines which are situated close together on a small projection; I should think they might possibly be considered as rudiments of a third pair of legs, but [ do not presume to have any definite opinion about the matter: The abdomen is small, with short, distinct, rather broad stylets, not set off by an articulation, and each provided with four spines, of which the two innermost are the longest and thickest. In Spheronella, as a rule, there is a great difference between the two pairs of trunk- legs and between each of these and the caudal stylets. The shape and size of the trunk- legs vary much according to the species, and the appearance of the legs as well as of the stylets presents very great variation on account of the very different length of their terminal sete. In this genus the first pair of legs usually originates outside and behind the basis of the maxillipeds, nearly in the middle between the laterai margins and the median line of the trunk, at the bottom of a pretty broad transverse depression. Each leg consists of a basal part, a peduncle, differing much in length and breadth (sometimes, as in Spher. intermedia (pl. V, fig. 3f and fig. 3g), this part in exceedingly large), and of two branches, one of which is generally longer than the other, sometimes the one is wanting altogether. As a rule, none of the branches are articulated on the peduncle, and frequently one or both of them terminate in sete, one of which is longer than the others, the one on the outer branch often exceedingly lony, occasionally half as long as the whole animal. The second pair of 40 legs is situated at or behind the middle of the trunk, and if seen from below, near or on the lateral margin; as a rule it is shorter than the first pair, the peduncle is not much thicker than the rather short inner branch, whereas the outer branch — if there is one — is triangular and often ends in a short seta. The terminal sete of the inner branch vary as much in length as those of the outer branches of the first pair of legs, and one of them can attain to half the length of the animal. The caudal stylets are usually situated rather close to each other, most frequently near the posterior end of the body, sometimes very much to the front about at the middle of the trunk, especially where it is very large, as in Spher. Bonniert and in S. frontalis (pl. VII); in the latter species they even appear in front of the second pair of legs. Both stylets as a rule are rather short, sometimes very short, nearly cylindrical; one terminal seta on each stylet is often rather or very long, occasionally excee- ding in length the above-mentioned long sete on the trunk-legs. — Several species of Spheronella deviate now in one, now in more respects, from this description which is based on the main bulk of the species. In Spher. modesta (pl. IX, fig. 2f and fig. 2h) the first, and especiaily the second pair of legs are considerably reduced in size etc., whereas the caudal stylets. are comparatively large and constructed like those of the female (see above). In S. decorata (pl. VIII, fig. 3f and fig. 3g) the two pairs of trunk-legs are very much alike, each lege consisting of a short basal part with two short branches, and each branch ending in a thick, but rather short seta; caudal stylets are wanting. Concerning S. microce- phala (pl. VIII, fig. 2g and 2h) which deviates considerably and is difficult to understand, I refer to my description in the systematic part of this work. In S. curtipes (pl. X, fig. 2f and fig. 2g) both pairs of legs are very small, slender and 2-jointed, the caudal stylets are quite minute and situated far from each other on the posterior margin of a long and broad, but not much projecting eminence, the posterior angles of which are decorated with peculiar rounded processes and knots (might the whole formation possibly be a reduced abdomen ?). It has been mentioned already that trunk-legs and caudal stylets have disappeared altogether in several species of Spheronella as well as in Mysidion and Aspidoecia. In the species belonging to Stenothocheres and Aspidoecia the. body of the male is quite naked, whereas all the other species are more or less clothed with hairs. In not a few species the frontal margin is furnished with very short hairs or with fine spiniform processes (the species of Spheronella which live as parasites on Cumacea). In nearly all species the margin of the lateral borders of the head are trimmed with a series or a stripe of hairs usually of medium length, which as a rule extends towards the front in a curve round the base of the antennule, ending just in front of it; posteriorly it follows the lateral margin up to its curved extremity, whence (or a little in front of it) the stripe continues across the sides and the back, now straight on, now curving or in a broken line, now advan- cing, now receding obliquely. This line I consider as forming the boundary between the head and the trunk, its hairs being now very long (pl. III, fig. 3i; pl. VI, fig. 1d), now of the same length as those which cover the sides of the trunk. Sometimes (e. g. pl. LV, fig. 3 h) 41 we see instead of a single line a wider or narrower band of hair going from the posterior extremity of the lateral margin up across the back, and behind this line or band is a larger or smaller transverse naked area, whereas the other parts of the trunk: the back, the sides, the posterior extremity and the ventral surface, are closely covered with shorter or longer hairs, with the exception of a transverse band in front between the first pair of legs — and frequently their surroundings — which is naked. These hairs are usually simple; in Spher. frontalis (pl. VI, fig. 6a. fig. 6b and fig. 6h) and in Mysidion commune (pl. XI, fig. 3g and fig. 3h) the trunk is closely covered all over with transverse minute knots, each of which bears several (in S. frontalis at least many of them ten) fine hairs. In Spher. Giardii the trunk is covered with 2- or 3-branched hairs similar to those mentioned above in the female of S. Calliopti and in S. irregularis. In the male of S. Calliopii (pl. IL, fig. 3h and fig. 3 i) the hairs of the boundary line between the head and the trunk are particularly long, whereas the dorsal surface and the posterior extremity are covered with fine dots resembling the roots of hairs, though I have been unable to find any hairs, and across the back to the exterior angles of the first pair of legs we find a narrow, naked band; the ventral surface behind the caudal stylets is provided with ordinary hairs. In Spher. microce- phala (pl. VII, fig. 2g and fig. 2h) the hair-covering is less developed than in any of the other species, as only the hindmost part of the lateral borders of the head and the ventral surface of the trunk are covered with hairs, the other parts of the body being naked. Ina few species we find hairs in front of the base of the maxillipeds, and in Spher. chinensis (pl. IIL., fig. 1a) and kindred species there is a bunch or a short band of hairs outside the base of the maxillule. As to the internal structure of the male I confine myself to the following observa- tions. In a well-preserved specimen we usually perceive through the transparent skin two larger or smaller globular bodies in the middle of the trunk or somewhat more to the front; in Stenothocheres they are situated close to the front of the abdomen behind the base of the second pair of legs. I will call these bodies spermatothece, though I cannot make out whether they have really the function of such organs, or whether they are the testicles them- selves. I have illustrated them in several forms, as Spher. paradoxa (pl. II, fig. 4 h, q), S. capensis (pl. VI, fig. 1c), S. Bonnier (pl. VU, fig. 1a, q), S. frontalis (pl. VII, fig. 6 a, q) and Mysidion commune (pl. XI, fig. 3 g,q); in this last species the spermatothece are particularly large and obliquely situated (probably a case of anomaly or of accidental pressure in the figured specimen, for in the next species: Mysidion abyssorum (pl. XII, fig. 2 ¢, q) they are normal). In Aspidoecia Normani the specimen illustrated (pl. XII, fig. 3k) showed a single, but very large spermatotheca (q), but in a couple of other specimens I saw two considerably smaller and normally situated spermatothece. In Spher. paradoxa I succeeded in finding the genital apertures very close to each other on the posterior wall of the depression which runs across the front part of the trunk on its ventral surface. From each spermato- theca a rather short duct goes forward and obliquely towards the median line to its aperture. 6 42 ; In Homoeoscelis minuta (pl. II, fig. 11 and fig. 1k) I have found the rather small spermato- thece (though they are not illustrated) close together in the line between the hindmost pair of trunk-legs, and I think I have found the two genital apertures in close proximity in the posterior wall of the ventral depression — which is particularly conspicuous in fig. 1k — in a line between the first pair of legs. In a number of species belonging to Spheronella I have found a most peculiar struc- ture beneath the larger part of the skin of the head at its back and sides. It appeared most distinctly in S. paradowa, where I saw very plainly beneath the skin a single layer of rather large hollow spaces; fig. 4k in pl. III is drawn to the same scale of enlargement as fig. 4i and shows the skin and two rows of the afore-mentioned hollow spaces beside each other. In S. Metope (pl. IV, fig. 31) the spaces were filled and appeared in outline as shown in the illustration. The males are sometimes hinged on the females, but much more frequently on the gills or on the marsupial plates of the host by a thread which proceeds from the median line of the front close in advance of the rostrum. ‘This thread is secreted by a gland or glands and can presumably be produced by the males of all species. The shortest thread I found in Homoeoscelis minuta, in two specimens, in one of which its length was similar to that of the first joint of the maxilliped, in the other somewhat shorter. In e. g. Spher. para doxa (pl. III, fig. 4h,s) the thread is about #/s of the length of the animal, in Stenothocheres Sarsivi (pl. I, fig. 2k, s) a little shorter and in S. abyssi (pl. IV, fig. 2d) even a little longer than the whole animal. I found the longest thread in a specimen of Aspidoecia Normani, Where it was between twice and three times as long as the animal, whereas in the spe- cimen illustrated in pl. XII, fig. 3k it was scarcely half as long as the male. This last instance shows that the length of the thread can vary very much in the same species, but this is not usually the case, as in some species a shorter, in others a long thread is always found. In all the above-mentioned and in several other species the thread is always simple and cylindrical, generally a little dilated towards the distal end by which it attaches itself, - and not unfrequently the end is expanded into a disk. Deviating forms of this thread are met with in the species of the genus Mysidion, and especially in the species which I have placed together below under the heading of the Spheronella Leuckartii-group. In Mysidion the proximal part of the thread is simple, the distal part appears in two varieties; either, as in Mysidion abyssorum (pl. XII, fig. 2g), it shows two considerable fusiform expansions, the middle parts of which are each surrounded by a peculiar collar-shaped ring, or, as in Mysidion commune (pl. XI, fig. 3h,s), the apical part is very thick and above it the thread dilates still more and becomes fusiform; its widest part has a collar-like ring, and a similar ring surrounds it somewhat higher up, where the thread is only half as wide. In the species belonging to the Spher. Leuckartii-group we often find the male hinged by a thread which varies in extent between nearly half and almost the full length of the animal, and is constructed in the following way: it is divided into two parts, either of equal length, or 43 the proximal part longer than the distal one, the former ending in a thick, bell-shaped and thick-skinned, hairy expansion (pl. I, fig. 4b); the distal part comes out of the bell in which its extremity forms a little ball; its other extremity expands into a disk which is glued on to the Amphipod. I have found several of such threads and examined them as carefully as possible, but it is quite incomprehensible to me how the animal has been capable of producing them. ec. The Ovisacs and the Development of the Eggs. 1. The Ovisacs. Of forty-one species the ovisacs have been found, and only in two species of Spheronella they are wnknown to me. In the two species of the genus Stenothocheres they differ so much from those met with in the other genera, that I prefer to leave out this genus for the present, setting it aside for separate treatment. In Homoeoscelis, Spheronella, Choniostoma, Mysidion and Aspidoecia each female deposes several — no doubt at least four or five — or many ovisacs, which, if not deformed by pressure, are sub-globular, oval or, in Mysidion, of a short pyriform shape. In Homoeo- scelis minuta, of which I have examined a large material, I can assert that the female deposes at most eight ovisacs, though usually but five to seven are found; in Choniostoma the maximum seems to be twelve, in Aspidoecia thirteen to fourteen, in Mysidion seventeen or still more. Im the numerous species of Spheronella which live in the marsupium of Amphipoda, I cannot indicate the maximum of the ovisacs, partly because my material of each particular species is too small, or because not unfrequently a couple or more of females are lodged in the same marsupium, partly because, in many cases, one cannot be certain that some of the sacs have not been washed away. Better information can be given about some species living in the marsupium of [sopoda and Cumacea: in Spher. Munnopsidis I found one female with twenty ovisacs, in S. decorata the same number, in one specimen of S. modesta twenty-two, in another twenty-eight ovisacs, all laid by one single female. This latter number may be supposed to be about the maximum, not only in the above- mentioned species, but in the whole family. It is very difficult to indicate the smallest number of ovisacs made by normal females of the different species, as, for one thing, it has to be ascertained, whether the specimen in hand has altogether finished laying eggs, and a considerable material has to be examined for this purpose alone; still, though I have not done this, I think I can say that the number is never less than four or five, perhaps seldom less than five or six. In all five genera each ovisac is smoothly rounded, its eggs being as usual enclosed in a common membrane. In Homoeoscelis, Spheronella and Choniostoma all ovisacs are deposed freely without being attached to the female or having any real connection with each other. Indeed, we see rather frequently some, or many, of the ovisacs sticking together, or one, or several of them, adhering somewhere to the body of the female; however, this kind of adhesion is of a secondary, quite unimportant nature, G? 44 and is certainly owing to the fact that the membrane surrounding the ovisac was not sufficiently stiffened, when — or shortly after — the ovisac was laid. In DMysidion and Aspidoecia all ovisacs, from the moment they are laid till the larve have swum out, are hinged to the lips of the genital apertures by stalks, which are rather short (in Mysidion), or very short (in Aspidoecia), so that we see pretty frequently six to seven, or sometimes more ovisacs hinged at one genital aperture (pl. XI, fig. 3b); fig. 2b in pl. XII shows how the lips of the genital aperture and the part behind them are covered by a plate formed of a coagulated viscous substance, from which the stalks of the ovisacs proceed; the plate must be considered as the coalescent basal parts of the stalks. The genus Stenothocheres deviates considerably from the other genera, but unfor- tunately I am not prepared to represent its conditions as precisely as I should like to do. Of one of its two species I have seen but one single female with eggs, of the other (S¢. egregius) wy material is indeed very abundant, but not particularly good, some of it being very old, and most of it, though of later date, having shrunk somewhat, because its hosts — while still alive — had been put into too strong spirit. In many cases I only found a single lump of eggs, which was rather larger, or considerably larger than the female, had no regular form, and was not surrounded by a common membrane. Sometimes, but not always, this lump seems to consist of two — seldom three — smaller coalescent lumps; four instances were of a more instructive nature. A female of St. Sarsii showed two lumps of about equal size, one of which was free, the other (no doubt accidentally) adhering to the female. One of them is illustrated in pl. I, fig. 2c, which shows the irregular shape and want of a common membrane, as well as the size in proportion to the largest female which is illustrated in fig. 2a and magnified to the same scale. The three other cases have been observed in Sf. egregius; in all of them the female had doubtless finished laying eggs. In one case two smaller, short oval lumps were glued together at their extremities, one con- taining seven, the other nine eggs; in the second there were two lumps, one of them a little larger, the other a little smaller than the female; in the third case there was an oblong lump containing ten eggs, the young animals of which were a little more than half- developed, six to seven larve which were about to break out of the shells, and thirteen free larve; all this indicated that the eggs had been laid at intervals. It seems probable, on the whole, that the eggs are not laid all at the same time, but successively, though the intervals must be rather short, whereas the ovisacs, at least in most and probably in all the other genera, are deposed at rather considerable intervals; this is easily seen from the fact that among ovisacs deposed by the same female, we often find one or two which have evidently been laid recently, whereas some others contain more or less developed larve. We sum up our observations in the following statement, that in Stenothocheres the eggs are deposed in one single large and free lump, or in a couple of smaller and free lumps of irregular form and without the common membrane which belongs to a proper ovisac; and finally, it seems rather probable that the eggs are not laid all at the same time. 45 We have now to deal with several other questions, some of which are difficult to answer definitely, namely: the size of the ovisacs compared with that of the females in the different species, their relative size in females of the same species and of different species, the number and size of the eggs in the ovisacs of the different species, and the fertility of each species. Here, however, we at once meet two difficulties: the one mentioned above on p. 24, that shortly before her laying eggs the female is always somewhat (and no doubt usually much) larger than after it, and in most cases it is quite impossible to procure spe- cimens which are going to lay eggs, such as have half done and such as have quite done laying eggs; most frequently one only finds from two to five specimens altogether, all of which have half done or quite done, or else one or two of them are not full-erown; besides, an ovisac is somewhat smaller when it has just been laid, than later on, when the larve break out of it, for durmg the development the ovisac increases somewhat in size, getting at the same time less firm, as each egg, which is always globular or polyhedrous at the beginning, becomes elongated. Making allowance for this fact, it is seen that in most species there is not usually much difference between the sizes of the ovisacs deposed by the same female, whereas in some species the ovisacs often, though not always, differ very much in dimension (pl. X, fig. 4a; pl. XI, fig. 3c). -There is a great difference, on the other hand, between the average size of the ovisacs compared with the adult females of each species; comp. e. g. the proportion between fig. 3a and fig. 3c in pl. ILI with that between fig. 4a and fig. 4d in the same pl. Of course, we may say that as a rule the ovisacs are comparatively smaller in the species which depose a very large number of them than in those which lay rather few, yet even in these the ovisacs sometimes do not exceed middle size. In the species whose females are large, as a matter of course, the ovisacs are much larger than in the small species. The number of eggs contained in the ovisacs naturally depends on the size both of the eggs and of the sacs. It is true, I have not measured the eggs of various forms, but as I know the larve that come out of the eggs in more than half of my species, as the length of these larve varies between about ‘15 mm. and °30 mm. only, and as they show no relatively great differences either of breadth or thickness, I possess a pretty accurate standard for judging the relative size of the eggs, for, evidently, the largest eggs (judging from the larve: those of Spher. decorata and of the genus Choniostoma) cannot be much more than double the diameter of the smallest (in the genus Homoeoscelis). The further result is, that in the species whose females are very small, as Stenothocheres egregius (pl. 1), Homoeoscelis minuta (pl. IL), and especially in Hom. mediterranea (pl. XIII), the eggs must by very large compared with the females, whereas the eggs must be proportionally small where the females are very large, as in Spher. Calliopii (pl. III) and in the two species of Choniostoma (pl. X and p]. XI). That these statements agree with facts appears very clearly from the illustrations of the eggs and females of the above-mentioned species; — it must be borne 46 in mind that the female and eggs of the same species are always illustrated enlarged on the same scale. As the difference of size between the very small species (or rather those whose adult females are very small), as Homoeoscelis minuta and H. mediterranea, and the very large ones, as Choniostoma mirabile and Ch. Hansenii, is exceedingly great (s. above p. 24), and the difference between the eggs not being greater as just stated, it follows that, with equal proportion between fertility and volume, the large species lays manifold more eggs than the small one. To abide by our example: the number of ovisacs in Homoeoscelis amounting to about eight, in Choniostoma to about twelve, it is evident that in the small species we find few, in the large ones numerous eggs in each ovisac, and this fact is indeed proved by the following figures: in an ovisac of Homoeoscelis minuta are found only about 14 to 18 eggs, in H. mediterranea no more than about 6 to 10, whereas in a middle-sized ovisac of Choniostoma mirabile I have counted 1057 eggs. If, in a smaller species, as e. g. Spher. dispar and S. modesta, the number of ovisacs increases to about twenty, or, as in the latter species, to about twenty-eight, the quantity of eggs contained in an ovisac is naturally rather small (pl. LX, fig. 3e, fig. 3c and fig. 2c), whereas‘in the gigantic Spher. Munnopsidis, of which species one specimen — consisting only of a half-emptied skin — was about 5mm. in diameter, the number of ovisacs may indeed amount to twenty, still the average number of eggs in each ovisac (in this species I have found great variety in the size of ovisacs of the same specimen) is nevertheless very great, as is shown quite distinctly in fig.4a on pl. X. The entire bulk of eggs deposed by a female — as stated above — is always larger than the animal itself after it has laid them, and it is often so marvellously large, compared with the female, that we hardly understand the possibility of it (s. pl. XI, fig. 3a). This state of things, however, may be partly explained by the fact that the ovisacs are deposed at certain intervals (about a possible deviation in Stenothocheres, see above), and that conse- ~ quently the eggs can be gradually developed in the female. If we find seven or eight ovisacs in a female, the development of at least one or two of these is nearly always so far advanced, that the larve are in the Nauwplius stage; where ten or eleven, ‘or. still more, ovisacs are found, one or two of these usually contain almost or quite full-grown larve. The length of time which elapses between the laying of the first and of the last ovisac im specimens containing a large number of these sacs, as Spher. decorata and the other species living in the marsupium of Cumacea, seems to be about equal to that which the first laid ovisac requires for its development: the division of the germ, the Nawplius stage, and the development of the larva with numerous limbs, though I cannot tell how many days are required for this process. The two species which lay the smallest number of eggs are the diminutive forms Homocoscelis mediterranea and Stenothocheres egregius, the former has as many as eight ovisacs containing in all 60 to 70 eggs, whereas St. egregius, as a rule, only lays about 30 eggs (I have found between 16 and 42 eggs, the latter number in an exceptionally large specimen). The largest species, AT viz. Spher. Munnopsidis and the two species of Choniostoma, lay the largest number of eggs. In a specimen of Ch. mirabile with eleven ovisacs I counted the eggs in one of these presumably of middle size, and I found 1057 eggs, so the number of eggs laid by this specimen may be said to amount to 11,620, and if we take this figure as the normal quantity in Ch. mirabile of the Kara Sea, at any rate we do not exaggerate. In Spher. Munnopsidis the number of eggs seems to be even much higher, however, it would scarcely be possible to calculate the exact amount. Between these last-mentioned species and Stenothocheres egregius the other species present a variety of transitions, as far as fertility is concerned. In a following chapter about distribution etc. I shall have an opportunity of making some further observations on these rather remarkable differences. Il. The development of the eggs. As for the division of the germ and the earlier part of the embryological development, which I have not studied myself, I shall refer to the representation of Salensky. As in all Copepoda a Nauplius stage is developed (pl. XI, fig. 1¢ and fig. 1d); but this stage never becomes free, it evolves itself into the stage of a highly organised larva, of which a detailed description is given below. When this larva, which corresponds with the first Cyclops stage in other parasitic Copepoda, is full-grown, it breaks out of the egg-membrane and of the ovisac. As for the details concerning the development of the Nauplius stage and of the larva, I must again refer to Salensky. It may be added that in material preserved in spirit (and I have seen no other) the ovisacs, when younger, are of a light yellowish colour, but they gradually get whiter, as the larve are developed. d. The free Larva. I. The Material. Of several species 1 have found free larve, which were either swimning out of, or had recently swum into the marsupium, and these specimens, of course, were excellent, showing the normal shape of the larve. Of a number of species I have procured a rich material of larva by pulling them out of an ovisac; they were good enough When taken while about to break out of the egg-membrane, though the body might be somewhat soft and not extended in its full length, thus showing a vaguer outline which did not quite correspond with that of the swimming specimens. Sometimes I had to content myself with younger animals, which had to be pulled out of their egg-membranes, and which had indeed a well-developed mouth, maxillz, maxillipeds etc., but whose cephalothorax was decidedly shorter and stouter than those of the full-grown larve, and which also showed other signs of unfinished development, so that no reliable observations could be made of difficult parts, as e.g. the branches of the maxillule. Finally, of a few species I had only larve which had swum into the marsupium of a new host, where they had attached them- selves (9: below), and in these the cephalothorax, as a rule, was shorter and broader than in the free specimens. Of some species I had larve of this kind as well as of others, 48 that had been taken in a free state, or had been pulled out of an ovisac. All this put together gives the result, that I have been able to examine the larvee of twenty-three species, lacking only the larva of Stenothocheres Sarsii and nineteen species of Spheronella; thus I possess the larva of Sten. egregius, of fifteen species of Spheronella and of all seven species of the four remaining genera. Fortunately the fifteen species of Spheronella represent nearly all the more important types of this large genus. As for the illustrations, I beg to notice that I have frequently omitted the two pairs of natatory legs, or at least their branches, as their representation, as a rule, would have been exceedingly difficult, and the omission is of little consequence, as the number and the arrangement of the natatory hairs are very much alike in the different species. In some cases the abdomen is also left out. My representations of the maxillule include all that a careful study enabled me to observe; however, I am inclined to think that a better material would sometimes have allowed me to discover one — occasionally two — more branches. IL. Structure of the Larve. The length of the body usually varies between 20 and *25mm.; the longest larva I found belonged to Spher. decorata, and it is ‘30mm. long, the shortest, 15 mm. in length, belongs to Homoeoscelis minuta. The body is divided into two parts: the cephalothorax and the abdomen. The cephalothorax is somewhat depressed, usually oval and about I'/2 time as long as it is broad, sometimes (Mysidion, pl. XII, fig. 2h) more elongated, almost double as long as it is broad; it consists of two divisions, namely, the cephalothorax properly speaking, and a single trunk-segment (pl. III, fig. 3k), which are joined by a rather sinuate articulation, whereas the trunk-segment is between five and eight times (in a single case about eleven or twelve times) shorter than the anterior division. I have found behind the segment mentioned a very short portion which looked like the rudiment of a second free segment (pl. III, fig. 3k) and belonged to the cephalothorax, not to the abdomen; I cannot, however, say anything definite about this part and will content myself with stating what I have observed. The abdomen is narrow, and its length varies between a little more than one sixth and rather more than one third of that of the cephalo- thorax; it always consists of three distinctly separated segments and has two caudal stylets, which as a rule are plainly articulated on the third segment, but sometimes are coalescent with it (e.g. pl. I, fig. 11). The foremost half of the large anterior division of the cephalothorax is always provided with antennulz, antennz, rostrum with mandibles, maxillule, maxille and maxillipeds; the hindmost half has a longer or shorter odd pouch, which turns backward, decreasing in width towards its distal end (pl. I. fig. 11, 1); its posterior part forms a free, either pointed or rounded bag along the ventral surface, often covering the transversal band which unites the first pair of natatory legs (pl. III, fig. 31), sometimes even the band between the second pair of natatory legs (pl. II, fig. 11). The first pair of natatory legs is situated at the posterior extremity of the first division of the cephalothorax, whereas the second pair proceeds from the free trunk-segment. It may be mentioned finally, that Salensky has 49 shown the existence of an eye (op. cit. p. 314, Taf. X, fig. 21 and 23, 0e), which in the illu- stration is drawn as situated on the ventral side of the forehead »in Form von zweien am oberen Theile etwas verdickten sichelformigen Pigmentflecken, welche in der Mitte sich beriihren und eine x-formige Figur darstellen«. As a matter of course, the eye must be found on the dorsal surface, but I have been unable to find it on my larve, probably bécause the spirit had dissolved the pigment. The Antennule are always rather short; they consist typically of three joints, the second of which is usually short and not unfrequently coalescent with the first, in which case we only perceive two joints. The first, and particularly the third joint, are provided with pretty long sete; the terminal seta of the third joint is very long, and from the lower side of this joint proceeds always a single, particularly long olfactory seta (pl. I, fig. 11, b), which is at least double, usually several times, the length of the whole antennula; sometimes this seta is exceedingly long, as e.g. in Spher. dispar (pl. [X, fig. 3k) and in Spher. insignis (pl. X, fig. le), where it reaches further than the middle of the abdomen, nay in the last- mentioned species the olfactory seta in itself is longer than the whole cephalothorax. The antennule are always attached pretty far from each other at the edge of the cephalothorax, the area between them forming what I call the front. Close to the inner margin of the base of the antennula we often see an oblique list; moreover, in nearly all species of Spheronella which are parasitic on Cumacea, the front is decorated with one or several rows of delicate and peculiar processes, which decoration reaches its highest development in Spher. decorata (pl. VIII, fig. 3i and fig. 31). In Spher. modesta these processes are replaced by transverse lists (pl. [X, fig. 21). The Antenne proceed behind, and usually at the same time somewhat obliquely inside the antennule, but never from the base of the rostrum. Sometimes they are conside- rably shorter than the antennule, sometimes about the same length, and in the genus Homoeoscelis (pl. I, fig. 11 and pl. XIII, fig. 1h) more than double the length. In Spher. marginata (pl. XIII, fig. 6g) only two joints are found, in all other species they consist of three or four joints, three of which are always distinct, but it is often difficult to make out for certain, whether the eminence from which the supposed second joint proceeds, is a real joint, or in other words, if the apparent basal joint consists of two joints; as, however, Spher. antillensis has four very distinct joints (pl. III, fig.2e), this is probably the typical number. The terminal joint is nearly always short and usually ends in a long, thick seta, beside which we frequently find a shorter one. The next joint is now rather short, now long, or very long, and where the antenna is long, it is on account of the length of this joint, as the basal joint, or where there are four joints, the two first of these are never elongate, but sometimes (pl. IX, fig. 2i and fig. 3k) comparatively broad. In Mysidion (pl. XII, fig. 2h) the antenne are very small, and in Spher. microcephala (pl. VIII, fig. 2i) almost rudimentary, in both cases 3-jointed, with an exceedingly short terminal seta. ~The Rostrum seems to correspond only with the more distal part of this organ in ‘ 50 the female and the male, for not only are the antennz situated a little outside its base, but the maxillule are found quite, or nearly quite, outside it. The general structure of the mouth is like that of the adult animal; the mandibles are frequently seen in the opening, but the hairs of the mouth-border are always short, frequently so short, that they can only be discovered with the greatest difficulty. The Maxillule are difficult to understand, and it is very difficult to discover all their sete. As mentioned above, in the various illustrations of the larve I have drawn what I have found, but I am pretty sure that I have not everywhere found all the elements. The highest development I met in Spher. Calliopii (pl. III, fig. 31) as well as in the species which are parasitic in the marsupium of Cumacea, as Spher. decorata (pl. VILL, fig. 31), S. modesta (pl. LX, fig. 2i) and S. dispar (pl. [X, fig. 3k). In these figures we see obliquely behind and outside the rostrum on each side four sete, the two hindmost of which are coalescent or adjacent at their base, one or both of them being plumose. Obliquely from these and somewhat nearer to the rostrum proceeds a third seta, which turns straight towards the front, and obliquely before this one again, there is a fourth, shorter seta. I suppose that all these four setz belong to one maxillula, the basal part of which is not separated from the ventral side of the head. In Spher. marginata (pl. XIII, fig. 6g) the innermost seta is reduced to a short process. In most of the other forms I have only been able to find three sete, e. g. in Spher. microcephala (pl. VIII, fig. 21) and in Choniostoma mirabile (pl. XI, fig. le), or two, or only one. In Stenothocheres egregius (pl. I, fig. 11) I have found a maxillula (e) which reminds me much of those in the adult animals, as it consists of a short, basal part, from which proceed two thick sete, of which the hindmost is shorter. the foremost very long. In the genus Homoeoscelis (pl. Il, fig. 11 and pl. XIII, fig. 1h) I have been unable to find vestiges of maxillule. The Mazxille nearly always consist of three distinct joints, and are very much like those of the adult animals, though the stout basal joint, as a rule, is narrower and the two next joints are longer than in the adult; these two joints are slender, and the last one somewhat curved and claw-like. In Homoeoscelis only the two last joints are entirely coalescent, forming one curved joint, which moreover along the larger part of both margins is provided with exceedingly fine and short setiform processes (pl. XIII, fig. 1h). As a rule all three joints are simple and smooth, but in those species of Spheronella which live in the marsupium of Cumacea, the inner margin of the third joint is coarsely or finely serrated, and the first joint has on its inner margin, against which the second joint can be folded up, a double row of fine cylindrical processes (pl. VIII, fig. 3n and fig. 30), and a similar decoration is seen in Spher. Munnopsidis (pl. X, fig. 4d). — The two maxille are always situated at some — usually at a considerable — distance from each other. The Masillipeds, as a rule, are placed close behind the maxille, and also generally somewhat closer to the median line than these; in Homoeoscelis only (pl. I, fig. 11 and pl. XII, fig. 1h) there is a great distance between these two pairs of appendages. The 51 maxillipeds are always of considerable size, they are very like those of the males, and always consist of four distinct joints. Their rather stout basal joint is nearly as long as. or somewhat longer than, the three following joints together; the second and the third are very slender, and their joint length is a little longer, or somewhat shorter, than the last joint, which is extremely slender, almost setaceous, slightly curved and pointed. All the joints are nearly always smooth and naked; in Spher. Munnopsidis the fourth joint has towards its apex three spiniform processes (pl. X, fig.4d), and in Mysidion abyssorum there are five or six somewhat similar processes along the more central part of the inner margin (pl. XII, fig. 2h); the third joint at its extremity is always furnished with a spine inside the articulation of the fourth joint. _ In most figures I have carefully illustrated the sub-median skeleton, which consists of lists running backward from the base of the rostrum and the maxillule, surrounding the base and forming the articulation of the maxille and the maxillipeds. The two pairs of natatory legs of each specimen are very much alike, and they differ very little, comparatively, in the various species. Each leg consists of a good-sized peduncle, which is particularly broad in Stenothocheres egregius (pl. I, fig. 11, m and n), somewhat nar- rower, or rather narrow, in the other species (see particularly pl. VIII, fig. 2i); from the posterior margin proceed at some distance from each other two one-jointed, about equally long branches, and the outer branch, which proceeds from the end of the peduncle, is broader than the inner one and rather dissimilar in outline. The outer branch of the first pair of legs, as a rule, has four rather short, naked sete on its outer margin, two very long plumose setz on its terminal margin; on the inner margin it has either two very long plumose sete (pl. [X, fig. 3k) or one of this kind and one much shorter, naked seta (pl. I, fig. 11). The outer branch of the second pair of legs is very like that of the first pair, but its outer margin has only three shorter, simple sete, its terminal margin two, and its inner margin two exceedingly long plumose sete, all four of which are longer than in the first pair of legs. ~The inner branch of the first pair of legs has four, of the second pair three very long, plumose setz on its inner margin, and in both pair of legs two similar sete on its terminal margin (all these sete are longest in the second pair), whereas the outer margin has only one single seta, which is either short and naked, or very long and plumose. These are the results of my researches in the few species whose natatory legs have been examined with special care, but it must be observed that these species belong to three genera: Stenothocheres, Spheronella and Choniostoma. Even if an examination of more species should show greater variety in the number of sete, such differences are not at all likely to be considerable, and furthermore, it is in most cases exceedingly difficult to count the sete accurately, as the legs are very frequently folded up or standing on edge; therefore it would be all but impossible to make any practical use of the presence or absence of such a seta as characteristic mark of species or genera. — The two legs belonging to each pair are, as’ usual, united by a prominent, movable transverse band (pl. 1, fig. 11, m’ and n’‘). 7* ‘ 52 The first abdominal segment always dilates considerably from its base towards its end, and the free posterior angle has a powerful, often spiniform, seta, the length of which varies between being a little longer than the following segment (Stenothocheres, pl. I, fig. 11) and being longer than the whole abdomen, and plumose in its distal half (Choniostoma, pl. XI, fig. le). Inside or outside this seta and close to it there is always another seta, which, as a rule, is much shorter, and only in Spher. microcephala (pl. VIII, fig. 2i) is remarkably long, though somewhat shorter than the first one. The second segment is sometimes shorter, and in this case not unfrequently somewhat narrower, than the first one, e.g. in Choniostoma, sometimes quite as long, and always without seta. In Stenothocheres (pl. I, fig. 11) the third segment, together with the not separated caudal stylets, forms a large and broad, elongate segment, much larger than any of the preceding ones, and incised posteriorly in the median line. In all other species the third segment, together with the caudal stylets, is nearly always somewhat, and generally much, smaller than the second segment, and the stylets are sometimes not set off from the segment, but most frequently distinctly articulated on it as two short, almost cylindrical joints. Each stylet has always a very long and thick, sometimes plumose, seta, which in Stenothocheres egregius is only a little longer than the abdomen, in Mysidion and Aspidoecia somewhat longer, though not nearly half as long as the cephalothorax, in Homoeoscelis, Spheronella and Choniostoma longer than half the length of the cephalothorax, and sometimes attaining to three quarters the length of this part, e.g. in Spher. dispar (pl. [X, fig. 3k). Outside this long seta each stylet has in Stenothocheres four, in the other species two or three, comparatively short sete. Whereas there were great differences between the females among themselves and between the males among themselves in the different genera and species, we see from the detailed description given above, that all larva I know are surprisingly uniform, so much so, that I have been able to find rather insignificant generic characters only in Stenothocheres and Homoeoscelis, as distinct from the four other genera; at the same time the larve of Mysidion and Aspidoecia — whose females deviate much from those of the other genera with regard to receptaculum seminis, the position of the genital apertures and the hingement of the ovisacs — deviate less from various larve of Spheronella, than these ditfer from each other. Ill. Further Development of the Larve. The larve, after making their way out of the ovisac, — at least as a rule, swim out and seek a new host. I cannot deny the pos- sibility that one or a few of the larve may remain in the branchial cavity or in the marsupium of the mother’s host, though I doubt it very much. In a marsupium which was infested beforehand I have repeatedly found one or several larve, which were decidedly invaders. [ met with the greatest invasion in a specimen of Hippolyte Gaimardii, where in one of the branchial cavities I discovered a very young female of Choniostoma mirabile and certainly more than fifty larve and pup hinged on the gill-fibres (s. the special description below). 53 When the larva has found its new host, it attaches itself either beneath the cara- pace to the branchiz or in the marsupium to one of its plates, to one of the gills, or simply to the ventral surface of the body, or to the basal part of a leg, and in case it attaches itself to a not full-grown female of Amphipoda, of course it must content itself with one of the three last-mentioned places. The larve of Aspidoecia fasten themselves either outside on the dorsal surface of the host (an Lrythrops), if they are growing into females, or on a female of their own species (pl. XII, fig. 3b), if they are going to be males. The fixation, which is very solid, is effected by a viscous substance, which expands itself so as to form a larger or smaller plate on the front (pl. IV, fig. 1e,s; pl. VIII, fig. 21; pl. XI, fig. 1e,s). This viscous substance must be secreted of a gland in the front part of the head, the orifice of which, however, I have tried in vain to find; the gland itself must be studied from fresh material. (It was pointed out long ago by several authors, that the larve of various Caligide, of Achtheres etc., in their first stage fasten themselves in a somewhat similar manner by a »Stirnband« (Claus).) The larva, after hinging itself in this way. relaxes the grip of its limbs and hangs quite free; thereupon it begins to change form, bending forward the last joint of the maxille, and its cephalothorax getting gradually shorter and broader (pl. XJ, fig. le; pl. IV, fig. 1 e); how short and broad it may occasionally become, may be seen on pl. IV, by comparing fig. 1c, which represents a free larva, and fig. 1d, which represents a larva that has reached its full breadth, and which no doubt is going to develop into a male; fig. 1e is an intermediate form shortly after the fixation. Then the muscles etc. in all the limbs and in the abdomen dissolve themselves, and the contents of these organs are transferred into the cephalothorax, the muscles of which have also been dissolved, and finally this united substance is surrounded by a new skin under the old one. The subsequent development will be treated in the next division. e. The post-larval Development; the Pupe. The post-larval development, which takes place between the larval stage just described and the appearance of adult males and of females (which, though very small, in all important features resemble the egg-laying specimens), offer the greatest deviations between the different forms, but, unfortunately, the representation I am capable of giving is very fragmentary. The larva in many cases develops into a pupa, out of which evolves the female, and, in some cases, the male; in other cases the male is developed immediately from the larva etc., and in Mysidion the metamorphosis is more complicated. I know the complete development only of two species of Spheronella and of Homoeoscelis minuta, but the two first-mentioned, in particular, are fortunately very different from one another; I know, moreover, the develop- ment of the male in Aspidoecia, and of the females belonging to the species of the Spher. Leuckartii-group. Finally I have found a pupa of each of four other species of Spheronella 54 (S. Argisse, S. longipes, S. microcephala and S. insignis), numerous pupe of Choniostoma mirabile, and three different stages of development of Mysidion commune. Let us begin with the simplest form of development. Ona completely adult female of Aspidoecia Normani, sitting on the dorsal surface of the abdomen of Lrythrops serratus G. O. Sars, I found four larve attached (pl. XII, fig. 3b, where one of the larve is left out and two of the others designated by a v.) One of these came off easily, when touched, and it appeared that its skin had begun to burst along the lateral margin; on a closer examination it burst somewhat more, and it proved to contain a full-grown male. In this state the preparation is illustrated in pl. XII, fig. 3m. The spermatothecz of the male were filled, and it had a very short and very thick frontal thread (s). This instance suffi- ciently proves, that the male develops itself directly in the larva, without the intermediate pupa stage and without a second moult. I also found a single larva, which had placed itself directly on the carapace of the host, but its examination gave no result, its development not being far enough advanced. However, I have found exceedingly small females sitting on different parts of the host, and I have not the slightest doubt that the larve which are going to become females, attach themselves directly to the host, though I do not know, whether they ‘pass through the pupa stage or change directly into females, yet I feel inclined to accept the latter alternative, seeing that the parasite sits on the outside of its host, thereby running the risk of falling off very easily. We now come to a species in which both sexes pass through a pupa stage, namely Spher. Giardii (pl. VI. fig. 3i, fig. 3k and fig. 31). I have found three male pupe and not a few female ones. The body is of a short ovate shape; the males are more oblong and smaller, -125 mm. in length (fig. 3k), the female pupz stouter and varying in length between ‘125 mm. and -142 mm. (fig. 31, which is drawn on the same scale of enlargement as fig. 3 k). In the female the foremost two thirds of the ventral surface form a broad, naked, odd area, whereas the whole remainder of the body is closely covered with rather short hairs. The male pupa is covered with hairs like the female, except that part of the back which corre- sponds to the head and the upper parts of its lateral surfaces, which are naked (fig. 3k). Fig. 3i shows a male pupa from below, fig. 3k from the side; the antennulz (a), the antennz (c), the maxillule, the maxilla, the maxillipeds, the trunk-legs and the caudal sty- lets are seen as naked, smooth, pouch-shaped processes, whereas the mouth is developed as in the adult animal; in front of the mouth we notice a semicircular, rather vaulted and prominent area, and between the maxilla and the maxillipeds two pairs of oblong, somewhat prominent knots, corresponding with the sub-median skeleton. The female pupa has also a free mouth and all the same processes and projecting parts as the male, from which, however, it is easily distinguished, not only by the above-mentioned difference in the extent of the hair- coat, but also by the size of the processes which stand in the place of the trunk-legs, and which are very small, whereas in the male pupx these organs — especially the first pair —- are much larger and furcate at the end This difference between the trunk legs, which 55 entirely agrees with that in the adult animals, proves in a satisfactory way that my suggestion with regard to the sexes is correct. One of the illustrated male pupe (fig. 3 k) is fastened by a frontal thread, which is about one third the length of the body and conside- rably dilated towards the end, but its extreme expanded part is of a different quality and forms a disk-like plate. In the other male pupa (fig. 3i) the thread is shorter and some- what thicker, but its distal end is broken off. The fixation of the female pupa is effected by a thread, which is so short that the front part of the animal is pressed against the gill, or the plate of the marsupium, to which it is attached. — It is stated above that the pups have a well-developed mouth, and it would seem probable that at least the female pupz take food and grow a little. Undoubtedly the males and females come out directly of their respective pup, like the females of the species belonging to the group of Spher. Leuckartii, in which [ have observed the fact myself. Only one point seems to present some difficulty, namely, that my male pupe are only -125 mm. in length, whereas the male animals are between ‘17 and ‘21 mm. long and of a similar shape. With regard to this point I refer to my observation of the growth of the male of Spher. paradoza mentioned below on p. 57—58. Homoeoscelis minuta. A single pupa (pl. I, fig. 3b) was found hinged by a -frontal thread to the gill-bearing epipod. The pupa is 18 mm. long, of an elongate oval shape and naked all over. We see the pouch-shaped processes in which the antennulz (a), the antennz (c), the maxille (f), the maxillipeds (g), the first pair of trunk-legs (m), the second pair (n) and the caudal stylets (p) are developed; but besides all these, we notice between the second pair of trunk-legs and the caudal stylets a pair of very small, most peculiar processes (x), which are possibly a rudimentary third pair of legs that do not develop any further, and which disappear again. The mouth with the mandibles is like the pupa of Spher. Argisse (s. below); the frontal thread is scarcely a fourth of the length of the animal, it is simple, with a discoid expansion at the end. This pupa was hinged in the branchial cavity between two adult males attached in the same way, but there was no female, and these two circum- stances make it more than probable that the pupa was a female, especially, as in a large material of this species I have seldom found more than one male, and never more than two males and one female in the same branchial cavity. Later on I found in two specimens respectively two and three pup, one among the latter of which — being no doubt younger than the others — was somewhat smaller and had less developed rudiments of limbs, though otherwise it was similar to the other four, all of which agreed with the above-described specimen. (The frontal thread in one of the specimens was half as long as the body). Considering that (as stated above) I have never found more than one female and two males in the same branchial cavity, the four large pup must either all be males, or — which is probable — be male and female pup. So, judging from the sex of the minutely described pupa, there is no difference between the development of the two sexes, and this agrees very well with the fact that recently hatched females can sometimes be distinguished from the males only by possessing genital apertures, as in several males the spermatothecz are not distinctly seen. 56 At least with regard to the structure of the pupa we know, Spheronella Argisse, Spher. insignis and Choniostoma mirabile come rather close to the two preceding species ; each of them will now be treated separately. Spheronella Argisse. Of this species I have found a single female pupa (pl. TV. fig. 31 and fig. 3m), which in the most important features agrees with the female pupa of S. Giardii. The body is ovate, naked all over, 15 mm. in length, and, like the last-men- tioned species, provided with the pouch-like processes, in which antennule (a), antenne (c), maxillule (e) maxille (f) maxillipeds (g), trunk-legs and caudal stylets (p) are developed, and all these processes show about the same relative size as the corresponding organs in the females compared with each other; moreover, the mouth is distinct, and the mandibles are also seen in fig. 31. The animal is attached by a very peculiar crooked funnel of conside- rable size (s), which is of course made of a viscous substance, and its narrower, though comparatively thick, base proceeds from the front. ; Spheronella insignis. Of Diastylis cornutau Boeck a single pupa was found (pl. X, fig. 1h), 14 mm. in length and sub-globular in shape. The mouth is well developed, as in an adult female; antennulz, maxillule, maxilla and maxillipeds are seen as protruding naked pouches of somewhat irregular shape; in front of the mouth are found two rather large, odd, connected areas, and close behind them a smaller spot on each side of the median line; in front of, and in an oblique direction from, the maxillule, there are several small areas, and finally a large transverse area, expanded in the middle, which joins the bases of the maxilli- peds. All these areas are naked, and more or less prominent compared with the surroun- ding skin. Except these organs and areas and the surface between them, the whole ventral surface of the pupa and its sides are covered with hairs of medium length; the back is covered in the same way on its most anterior part and on its posterior half, whereas a broad, transverse area extends over the larger part of its anterior half; this area is naked, and closely covered all over with irregular projections of a comparatively rather considerable size, the real shape of which can only be recognised when seen obliquely, whereas a perpen- dicular view gives the impression that this surface is covered with numerous irregular lines, forming ovals, oblong spots or simply flourishes. I have been unable to find out with certainty the sex of this interesting pupa, as the adult male of this species is devoid of trunk-legs; however, I consider it to be a female. Choniostoma mirabile. On the gills of a Hippolyte Gaimardii M.-Edw. I have found a number af larve, which showed all the stages transitional to that of the pupa, besides numerous fully developed pupx (pl. XI, fig. 1g—1k). A comparison between the figures 1h, 1i og 1k, which are drawn to the same scale of enlargement, will show that the pup differ much in size and somewhat in shape; the largest specimen I have measured is -38 mm. broad and -27 mm. long, the smallest only -27 mm. broad and ‘19 mm. long, and one specimen, which was still enclosed in the skin of the larva, is only ‘19 mm. broad and -18 mm. long. So the pupe grow very considerably after breaking out of the larval 57 skin, and there can be no doubt that they take nourishment through their well-developed mouth (s. later on under Spher. danica.) They are always considerably depressed, the anterior margin is long, now quite straight, now a little concave, in the younger specimens occasio- nally a little convex; the lateral margins converge rather considerably backward and merge more or less evenly into the posterior margin. The mouth is pretty well developed, with distinct mandibles (fig. 1g), and is situated much to the front on the ventral surface; it is surrounded by rather small pouch-like processes: antennulz (a), antenne and maxille, and behind it is found a rather large, peculiar area. Near the anterior margin on the ventral surface is seen a body with an irregularly curved posterior margin, and this is the adhesive plate (s), by which the animal is hinged. The parts just described on the ventral surface are surrounded by a naked area, whereas the greater part of the remainder of its surface is provided with hairs, which are sometimes arranged in rows of two or three or more, sometimes are more scattered, and the hairs which grow close to the edge are rather long, the more central ones are short; the larger part of the dorsal surface is naked. Fig. 1i and fig. 1k reveal that the contents of the pupa — perhaps on account of alcoholic influence — do not reach the outer skin, whereas fig. 1h shows a very large and entirely filled pupa; fig. 1i, the specimen represented on a larger scale, so far shows the same as the two illustrations just mentioned, but it reveals at the same time that the inner body has its own skin with distinct sete at its margin. The male of Choniostoma being wnknown, I cannot decide how far all the specimens found are only female, or both female and male pupe. We now come to a species — Spheronella paradoxa — whose development differs very much from that of Spher. Giardii, or that of Homoeoscelis minuta, and the. forms which, according to my just stated (though rather defective) knowledge, are related to these species. In Spher. paradoxa both sexes develop themselves without passing through the stage of an independent pupa properly speaking, nevertheless there is a considerable. difference of aspect between the development of the sexes. Let us first, examine the development of the male. I had a very considerable material of larva, which-were hinged by their front, and in which the cephalothorax was nearly as broad as long, besides being very thick (pl. LV, fig. 1d); however, as it was impossible to me, in spite of careful searching, to find a single male pupa, I conceived the idea of submitting my big larve to a thorough examination, and in a single specimen I fancied that I saw two spermatothece through the skin. I succeeded in taking away the skin of this larva, and I found a young male with hairs on its abdomen, rather short caudal stylets, and two spermatothecz, whereas the limbs were still for the most part rather indistinct. Hereby, then, we havé got the. proof that the males come out directly of these big larve (comp. Aspidoecia, p. 54), still there remains a single point which is not fully explained. My rather numerous males are — with very rare exceptions — about ‘245 and -27 mm. in length, though one of the big larve is only 19 mm. long, the empty abdomen included, and the cephalothorax of one of the largest of these larvee is only *16 mm. in length (in pl. IL, fig. 4c I have represented a male, in fig. 4e such a larva enlarged on 8 58 the same scale, and a comparison between these two figures offers a pretty good illustration of the different sizes seen from below). Fortunately I have found a single male which without any doubt, judging from its inward and outward condition, is quite recently hatched, and which is only -18 mm. in length, consequently ‘02 mm. longer than the above-mentioned cephalothorax; now, if we consider the prominent frontal border in the male and its rather more elongate shape, this slight difference is accounted for. The result is, that ammediately after hatching, the male must grow to some extent, for, as males of small size in this (as in other) species are pretty rare, we have good reason to suppose that this growth is compara- tively rapid. The female apparently passes through a pupa stage. I have found three such »pupz« altogether, which were all about the same size; the specimen illustrated in pl. LV, fig. 1f is 174 mm. long. The body is ovate, somewhat flattened and attached at the front by a broad adhesive plate (s). In the illustration several limbs are seen, but, on closer examina- tion, it appears that all these organs are those of the larva: antennule (a), antennz (c), maxilla, maxillipeds (g), first pair of natatory legs (m), second pair of natatory legs (n) and abdomen (0), in other words, the animal is enclosed in the skin of the larva, whose appen- dages and abdomen are not only emptied of their contents, but have shrunk, so as to be almost unrecognisable. There is no mouth. Under the skin we see the scarcely developed mouth, the maxille and the folded mazxillipeds of the young female. So the skin of the larva has acquired the appearance of a pupa; a real pupa does not exist. The animal cannot possibly take any nourishment. Fig. 1g in pl. IV represents a young female that has just burst the ventral side of the »skin of the pupa«, whereas its ragged dorsal part still hangs on to it; this specimen was only ‘207 mm. in length, consequently only 034 mm. longer than the pupa represented. This young female was still attached by the adhesive plate (s) of the skin of the larva. A pupa deviating from those of the above-mentioned types is found in Spheronella danica, Spher. chinensis and closely allied species, which, together with Spher. Leuckartii, form a small group, which I have named after this species. Salensky (in his op. cit.) has described and illustrated several stages of development of Spher. Leuckartii, and his obser- vations agree very well with mine, only I have been able to make some additional statements. The pupa is ovate, sometimes naked on its anterior part (pl. III, fig. 2f), though, as a rule, it has only a smaller naked spot in the midst of the ventral surface (pl. II, several figures); otherwise it is all over pretty closely covered with rather short hairs; from the anterior end, which is always narrower or more pointed, proceeds a tuft of longer hairs, and in the midst of these is a rather short thread, which ends in a disk (pl. LI, fig. 6 e), by which the pupa is hinged, either to one of the plates of the marsupium, to the inner side of the basal joint of a leg, or to a gill. (Usually this frontal thread proceeds from a small depression with flat bottom, however, in one case, I have noticed that it proceeded from a stouter, short, cylindrical eminence (pl. IL, fig. 4d and fig. 4e). On the posterior half of the above- 59 mentioned naked spot we find a somewhat prominent mouth (pl. IT, fig. 6e and fig. 6f; pl. III, fig. 2c), which is supported by some chitinous lists, the two longest of which point straight forward, are as long as, or somewhat longer than, the diameter of the mouth and enclose the anterior half of the naked spot. Of mouth-appendages we only see the well- developed mandibles, the points of which project in the orifice of the mouth; the hairs round the mouth-border are apparently wanting (but perhaps the magnifying power of my micro- scope does not suffice to discover them). Outside the anterior half of the mouth, and beneath the naked area in front of it, several muscles are seen, which evidently serve the action of the mouth. The pupz of the same species show great difference of size. In Spher. chinensis a small pupa is 146 mm.. a large one -24 mm. in length, and both are illustrated in pl. I, fig.6¢ and fig.6d magnified on the same scale. But I have found a much greater difference in Spher. danica, of which species I have seen nine pup: the smallest (pl. II, fig. 4d) is only ‘115 mm. long and ‘085 mm. broad, an other (fig. 4c) is 185 mm. long, and the largest is ‘25 mm. long and 194 mm. broad, thus somewhat more than double the length and double the breadth of the smallest specimen. Jn the small pupe I have not been able to find the vestige of any organ, and their contents — except the muscles of the mouth — consisted of a granular substance. Salensky writes (op. cit. p. 317): »Weder innere Organe, noch selbst irgend welche Formelemente liessen sich im Innern wahrnehmen; doch kénnte méglicher Weise der Darmkanal vorhanden sein und nur durch die zahlreichen Kérnchen verdunkelt werden« —; later, on p. 318, he describes the mouth, but he has seen neither the man- dibles uor the muscles, nor has he been aware that he had a mouth before him, and he concludes his statement about his youngest specimen, which is “12 mm, long, in the following words: »Das Wachsthum geht in dieser Periode sehr rasch vor siche. This last observation I consider to be correct, and as it appears from the above-stated measurements, that the pupa during its development grows to a manifold larger volume, if 7s clear that it must tale nourishment through its mouth, of which we have given a description, but it must be left to new examinations of fresh material to explain how this is effected, if there really exists an intestinal tube — which we must naturally suppose — but the contents of young and half- grown pup preserved in alcohol have made the impression on me of being evenly granulous, without organs. In the above-mentioned largest (24 mm. long) pupa of Spher. chinensis (pl. U1, fig. 6b) and in the largest (‘25 mm. long) pupa of Spher. danica I found young females, and Salensky writes (p. 319): »Das letzte von mir geselhene Stadium zeigte das unter der Puppenhaut schon vollkommen ausgebildete Thier mit allen Anhiingen. Die Puppe hat eine Linge von 0,27 Mm. erreicht....; an ihrer Oberfliiche sind keine Verinderungen sichtbar«; whereupon he describes the animal, though without mentioning that it is a female, however, this is easily seen from his illustration. My specimen of Spher. chinensis (pl. IL, fig. 6 f) shows antennule, mouth, maxilla, maxillipeds, the two pairs of trunk-legs, the not yet fully developed genital apertures and the caudal stylets, which are seen through the hairy skin. It is seen that the mouth of the young female is situated far more to the front 8* 60 than that of the pupa, and is much larger. The largest pupa of Spher. danica agreed entirely whith fig. 6f, except in one remarkable point: the mouth of the female projected freely through the skin of the pupa (though I was unable to discover any rent or larger opening in the skin), thus presenting the extraordinary sight of two mouths protruding beyond the ventral surface of the pupa; the mouth of the female is situated a little behind the front extremity, that of the pupa nearly in the middle of the surface; the diameter of the latter in proportion to that of the former being as seven to ten. — Fig. 2¢ in pl. III illustrates such a young female belonging to Spher. antillensis, which has rent and thrown off about the front third of the skin of the pupa, whereas the hindmost part of the body is still enclosed in the larger posterior part of the skin, including the mouth of the pupa, and as the same animal is illustrated in fig. 2a magnified on the same scale as the ovisac exhibited in fig. 2b,a comparison of these two figures will show clearly, how much the animal grows from its stage as an egg up to the moment when it comes out of the pupa. But what about the male? The animal which Salensky found in its earliest stage as pupa, and of which he has had several specimens, according to his description and illustration is only the female, and of the pupz I have seen, the smaller ones contained no animals, and the very large ones contained females. As mentioned above, a pupa containing a female of Spher. danica is -25 mm. long, but a male which, judging from its appearance, seemed to be recently hatched, was only +15 mm. long, whereas a full-grown male of the same species was °24 mm. in length. So we notice here the same growth of the young male as mentioned above in Spher. paradoxa, but at the same time we find that the full- grown male is somewhat smaller, and the recently hatched male only a little more than half as long as the large pupa (25 mm.). As the small pupz examined by Salensky and by myself never contained any animal, it is very probable that the male of this species, as well as that of Spher. paradoxa, is developed directly from the larva, and the size of the recently hatched male agrees perfectly with this supposition. Unfortunately, of all my eight species of this group I have only seen very few larve, however, the contents of a single specimen of these seemed to indicate that a male, not a pupa, was developing in it, still its growth was not sufficiently advanced to decide the question with absolute certainty. Of Spher. longipes I have found a single specimen of a pupa (pl. VII, fig. 2 g), which essentially coincides with those belonging to the group Spher. Leuckartii. This pupa is ab. ‘17mm. long and -11_mm. broad, rather elongate, as we see; the lateral margins run almost parallel in part of their length, the front extremity is somewhat pointed and has a small orifice, out of which a short broken thread is protruding. The pupa is quite naked and possesses in the centre of its ventral surface a small, but well developed, projecting mouth, at the front and at the sides of which some chitinous lists and muscles are noticed. Its contents consist of a granular substance, in which no organs are indicated; only towards the front extremity we see — as shown in the illustration — some vague indications of an organic structure. 61 In this place I will briefly mention an organism, which I found on the glass after having prepared Spher. microcephala, and which, I suppose, is the pupa of this rather deviating species. It is somewhat depressed (pl. VIII, fig. 2k); its outline is ovate with a straight posterior margin, whereas its rounded front margin bears a somewhat protruding adhesive plate (the stripes of which are too strongly marked in the illustration); its dorsal surface is provided with some short hairs, the ventral surface is naked. It is -20 mm. long, has neither mouth nor other outer organs, nor do we find distinct indications of internal organs. In Mysidion commune the metamorphosis is more complicated than in the preceding forms, but, unfortunately, my material is not large enough to allow me to elucidate it in all details, besides, the forms in hand present several features which I do not understand. I have found altogether three stages of development, two instances of the earliest, one of the medium, and two of the last and largest stage. I will begin with this last stage, which indeed presents a kind of semi-pupa, or a young female in possession of features which it afterwards loses. The two specimens found are of about equal size, the one illustrated in pl. XII, fig. 1d is -31 mm. long. The body is elongate ovate, rather pointed at the front extremity, which has a mouth provided with a border and its surrounding hairs; on its sides are the maxillule, and on the ventral surface, a little behind the mouth, are the maxille and the maxillipeds, which, though well developed, in some small, unimportant points deviate from those of the adult female; on the dorsal side, at a rather good distance from the mouth, we find the one-jointed organs, which for a long time I considered to be the antennulz (a), but which no doubt are better explained as being the antenne. On the ventral surface, at a considerable distance from the posterior extremity, we see an odd, strongly protruding, elongate and somewhat pointed process (x), and nearer the posterior margin, somewhat up on the back, the scarcely fully developed crescent (r) which surrounds the future genital aperture. In the middle of the back appears an odd, rather low, blunt excrescence, from which proceeds a most peculiar fixation-thread, consisting of two divisions. The first part (u) is somewhat shorter and thinner than the basal joint of the maxilliped, and its distal part is tubular; from the inside of this tube the second division comes out as a thread, which is thin in a considerable part of its length, then dilates rapidly and widely (v), forming a low collar at its widest expansion; it continues beyond the collar rather thick, in the middle somewhat thinner; this part is hollow, very light and is no doubt furnished with very thin walls, and its end is fastened to a plate of the marsupium of the host. Can this singular fixation-thread be considered as homologous with the frontal thread of other pupe? This would seem probable, though it is placed rather far backward; how it is produced is incomprehensible to me, but its distal end is very like the thread I have de- scribed in the male of Mysidion abyssorum, and its proximal part exhibits great likeness to the frontal thread in the male of Spher. danica. Somewhat in front of this thread, on each side, at a short distance from the outline of the back, we see a conical process (t), which for a long time was inexplicable to me, but which I suppose must be explained as the 62 antennule; according to this interpretation they ought to have been marked »a« in the illustration, and the antenne not »a«, but »c«. Somewhat behind the fixation-thread I found a pair of bodies (z) which were situated at a short distance from each other and looked as if they were pasted on; whether they belong to the animal I do not know, though I suppose they do, as one of the specimens was provided with both these small bodies, the other with one of them. The hindmost part of the body and a smaller part of the surface surrounding the large ventral process, are provided with a number of rather short hairs. Fig. 1¢ in pl. XIL shows the stage of development which immediately precedes the afore-described stage. The only specimen found is ‘24 mm. long; the scale of enlargement is like that of the last stage. The body has about the same shape, and the mouth, the maxillule, maxille (f), maxillipeds (g) and antenn (a) have a similar structure and position; but the antemnule (t) are placed further backward than in the more advanced stage, almost behind the middle of the animal; their form is somewhat vague. The dorsal fixation-thread (u and v) proceeds at some distance behind the middle of the median line of the back; it is almost constructed as in the older stage, but is considerably longer, and its distal part is much less stout. The crescent of the genital aperture (r) is not found on the dorsal, but on the ventral surface, near the posterior extremity of the body. The odd ventral process (x) is situated almost in the middle between the base of the maxillipeds and the posterior end of the body; it is rather short and very stout and broad; it is indeed a projection, which has on its top a well-developed mouth (y) with mandibles, and whose sides and front part are provided with chitinous lists, like those which surround the mouth of the pupa in the group Spher. Leuckartii; on the side this skeleton forms a figure (y’), which, seen as in the drawing obliquely sideways, looks very much like a maxillula; however, on closer examination it turns out not to be any appendage. Now we should feel much inclined to think, that the pupa just described does indeed represent an animal provided with an apical mouth in the act of breaking out of the skin of the pupa, the mouth of which is marked »y«, however, repeated and careful examinations of the admirably preserved pupa, which [ have been able to roll under a glass-cover, and to study from all sides, give the result, that a is really an animal with a larger apical and a smaller, but very well developed ventral mouth, the latter of which disappears in the following stage, only leaving the odd ventral process x in fig. 1d. That the existence of two mouths in the same animal is per- fectly incomprehensible to me. goes without saying. — The whole animal is naked. Finally I have found two badly preserved, infinitety small pup of about equal size, one of which is illustrated in pl. XI, fig. 3i. The body is -136 mm. long, shortly ovate, with a well-developed mouth at the rather pointed front extremity, and a little more back- ward on the ventral side maxilla and maxillipeds, the former of which being of pretty good size, but with thin walls and-of a somewhat vague form, whereas the maxillipeds are almost smaller than the maxille, 2-jointed, and very weak. Beneath the skin, between the maxille, is seen a pretty large, anteriorly inflexed ring (x), which seems to be the beginning of a 63 mouth. The pupa appears to have been attached by a dorsal thread, as in the preceding stage, however, it is too badly preserved to allow of a more precise definition. I am unable to give any more details about this stage; I do not see at all how it can be an earlier stage in the development of the female, and consequently be followed by the two above-described stages; so it may possibly be a male pupa; however, it must be left to the future to solve these and other problems in the remarkable development of Mysidion commune. I have now communicated in detail all [ know about the post-larval development of the forms of this family. Being unable, on account of the great gaps, to generalize very much, I have preferred to collect all I know in this place, instead of contenting myself with making a shorter extract and distributing the greater part among the forms in question in the later systematic representation. Though I think I have found a series of rather intere- sting facts, this is only the beginning of a complete elucidation of the very peculiar meta- morphosis of these animals with their extraordinary variations in the different species. It would indeed repay the trouble to carry out such ai investigation in numerous representatives of this family, but it would at the same time present enormous difficulties, on account of the nature, as well as of the rarity, of the material. B. Habitation, Biology and .Distribution. a. The Place of the Hosts in the System and the Habitation of the Parasites. Of the forty-three species examined by me, two (the genus Choniostoma) live in the branchial cavity of two species of the genus Hippolyte Leach, which belongs to the tribe Caridea of the order Decapoda; two species (the genus Homoeoscelis), live in the branchial cavity of two species belonging respectively to the genera Diastylis Say and Iphinoé Sp. Bate, which two genera belong to widely differing families of the order Cumacea; one species (the genus Aspidoecia) lives on the outside of the body (on the carapace, on the back and the sides of the last free thoracic segment and of the six first abdominal segments, as well as on the eye-stalks) of the species of the genus Hrythrops G. O. Sars, which belongs to Myside vere. All the remainder — thirty-eight species — live in the marsupium of species belonging to the following orders: Mysidacea, Cumacea, Isopoda and Amphipoda; however, their distribution within these orders is rather interesting. In Mysidacea I have only found two species (the genus Mysidion) on the genera Erythrops G. O. Sars and Parerythrops G. O. Sars, belonging to Mysidz vere, and the three species on which they are found live — according to G. O. Sars — in a depth varying from 30 to 300 fathoms. An examina- (64 tion of a very large material of Danish species of the genus Mysis Latr. (sens. Sars in 1879), and of Macropsis Slabberi (v. Ben.), as well as of a number of specimens of Gastro- saccus Norm. — all shallow water species, — gave a negative result, neither did I find any parasite on numerous specimens of a species of Mysidz taken in shallow water in the West- Indies, nor on a great number of specimens belonging to two pelagic species of the genus Siriella Dana. In Cumacea I have found altogether five species of the genus Spheronella on six species belonging to the genera Diastylis Say, Eudorella Norm. and Iphinoé Sp. Bate, which three genera belong each to one of the eight families established by G. O. Sars, whereas an investigation of numerous other species, among which a very large material of several species which were taken in Denmark in very considerable quantity, as Cuma scor- pioides (Mont.), Lamprops fasciata G.O. Sars, Leucon nasicus Kr. Leucon nasicoides Lilljbg., Leucon acutirostris G. O. Sars, Eudorellopsis deformis (Kr.) Diastylis resima (Ky.) and Leptostylis ampullacea (Lilljbg.), gave a negative result. Within the order [sopoda, these parasites — altogether three species of the genus Spheronella — are only found in two species of the genus Janira Leach and in Munnopsis typica M. Sars, all three forms belonging to the large tribe Asellota, whereas an investigation of numerous forms belonging to other families, among which some species of Idothea F., Astacilla Cordiner, and Gnathia Leach, were represented by a great number of specimens, gave no result. (Of the small order Tanaidacea I have examined a good number of specimens from Denmark and numerous specimens of several species from Sicily, without finding a single parasite). The chief quantity of the parasites, namely twenty-eight species (the genus Stenothocheres and about three quarters of the genus Spheronella), were found on Amphipoda, and within this order exclusively on Gammaridea (on twenty-eight species). An inspection of a great number of specimens of different species of Caprellidaee and of some specimens of Cyamidz gave no result. (Among the material from the »Ingolf« expedition I found in 1895 on an AZgina Kr. a species, which will be described in the report on the results of this expedition). Of Hyperiidea I have only examined a few species. In giving a short general view of the occurrence of the parasites within the Gammaridea-group, I will avail myself of the twenty- five families adopted by G. O. Sars in his new important work: »An Account of the Cru- stacea of Norway, Vol. I, Amphipoda«; they are found in one or more representatives of the following fourteen families: Pontoporeiide (the genera Bathyporeia Lindstr. and Argissa Boeck), Ampeliscidee (the genus Ampelisca Kyr.), Amphilochide (the genera Astyra Boeck, Amphilochoides G. O. Sars and Gitanopsis G. O. Sars), Stenothoide (the genera Stenothoé Dana and Metopa Boeck), Oediceride (the genus Perioculodes G. O. Sars), Paramphithoide (the genus Paramphithoé Bruz.), Epimeride (the genus Acanthozone Boeck), Syrrhoide (the genus Bruzelia Boeck), Calliopiide (the genus Calliopius Lilljbg.), Atylidee (the genus Paratylus G. O. Sars), Gammaride (the genus Cheirocratus Norm.), Photide (the genera Lemboides Stebb., Protomedeia Kr., Leptocheirus Zadd., Gammaropsis Lilljbg. and Microprotopus Norm.), Corophiidx (the genus Corophium Latr.) and Dulichiide (the genus Dulichia Ky.). Of the remaining eleven families the 65 majority are small, with one or very few genera and rather few species, but the family Phoxocephalidze is pretty considerable, the family Podoceridxe is large, and the family Lysianasside is exceedingly large (in Sars’ work thirty-three genera), wherefore it seems interesting to me that no parasite of our Choniostomatide has been found on any species belonging to these families. One species was found in the Mediterranean in the genus Microdeutopus Costa, belonging to the Photidg, another species has been quite recently discovered in Cyclaspis G. O. 8., belonging to the family Cumide, and these genera are the only two mentioned in the literature of the subject, in which I have not personally observed the parasites of this family. To give an account of the forms the examination of which led to no result, would be too tedious, neither would it prove much; I will only say that I have examined a good number of exotic species, most of which were only represented by a few specimens, besides nearly all of the large material our museun: possesses of Gammaridea from Denmark, Greenland and the Kara Sea, and many of these species were represented by from fifty to hundreds of specimens. In F. Meinert’s three papers, of 1877, 1880 and 1890 respectively, about Danish Malacostraca, and in my own similar papers about the fauna of Western Greenland and of the Kara Sea, will be found the names of most of the northern and arctic species examined, of which I have had a large material. In a later paragraph I shall mention a little more in detail the following pheno- menon which stands in a certain connection with the matter above, namely, that of several species a considerable material from a large sea can be examined without showing a single parasite, whereas sometimes a smaller material of the same species from another sea reveals several parasites. This proves that we cannot conclude that a species is not infested, from the fact that an investigation of hundreds of specimens from different localities of a certain country has not led to the discovery of any parasite. In most cases such examinations must be undertaken on a much larger scale than I have been able to do, before any value can be attached to the negative results. b. Age and Sex of the Hosts. It serves our purpose best to divide the hosts into two sections according to their parasites, viz. whether the typical residence of these animals is in the marsupium or in other places. I will begin by the latter section, repeating my above statement that I only have examined five species of parasites which do not live in the marsupium. Aspidoecia Normani, which, as has just been said, lives on the outside of the body of species belonging to the genus Lrythrops, I have found on young specimens as well as on adult males and females, but in the latter the marsupium was either empty, or occupied by a species of the genus Mysidion. The two species of the genus Choniostoma live in the branchial cavity of two species of the genus Hippolyte. From the Kara Sea I have seen 9 66 altogether seven infested specimens, four of which were adult females without eggs, and each of these females was infested with one adult parasite, three of which having laid numerous ovisacs; two of the hosts were males: one was an adult infested with a parasite with numerous ovisacs, the other was a little smaller with eight smaller parasites; and finally, the seventh host was a female with eggs containing halfdeveloped young ones, on its right side was an empty swelling on the carapace about two thirds of the normal size, on the left side a very small and quite young female, and besides numerous larve and pup hinged on the gills. - The two species belonging to the genus Homocoscelis live in the branchial cavity of two species of Cumacea; of one of these: Iphinoé trispinosa (Goods.), I have seen seven infested specimens: one female, whose marsupium contained a Spheronella, three not quite fullgrown females, one of which — whose marsupium was in an early stage of development — was infested on one side with a female, a male, and eight ovisacs, in one of which were full-grown larve. The three last specimens were a male before the last moulting and two adult males, one of which with an adult female and two ovisacs, a young female and a male in one branchial cavity, the other containing only a half-grown female. Of the other species, Diastylis lucifera (Kr.), I have seen sixty-six infested specimens — most of them females, in at least three cases young males, but not a single adult male, though this last circumstance is of less weight than might be expected, as our naturalists have neglected to throw out the surface-net at night and in the evening in order to catch the full-grown roving males. About three fourths of the females had a well-developed marsupium; in more than two thirds of these it was empty, but in at least thirteen cases is was filled with half or fully developed young ones, never with eggs. Females which had not yet begun laying eggs appeared in females of Diastylis with young ones in the marsupium, as well as in specimens without marsupium; there were found likewise female parasites with the full number of ovisacs in females of Diastylis with young ones in the marsupium, and in younger specimens without marsupium. — In a sub- sequent paragraph I shall have an opportunity of entering into further consideration of these statements; more special statistics are found in the systematic part. We now come to the thirty-eight species which live typically in the marsupium of forms belonging to four different orders. I may say at once that the ten species which appear in Mysidacea, Cumacea and Isopoda, I have only found in perfectly developed mar- supia, but it must be added that I have also constantly found at least one older female with ovisacs in such a marsupium, so I know nothing about the stage of development of the host at the time when the first (and often only) female attached itself to it as a larva; I have examined numerous specimens of Cumacea, in which the marsupium was beginning to develop itself (it appeared as small plates), but without finding any parasite. Amphipoda presented somewhat different facts. Salensky writes about Spher. Leuckartii (op. cit. p. 302): »Das Thier fand sich in der Bruthdhle der Weibchen und an der unteren Flache der ent- sprechenden Brustsegmente der Méinnchen und war an den iusseren Bedeckungen des Wirthes mittelst eines besonderen Saugapparates befestigt.« By this »sucking apparatus« 67 the author means the rostrum, but this observation is not just, for the attachment certainly takes place by a frontal thread; moreover, I suppose that the animals Salensky took for males were in fact somewhat younger females without marsupium, founding this suggestion partly on his (otherwise bad) description of the host »Amphitoé sp.«, — according to Della Valle: Microdeutopus gryllotalpa Costa, — in which he does not say a word about the very great difference in the »hand« of the first pair of trunk-legs between the two sexes, partly on the fact that I have never found a Spheronella on any adult male; whether some of the not full-grown specimens on which I found typical marsupium-parasites, were young males, I cannot tell, but I doubt it. The twenty-four of the species parasitic on Amphipoda I have found exclusively in marsupia, and though, in not full-grown animals, I may not unfrequently have overlooked larvee, pup or very diminutive females, in any case I cannot have overlooked many females with ovisacs. Only in the following four species of Amphi- poda: Metopa Bruzelii (Goés), Argissa typica Boeck, Protomedeia fasciata Kr. and Ampelisca tenuicornis Lilljbg., have I found parasites in specimens without or with half-developed marsupium. In a specimen with scarcely half-developed marsupium of Mefopa Bruzelii, two larvee were found, and in a still younger one without marsupium, a single larva. In two young females without marsupium of Argissa typica appeared respectively one pupa and a tiny female of Spher. Argisse. In a young specimen of Protomedeia fasciata Ky., from Greenland, were found a not half-grown female and a male of Spher. Bonniert. Spher. longipes I found in nine specimens of Ampelisca tenuwicornis; two of these only were females with fully developed marsupium, the third was a young female with half-developed marsupium, which contained a not half-grown female of the parasite; the six remaining specimens were young, without marsupium, and on each of the five of these I found a single female between not half-grown and very small, — in one case even recently hatched; in the sixth spe- cimen there were only two loose larve. The result hereof is, that in Amphipoda I have not found a single adult female in a specimen without entirely developed marsupium, and never ovisacs except in marsupia. It is probable that larve not unfrequently fix themselves to immature females, beginning their development there, and thus entailing the necessity that larve as well as young females, and rarely males, remain on the host, while it passes through its last moultings; however, as said above, not a single observation has been made of ovisacs being found in females not fully developed, which by the by, seems natural enough, as they would certainly be washed away, if they were laid. However, I cannot prove that most specimens are infested before the marsupium is fully developed. No doubt, the larve seek either perfectly mature females — and at least rather often those whose marsupium is already infested by at least one (half-grown or quite adult) female and a male —, or such younger specimens as are so far advanced, that they will have got their marsupium before. or at the time when the females that have developed themselves out of some of them, are ready to begin laying eggs. Whether the larve of species that live in 9* 68 the marsupium of Isopoda, Cumacea and Mysidacea sometimes, or often, fix themselves to not full-grown females, I repeat, I cannot tell. I will add that in three cases (in Ampelisca tenwicornis, Protomedeia fasciata from Denmark, and in the same from Greenland) I found in the marsupium, together with one female Spheronella without ovisacs (in two cases a male attached to it), four to six of the Amphipod’s own eggs, in one case with half-developed young ones. In the marsupium of an Ampelisca typica I found a large female and two ovisacs of Spher. microcephala and two of the Amphipod’s own eggs. In Hudorella truncatula I found one of its own eges together with an adult female, three ovisacs and a male; in another specimen were found no less than twenty-four of its own eggs together with an almost adult female and two larve which were invaders. c. Number of Parasites on each particular Host. In the systematic part of this work I give a kind of statistics of each species, accounting for my findings, and giving numerous data concerning the number of each sex, of the ovisacs and the stages of development found on each particular host. Of this con- siderable material I put down some extracts here, which will give a condensed view of this matter. Of Aspidoecia Normani which lives fixed on the outside of Hrythrops, I have often found one or several females of very different sizes on the same host, in one case as many as six females with ovisacs, three younger females and one larva on one single specimen. In a large material of Diastylis lucifera with Homoeoscelis minuta in the branchial cavity, I have never found more than one female, as a rule only one male, very seldom two males on one specimen, whereas of Iphinoé trispinosa, infested with Hom. mediterranea, only four specimens have been thoroughly examined by me; one of these had two females, six ovisacs and two males in the same branchial cavity, one had an adult and a young female, two ovisacs and a male in the same branchial cavity, one carried a male in one branchial cavity, a female with eight ovisacs and a male in the other. Of the species of Choniostoma which live in the branchial cavity of Hippolyte, I have found only one specimen of an adult female on a host, whereas of younger female parasites one specimen contained three in one branchial cavity, five in the other; in another specimen I saw an empty swelling covering one of the branchial cavities, whereas the other contained one female which was far from half-grown, besides certainly more than fifty larvae and pupe hinged on the gill-fibres; however, I doubt very much whether most of these would have been able to develop themselves into adult females (and perhaps males) on this shrimp; it seems to me rather doubtful that the animal should be able to afford the nourishment required, and still more so whether the parasites would find sufficient room to grow. In the parasites living in the marsupium we find the greatest differences as to the numbers of them on one host, but at the same time it must be observed that while 69 some species as a rule only contain a single female and a male, sometimes two males and seldom two females in the same marsupium, in other species we pretty frequently find two, three, or more females and several or many males on the same host. Of twenty-six out of twenty-eight adult specimens of Mefopa Bruzelii (Goés) infested with Stenothocheres egreyius which were examined and noted, there were found only one female and frequently also one, seldom two males on each specimen; in one specimen were found only eggs and young ones, in one two females and no male. In Cumacea and Isopoda the marsupium never contained more than one female, often also a male, sometimes two, and in a single case three males, besides, in one case [ found one pupa, in another, where no male existed, some invading larve. In Mysidacea some deviations are observed; of adult females we very rarely find more than one specimen, but rather frequently also one or two young females or tiny young ones which have not gone through the whole metamorphosis, besides frequently one or more, in one single case even ten males; (as for further details, s. statistics in the systematic part). In one specimen of Hrythrops serratus there had lived at least three females with ovisacs and one male. Several Amphipoda infested with species of Spheronella as a rule only show one single female (with one or two males), others not unfrequently two or three females, mostly of somewhat different age, but here I will mention some cases of peculiarly abundant invasion. In one specimen of Calliopius leviusculus (Kr.) were found one female with eight ovisacs and five males, in another specimen six females with twenty- Jive ovisacs and two males, but the richest finds were supplied to me by Spher. paradoxa in species of Bathyporeia. In one specimen I found one large female, two small females, no ovisacs, eight males, four broad larvee, sixteen »male pupe« and one »female pupa«; in another specimen four females of widely differing size, four ovisacs, eight males, two larve and one »male pupa«; in a third animal three large females, two very small females, three ovisacs, two larve and seventeen males, the largest number of the male sex I have ever found. — The result of a large infestation is that the marsupium of the host swells to the same extent as if it were filled with its own half or almost fully developed young ones. d. Number of Species of Parasites on the same Species of Host. On most species of hosts I have only found a single species of Choniostomatidee, yet in several cases I found two species of parasites, now of the same genus, now of different genera, on the same species of hosts, sometimes even on the same specimen; nay I have happened to discover three species of parasites, not only on the same species, but on one single specimen. As the particular cases are interesting in several respects, I will enumerate them here. On Metopa Bruzelii from Godthaab (through a renewed examination with the assistance of Sars’s new important work, I have made sure that all infested specimens really belonged to this species, and not partly to Mefopa sinuata G. O. Sars) were found in the marsupium of specimens from the same locality, now Stenothocheres egregius, 70 now Spheronella Metope, but never both forms in the same specimen. In a large material of the same host from another Greenlandish locality appeared a number of specimens of the former, not one of the Jast-mentioned parasites. On specimens of Hippolyte Gaimardii M.-Edw. from the Kara Sea appeared now Choniostoma mirabilc, now Ch. Hansenii, and the latter species was also found in Hipp. polaris (Sab.) from the same sea. In Danish specimens of Ampelisca tenuicornis Lilljby. I have found Spheroncila longipes; whereas Giard and Bonnier have found in specimens of the same species from le. Croisic (Brittany) the very deviating Sphav. microcephala. Danish specimens of Protomedeia fasciata Kr. were infested with Spher. Giardii, and Greenlandish specimens with the closely related species Spher. Bonnieri. On a specimen of Iphinoé trispinosa (Goods.) from Messina Homoeoscelis mediterranea was found in the branchial cavity and Spheronella marginata in the marsupium. In Norwegian specimens of Erythrops serratus G.O. Sars and Erythrops abyssorum G. O Sars we not unfrequently find a species of Mysidion in the marsupium, and one or several specimens of Aspidoecia Normani on the outside of the body of the same species. In one specimen I found two specimens of the latter species on the outside of the body, an adult and a young female of Mysidion abyssorum, and a young female of Mys. commune in the marsupium, which makes three species of parasites on one animal. e. Number of Species of Hosts of the different Parasites. The answer to this question presupposes the answer to another, namely that of the limitation of the parasitic species, and as this latter question is not treated in detail till later on in a separate chapter, I will content myself with mentioning the results of this examination, using them in answering the question indicated by the heading. Twenty-nine of my species, as a matter of fact, are only found each on one particular species. Of the remaining fourteen species eight (belonging to the group of Spher. Leuckartii) are also limited each to one particular species; however, it may perhaps be questioned whether these eight species can really be maintained as such, or must be regarded as chance varieties of a single species (see later on). The remaining six species must be mentioned each separately. Choniostoma Hansenii has been ascertained in two species: Hippolyte Gaimardii M.-Edw. and Hippolyte polaris (Sab.). Of Spheronella insignis indeed I have only had a small material which was taken on Diastylis cornuta Boeck and Diastylis levis Novm.; however, the difference between the parasites of the two species was so slight, that my experiences from other species led me to consider them as belonging to one species. The species found by Giard and Bonnier on Ampelisca tenuicornis Lilljbg. from le Croisic, which they have described under the name of Spher. microcephala, as far as I can see, is identical with the species described by me in this work under the same name, and this is taken on Ampelisca typica Sp. Bate in Danish waters. Spheronella paradoxa I have found on Bathyporeia norveyica G. O. Sars, Bathyporeia pelagica Sp. Bate and B. Robertsonzi Sp. Bate, and even if it were proved that the two 71 last-mentioned cannot be maintained as separate species, but must be considered as one, at any rate this parasite has been taken on two good species of the same genus. But now we come to a remarkable fact, viz. that three adult females haye been taken on three specimens of Perioculodes longimanus (Sp. Bate) belonging to a different family altogether, and these females I have not been able to distinguish from those taken in Bathyporeia. If some day the male of the parasite is found in Perioculodes, probably the interesting question will be settled, whether the same parasite can be found in animals belonging to such widely differing families. Mysidion commune I have found on the following three species: Pare- rythrops obesus G. O. Sars, Erythrops serratus G.O.Sars and Er. abyssorum G. O. Sars; moreover, it seems likely that it will be found in some other species of Hrythrops living in Norway. Finally, I will state as my opinion that the parasites living on all five Nor- Wegian species of the genus Hrythrops, belong to the same species: Aspidoecia Normani. I will sum up by stating what I consider as an established fact, namely that several species of Choniostomatidz live each on two or more species of the same genus or of two closely related genera, probably even to a considerably greater extent than I have been able to ascertain; perhaps, in exceptional cases, they may be found on animals of different families. On the other hand, I certainly think with regard to several species, that each of them infests only its particular species of hosts, and this result exactly agrees with what is known about the biology of other parasitic families belonging to Arthropoda. Only with regard to Epicaridea, Giard and Bonnier have made the assertion that each parasite has its particular host and is found on no other species, an assertion which I think is incorrect, seeing that this division also comes in under the rule which applies to Choniostomatidz'). f. Occurrence together with Parasites of other Orders or Classes. In nearly all orders of Malacostraca have been found species of the group or family Epicaridea belonging to Isopoda. Giard and Bonnier have expressed the opinion that there exists a certain connection between Choniostomatide and Hpicaridea; but this objectionable hypothesis I have mentioned above in detail (p. 11—12), and at the same time I have stated all I know from my own experience and most of what has appeared in literature about the occurrence of Epicaridea on the species which, according to our present knowledge, are infested by Choniostomatide. I have written that of Cumacea, Isopoda and Amphipoda together, I have seen about 240 specimens belonging to thirty-eight different species infested with Choniostomatide, but that neither in any of these 240 specimens, nor in any other of the thousands of animals belonging to these thirty-eight species, have I found one single specimen of any Epicarid. In Italy Della Valle has found two specimens of Ampelisca 1) It may be added here that in my treatment of the Malacostraca from the “Ingolf”expedition I shall give more detailed information about the arctic Epicaridea. 12 diadema Costa infested with a species of Spheronella (S. diadema Giard and Bonnier, without description) and two other specimens of the same Amphipod with Podascon Della Vallei Giard and Bonnier. Of Decapoda there have only been found as yet Choniostomatide on Hippolyte Gaimardii M.-Edw. and on Hipp. polaris (Sab.); both these species, we know, are not unfrequently infested, either with Hemiarthrus abdominalis (Kr.) under the abdomen, or with Gyge Hippolytes (Ky.) in the branchial cavity. In a specimen infested with Choniostoma J. Sparre Schneider has observed a specimen of Hemiarthrus, and on a specimen of Hipp. Garmardii, under a large swelling on the right side of the animal, I have found eleven ovisacs and an adult female of Chon. Hansenii, as well as an adult male and a tiny female (smaller than the male) of Gyge Hippolytes, and besides, under the apparently normal left side of the carapace, a male of Gyge. In the species of the genus Erythrops (order Mysidacea), G. O. Sars has found the Epicarid Aspidophryxus peltatus G. O.8., and Giard and Bonnier have reserved this name for the form found on Er. erythrophthalmus (Goés) (EL. Goésii G. O. 8.), and established a new species, A. Savsi7 G. and B., for the form which lives on 2. microphthalmus G.O.S. Giard and Bonnier have found their type specimens of the latter species (the value of which future examination will have. to decide) on the same specimen as their type specimens of Aspidoecia Normani, and I have found a specimen of Aspidophryxus on an Er. erythrophthalmus sent to me by Prof. Sars as infested with Aspidoecia, but this parasite must have fallen off before the animals were sent to me, if it was ever there at all. The genus Sylon Kr., belonging to Rhizocephala, lives, as we know, on the ventral side of the abdomen in some species of Hippolyte, but I have not found it on any specimen infested with Choniostoma. The remaining species belonging to Rhizocephala and Copepoda, which are parasitic on Malacostraca, have all been taken on forms on which no Choniostomatidee have been found. Together with Spheronella paradoxa 1 have found repeatedly in the marsupium of Bathyporeia several specimens of a species belonging to the family Tyroglyphide (the order Acarida). Of other parasites I have only detected some Protozoa on the branchiz and the marsupial plates in the material of Malacostraca with Choniostomatide examined by me. So I have arrived at the result that there exists no connection at all between Choniostomatidee and any of the other parasites of different orders found on the same species of hosts. g. Nourishment. I cut a bit of skin of an Hrythrops in the place where an Aspidoecia had been attached, cleaned it with caustic potash from muscles and viscous substance, and discovered a small hole, where the mouth of the parasite had its place. And this seems very natural indeed; in the first place, the mandibles appear quite well qualified for producing such a hole; secondly, the mouth is doubtless suctorial, and lastly, the female, which is permanently 73 attached, must draw her nourishment from the host in order to be able to grow to a bulk which, adding its own volume after having finished laying eggs, to that of the eggs it has produced, is frequently hundreds of times larger than the volume of the larva at the time when it attached itself, and after having fixed itself it is impossible for the animal to procure the nourishment necessary for this enormous growth in any other way than by a hole worked through the comparatively solid skin of the host, whose blood must form the food of the parasite. Hereby we have found a fixed starting-point in this question, and it is more than probable that the females of all the other Choniostomatide also grow and nourish themselves by sucking the blood of their host through a hole they have gnawed. At the same time, it seems rather probable that the females of many of these species, either voluntarily or invo- luntarily, e. g. by pressure of another specimen, or by the bulk of ovisacs, are pushed out of their place and have to gnaw a new hole for themselves. I have frequently found a female in such an attitude relatively to some of the ovisacs it had laid, or the ovisacs arranged in such a manner as to make me suppose that the animal had changed place. How far the males of this family take food, I do not know, but as their mouth is as well developed as that of the females, it seems likely that they do it while young, and perhaps not when they are old (about their growth, s. above on pag. 57—58). I consider it rather doubtful whether the larve take food, but I am quite certain that the pupxe, which are provided with a mouth, and about whose considerable growth several facts have been stated above, nourish themselves in a way similar to that of the females. h. The Influence of the Parasites on their Hosts. Giard and Bonnier have proved that parasitic Crustacea of different groups (as Entoniscine, Rhizocephala) cause a »castration parasitaire« in their hosts. In the last of their papers quoted above they mention Della Valle’s untenable hypothesis that Spheronella eats the eggs of its host, and they maintain that this suggestion is wrong, and that this is also a case of »castration parasitaire«, after which they continue: »Dans des cas trés rares, Vhote ayant été infesté tardivement, cette action [namely the castration] ne s’exerce pas aussi énergiquement, et quelques oeuis peuvent étre pondus et fécondés, comme nous l’ayons vu une fois chez Clypeoniscus ja genus belonging to Epicaridea which they have treated in the same paper, and which they use as example and parallel|, mais ce sont la des exceptions. En général, lhéte est infesté avant qa’il ne soit arrivé a l'état adulte. Sous l'influence du parasite. son développement génital est arrété sans que la croissance discontinue, de sorte qu’a l’époque ou devrait se produire normalement la maturité sexuelle, la progéniture ]égitime est remplacée par le parasite et les embryons de celui-cic«. This explanation, on the whole, agrees well with the numerous data which I have given above on p. 65— 68 in the division about the age and sex of the hosts, from my observations about my thirty-eight species 10 74 which live in marsupia. In the afore-mentioned place it is stated that only in six cases I found the eggs of the host together with a parasite, and it may be added here that in almost 160 cases I only found parasites (one or more specimens), but no eggs of the host, in the marsupium. But at the same time I have stated that I have found no parasite on any specimen of Cumacea or Isopoda in which the marsupium was wanting, and of Amphipoda I have only found altogether twelve specimens belonging to four species, whose marsupium was either wanting or only half developed, and on which, nevertheless, I found parasites; all these, without exception, were half-mature or tiny females, or larve (one male). However, this material is too small to allow me fully to adopt the opinion of the authors that the host »en général« is infested before its maturity; in the fully developed marsupium of several specimens of Corophium crassicorne Bruz. I have found only a half-developed or still younger female and either a male, a larva, or a pupa, but no adult females. But it is certainly an established fact that, as a rule, we notice a »castration parasitaire« in the hosts which have parasites in their marsupium. It is much more doubtful whether the four species living in the branchial cavity of Hippolyte and Cumacea, are usually, or sometimes, capable of causing a » castration parasitaire« of their hosts. In page 66, and particularly later on, in the systematic part, detailed infor- mation is given about the age and sex of these hosts, but I do not think we can draw definite conclusions from these statements. I have said that I have found specimens of Diastylis lucifera containing a female of Homoeoscelis with the full number of ovisacs in the branchial cavity, and the young ones of the host in the marsupium, but knowing neither the time required for the development in the marsupium of the eggs and young ones of the host, nor the time required by the parasite for its own growth and the laying of all its ovisacs, it seems impossible to me to draw a definite conclusion with regard to a »¢a- stration parasitaire«; however, that such a castration may take place, seems to follow from the above-stated fact that the marsupium of more than two thirds — namely thirty-one — of the infested adult females was empty, whereas in thirteen specimens it was filled with young ones; yet it must be observed that in the non-infested females, the marsupium was found to contain eggs or young ones in 182 specimens, whereas it was empty in 74 specimens. — It is rather probable that Aspidoecia Normani causes a castration, as the marsupia of the four infested females which were not occupied by DZysidion, were empty. The four species living in the branchial cavity of Hippolyte and Cumacea cause a swelling of the carapace of the host, which, to begin with, increases in size with the growth of the parasitic female, and thus — at least in Cumacea, and probably also in Hippolyte — continues gradually increasing with the number of ovisacs. In Hippolyte the swelling may reach the same size as if it enclosed a large Gyge, and in the subsequent description of Choniostoma Hansenii, it will be proved that this parasite itself produces the swelling. In Diastylis lucifera the swelling assumes a somewhat other shape, though it may be very conspicuous. 75 Finally it may be added, that in those specimens of Hippolyte in whose branchial cavity was found an adult female with ovisacs of Choniostoma, the gills belonging to the two foremost pairs of trunk-legs were either somewhat reduced, or had quite disappeared, and in one case the three other gills situated more behind appeared a little curled, as a sign of degeneration. In one specimen, which was infested with several smaller females, the gills were slightly curled. i. Geographical and Bathymetrical Distribution of the Family. Of the forty-three species described here, sixteen come from Denmark, and one of these was also found at le Croisic; eight are from Norway, eight from West-Greenland, and two of these also in the Kara Sea, three only from the Kara Sea, besides the two which are also found in Greenland, four from Sicily, two from the Pacific near the Hast-coast of Asia, one from the Cape, and one from the West-Indies. It may be added that in the most northern part of Norway and in the Barents Sea there has been found an undetermined species of Choniostoma, though no doubt one of those which occur in the Kara Sea, that one species (Spheronella Leuckartii Sal.) was taken at Naples, another species (Homoeoscelis sedentaria (Bonn.)) in the Atlantic off Gascogne, and besides, a non-described Spher. diadema G. and B. in the Mediterranean. Without counting Norway, whose species — with one exception — have been discovered and sent me by Prof. G.O.Sars, the number of species from the different countries and seas is nearly proportional to the material which our museum possesses of Amphipoda and Cumacea — the two orders in which most parasites have been found — from the same localities. From this fact we can pretty safely draw the conclusion that the examination of a manifold larger material than was at my disposal, e. g. from the Mediterranean and from eastern Asia, will bring numerous unknown forms to light. The above statements also suggest the probability that the family is distributed over all seas, though of course I cannot form any precise notion how far its geographical distribution will extend, when some day the chief bulk of the existing species has been discovered, for, as stated in the preface, I think there can be no doubt that their number amounts to hundreds. Unfortunately I am not prepared to give many data concerning the bathymetrical extent of my species, for in most cases I lack precise indication of the depth in which the special hosts have been found. Most of the sixteen Danish species and four species from the Mediterranean were taken in a depth of between a few and twenty-five fathoms, but unfortunately I cannot state precisely in how shallow water the most littoral species is generally found, or may sometimes be found; it may be stated, however, that a specimen of Bathyporeia with its parasite, Spheronella paradoxa, was taken ina depth of between eight and eighteen feet. Stenothocheres egregius was taken in a depth of forty fathoms, a few specimens in a depth of between forty and sixty fathoms; my only specimen of Spher. 10* 76 curtipes came from a depth of a hundred fathoms. Possibly a large part of the specimens of the three forms which live on Mysidacea were taken as far down as one to two hundred, some even to three hundred fathoms, but as the hosts in question may occur in a depth of less than a hundred fathoms, I cannot say anything definite about them. So it may be stated here, that in the material from the »Ingolf«expedition I found a couple of specimens parasitic in Cumacea, from a depth of respectively a thousand and thirteen hundred fathoms, which sufficiently proves that the family also occurs in a very considerable depth. j. Geographical Distribution of the particular Species relatively to that of their Hosts. It appears from the preceding paragraph that each of the forty of my species was taken in one particular country (as Denmark, Norway, Sicily etc.), or m a particular, com- paratively smaller sea. Only three species (Spheronella microcephala, S. decorata and Choniostoma Hansenii) show a wider distribution. As a matter of course, we shall gradually find that most species of Choniostomatide have a much wider range than is known at present, but from this I do not think we can draw the conclusion that they are as widely dispersed as the species on which they live. In a smaller material of Iphinoé trispinosa (Goods.) from Messina I found seven specimens with parasites, one specimen even with two parasitic species, but in a larger material of the same species from Denmark I have been unable to find a single parasite. In a pretty considerable material of Calliopius leviusculus (Ky.) from the Hast-coast of Asia between lat. 40° and 51°N. I found seven specimens with parasites (Spheronella Calliopii), but my examination of several specimens from West-Greenland, and numerous specimens from Denmark, gave a negative result. (Yet it must be mentioned that Sars, in his work on the Amphipoda, considers the Danish specimens to belong to another species, Call. Rathkei (Zadd.), but I doubt whether this will be maintained in the future). Several specimens of Hippolyte Gaimardii M.-Edw. from the Kara Sea were infested; at the most northern coast of Norway, Sp. Schneider, and at West-Greenland EK. Vanhéffen, have found Choniostoma in the same species, but I have examined without result a very large material of the same Hippolyte collected from numerous localities near the Danish coast, and I feel inclined to think that these parasites do not occur in the waters surrounding this country. The same may be said about Diastylis Rathkei (Ky.), of which in about half of the females with marsupium from West-Greenland and from the Kara Sea, I found Spheronella decorata, whereas of the same species from different Danish localities I have examined at least several scores of adult females without finding any parasite. k. Frequency of the Parasites in proportion to their Fertility. In mentioning the eggs of the parasites, I pointed out (p. 46—47) the enormous difference of fertility between the species with very small and those with very big females. Of 17 Hippolyte Gaimardii M.-Edw. I have had a large material from the Kara Sea, and one of similar size from West-Greenland; in the former I only found six infested specimens: four with Choniostoma mirabile, two with Ch. Hansenii; in the latter I found none at all, though a few specimens of Ch. Hansenii have been discovered there at least either on H. Gaimardii or on H. polaris. On the latter species [ found one Ch. Hansenii from the Kara Sea and one Chon. sp? from the coast of West-Greenland, though H. polaris is not unfrequent in the former, and of common occurrence in the latter locality. This shows distinctly enough that the genus Choniostoma with its two very large and particularly fertile species is pretty scarce in the Kara Sea and rare in West-Greenland. Most of the middle-sized and rather small species of parasites I have found in between very few and about seven specimens of their respective hosts, though my material of the latter was frequently very rich. Only of two parasites: Stenothocheres egregius and Homoeoscelis minuta have I found a great number of specimens in a large material of their respective hosts: Metopa Bruzelii (Goés) and Diastylis lucifera (Kr.). The first-mentioned parasite is the least prolific of all my species, and the second, in this respect, comes nearest to it among the species of which my material of infested hosts was sufficiently large; moreover, the hosts came from several localities. That the number of parasites cannot be determined only by the number of infested hosts, has been proved above, as of some species, e. g. the two afore-mentioned: Stenothocheres egregius and Hom. minuta, we seldom find more than one female and one male on each host, whereas on others we pretty frequently find several females and males as well as pup in one host, but the only parasite of which, from the last-mentioned reason, I have found as many specimens as of the above-named species, is Spheronella paradoxa (living on Bathyporeia, a genus of very frequent occurrence in Denmark); it belongs to the smaller species and, as it seems, does not lay more than four of five ovisacs, which are comparatively large. All these data decidedly point in one direction, but considering the insufficiency of my material, I will take good care not to lay down any rule or law which might possibly not prove quite tenable, and I will content myself with suggesting the direction. We might feel inclined to suppose that the conditions of life of most of these parasites are pretty similar, and that consequently the most prolific species would occur most frequently, the more so, as there is such an enormous difference in their fertility, that a species like Choniostoma mirabile lays at least more than three hundred times as many eggs as Stenothocheres egregius ; however, the above-mentioned examples prove in a striking manner that such a conclusion cannot be drawn. Consequently there must be circumstances to account for the fact that the two least prolific species: Sten. egregius — taken in two localities — and Hom. minuta — taken in several, probably in many places at considerable distance from each other — occur much more frequently than the prolific and very prolific species. An explanation of this fact is required, and I will attempt to explain it, at least partly. No doubt, the critical point in the life of the parasites must be the short period during which the full-grown, though very small, larve leave the ovisac and their mother’s host in order to seek a new 78 host for their own further development. and no doubt, a considerable number of the larvae of all species are destroyed while swimming about, partly because many of them cannot find their object in due time. Then again, there must be circumstances which cause a compara- tively smaller percentage of the brood of the least prolific species, and an enormous percen- tage of the brood of the most prolific species, to be destroyed during this period. This, again, must be supposed — at least partly — to have something to do with the difference of the number of specimens of the species which constitute the hosts. Now, as the larvee of the most prolific species seek large forms, those of the least prolific small forms, and as the large forms, as we know. are found on an average in much smaller number than the small forms, it follows that the larvee by which they are sought, have as a rule much less chance of finding them in due time, for, as previously stated in detail, there is no considerable ditference in the structure and size of the different larve, — e. g. the larva of Choniostoma mirabile is only about one eighth longer than that of Stenothocheres egregius, but, as far as 1 can see, scarcely so vigorous and so well adapted for swimming, (comp. the peduncles of the natatory legs in the two species; pl. I, fig. 11, and pl. XI, fig. 1 e). — On the base of my material of parasites and of my knowledge of the biology of the hosts, I might set forth several points, thus giving a wider scope to the discussion of these matters, but for various reasons | abstain from doing so. C. About Classification. a. Limitation and Characters of the Species. Of small Crustacea, such as Cladocera, Ostracoda and free-living Copepoda, there are in most cases some or many specimens of each species at the student's disposal for determination of the forms, and even where these animals are so small that the compound microscope has to be used in order to determine them, most of them can stand the pressure of a glass-cover, and as a rule it is unnecessary to submit the specimens to much parti- cular preparation, except where a description of them has to be given; finally, most species have a very fixed shape. All these factors help to facilitate the determination of the species. In Choniostomatide the circumstances are different. The animals are so rare, that of most species only a single specimen or a few specimens of each sex can be procured; neither males nor females can bear the pressure of a glass-cover; the males are so small, that they cannot be examined without high magnifying power, and before the examination particular care has often to be taken in placing them in the preparation, and though the females are much larger, the parts of their body which have to be investigated are exceedingly small] 79 and very difficult or impossible to examine in detail on intact specimens, so that in numerous cases one is obliged to undertake a difficult dissection, and to place the head and the genital area in a preparation; lastly, the general form of the body in both sexes, and particularly in the females, is far less fixed than in the free-living forms. On account of these circum- stances it is sometimes difficult to form a positive judgement about some species, e. g. in how far they present varieties of one species, or form separate species. It is sufficiently well known that a similar difficulty is not unfrequent with regard to the free forms, and trom what has been said about Choniostomatide, it is easy to understand that, with respect to this family, the difficulties are sometimes so great that a final settlement of some questions must be left to the future. It has been specially mentioned that most species of Choniostomatide have been found each on its particular species of Malacostraca, but, at the same time, I can prove to a certainty that the same species can be found on different species of the same genus (e. g. Choniostoma Hansenii on two species of Hippolyte), or even in forms of two different genera (Mysidion commune on Parerythrops and on two species of EHrythrops); and further, on the same species of host one may find two species of parasites in the branchial cavity or in the mar- supium, nay even two species in the same marsupium (Mysidion commune and Mysid. abys- sorum in Er. abyssorum). The result hereof is that we cannot absolutely take for granted that we know a parasite, because we have found it on a certain host, nor that a parasite belongs to an unknown species, because it is found in a host that is not mentioned in this work. All the same, in most cases the host is of the greatest importance in determining a parasite, and where parasites are found in new hosts, most frequently they will prove them- selves to be new species. Most of the species established in this work have been easy to distinguish from each other, and in the majority of cases there has been no hesitation at all in establishing the different species. It is mentioned above that on Perioculodes longimanus (Sp. Bate) I found females which were exactly like the Spheronella paradoxa living on species of Bathyporeia Lindstr., but as the male belonging to the females found on Perioculodes is wanting, I have not been able to decide whether the same species really lives on forms of different families. On account of rather small material, I have also had a little doubt concerning the identity of the forms found on Diastylis cornuta Boeck and D. levis Norm.; but with regard to this question, as well as to Aspidoecia Normani, I refer to the subsequent special representation. The greatest difficulty I met with in the species very closely allied to Spheronella Leuckartii Sal. Of these species I have established eight, taken in six genera belonging to four different families of Amphipoda, and four of these species of hosts (belonging to four different families) came from Denmark, two from Sicily, one from the West-Indies, one from Hong- Kong. The difficulties were so great, that I hesitated for a long time whether to establish them each separately, or as belonging all to one species. Though this question will be treated more thoroughly in the systematic part, I thought it right to call attention to it here. 80 We now come to the question concerning the characters of the species. It appears hat everywhere, except within the just mentioned group of Spher. Leuckartii, the males offer a considerable number of excellent and, as a rule, easily observed characters. In this respect they generally surpass the females; they are not only easier to examine, but the shape and decoration of the frontal margin, the processes from the sub-median skeleton, their often very peculiar two pairs of trunk-legs and the hair-coat of their trunk frequently afford excellent characters, which do not occur in the other sex. Other distinctive marks are not unfrequently found in the antennulz, the maxille and the maxillipeds, in the presence or in the want of caudal stylets, and in the former case, often in the length of their longest terminal seta etc. But, at the same time, we must point out that minor differences in the general shape of the body, in the distance of the caudal stylets from the posterior extremity, in the length of very long sete, are frequently seen in specimens of the same species; also, that such a feature as the rostrum protruding or receding may give a very different appear- ance to the animal. In the females the chief characteristics are found in the structure and the organs of the head, especially the antennule, the maxille and the maxillipeds, and in adult specimens usually in the genital area, (in Mysidion, however, in the arch round each genital aperture). As a rule the heads of the females are much more uniform than those of the males, and must be examined with great care; the size of the genital area compared with that of the head, its form, the extent of the solid chitine, as well as its hair-covering or want of hair-covering, often afford good characters; nevertheless, it is often necessary to make a prepara- tion, and as far as my experience goes, one must frequently pass over several smaller differences, as some variation may be found in the same species. The caudal stylets, their position etc., or the want of them, is always of importance. Within the genus Spheronella, the females of many species have trunk-legs, whereas these appendages are wanting in others, but in adult specimens they are often so hard to find, that the character drawn from their presence is not easy to make use of, and one has to examine the animals very carefully before being able to deny their existence. In the adult females of very few species we notice a peculiar haircoat, but it must be remembered that in Choniostoma Hansenvi the younger specimens are more hairy than the adults, and e. g. in Spheronella danica, the trunk of the young ones is closely covered all over with hair, that of the adults mostly or totally naked. In the females the general shape of the body is sometimes rather characteristic, but often rather variable according to chance circumstances, e. g. some kind of pressure, or the periods of beginning and ceasing to lay eggs. But, in order to find good characters of the species, we are not confined to the males and females only. In a large material it is not difficult to procure larve of a number of species, partly free specimens, partly in preparing those contained in one or some of the ovisacs. Of all the five species found in the marsupium of Cumacea, the larve are known to me, and these not only differ from all other larve yet found, they also show very distinct differences among themselves, particularly in the frontal decoration and in the two distal 81 joints of the maxilla; however, I may also observe that I have found very considerable difference in the length of the olfactory seta of the antennulze between larve of the same species (Spher. modesta). Most other larve we know also differ from each other in a number of features: length of the olfactory seta of the antennule, structure of the antenn, the relative size of the abdominal segments and of the caudal stylets, and the length of their long setw, sometimes (Mysidion) also in the presence or absence of fine processes on the terminal joint of the maxillipeds. However, in one case, namely in the genus Chonio- stoma, I have not been able to find any difference between the larve of the two closely allied, yet distinctly separate species. — The size of the ovisacs and the size of their eggs compared with the female present considerable differences between the species, yet they do not naturally form good distinguishing marks. Finally, the few pupze known to me differ very much according to species, except those belonging to the group of Spher. Leuckartii. b. Limitation and Characters of the Genera. All the species may be classed under six genera, which offer an almost regular eradation in the reduction of the females. The males too become considerably degraded, but not to such a degree as the females. This gradual reduction is combined with great changes in the way of laying the eggs, whereas there are very little differences and no reduction at all in the structure of the larve, and the post-larval development is too little known to allow of making general statements about it. Consequently, the arrangement of the genera in the systematic part is easily and naturally carried out in considering the gradual reduction indicated. The first genus, Stenothocheres, deviates from all the following by possessing a distinctly prominent abdomen, by the more conspicuous and distinctly two-branched trunk-legs of the females, and by the way they lay their eggs: in one or two (rarely three) free lumps of indefinite form — not in ovisacs, where the eggs are surrounded by a distinct common mem- brane. In the two first mentioned characters the genus approaches the less reduced forms of Copepoda. In the other five genera there is no abdomen, the trunk-legs of the females are small with at most one distinct branch, generally consisting of one single joint, or they are altogether wanting, and the eggs are deposed in several or in numerous ovisacs. These five genera are naturally divided into two groups: in the three first, viz. Homoeoscelis, Spheronella and Choniostoma, the genital apertures of the females are situated close together and surrounded by a more solid plate, ring or semicircular list, inside which are also found the entrances of the two receptacula seminis; the ovisacs, when laid, are free, not attached to the female. In the other group: JMJysidion and Aspidoecia, the genital apertures of the female are often placed at a greater distance or very far from each other; each has its crescent or ring, and far in front of them is situated one receptaculum seminis; the ovisacs are hinged on the lips of the genital apertures. In the first group most females 11 82 and most of the males known have 3-jointed antennule, distinct antennz, distinct and, in the males, often very considerable trunk-legs as well as caudal stylets, whereas a few species, through a reduction of these parts, form a transition to the genera of the last group, having antennule which are 2-jointed or quite indistinctly articulated, and no antenne, trunk- legs or caudal stylets in either of the two sexes. Another slight difference between the two groups may still be mentioned, namely, that in the males the maxillipeds, especially their distal part, is stouter and more normally developed in the first than in the second group. Homoeoscelis does not deviate very much from Spheronella, though it differs distinctly in having trunk-legs and caudal stylets, which are similarly shaped in the same individual, as well as in the two sexes, and whose form differs very much from that in the female of Spheronella; finally, the larve of this genus differ from all others in their very long antennz and in the great distance between the maxille and the maxillipeds. It is also a character of this genus that its species occur in the branchial cavity of Cumacea. In Choniostoma, unfortunately, the male is unknown; the female only differs from Spheronella in having rudimentary maxillipeds; however, this feature, as well as the fact that its species live in the branchial cavity of Hippolyte, appear to me sufficient to maintain the genus. Mysidion and Aspidoecia are distinguished most decidedly by differences in both sexes and in the mode of living, which it is hardly necessary to mention in detail, and Aspidoecia is the most reduced of all forms of the family, both sexes having 1-jointed antennule, no antenne, very small maxillule without additional branch, and, as a matter of course, no trunk-legs or caudal stylets; moreover, the maxillipeds are entirely wanting in the female, and their distal part is greatly reduced in the male. No less than thirty-four of the here described species are referred to the genus Spheronella, and these species differ very much from each other in several respects which, at least apparently, are of considerable importance: 1) Antennule mostly 3-jointed in both sexes, sometimes shorter and either 2-jointed or with indistinct articulation. 2) Antenne closely similar in both sexes, generally pretty well developed, in some species rudimentary, in others wanting. 3) Maxillule almost alike in both sexes, generally with an additional branch, some- times without it. 4) Maxille rudimentary in the female of S. marginata, well-developed in all other forms. 5) Trunk-legs and caudal stylets are good-sized in the males of most species, but are wanting in a few; these appendages are found in most of the females, though they are very small; they are wanting in some forms, and it may be said that where they are wanting in the male, they are also wanting in the female of the same species, though the reverse is not always the case. 6) The peculiar attachment of the female in S. paradoxa. — In spite of these salient differences I have not ventured to divide the genus into two or more genera, as I have been unable to discover any feature of sufficiently decisive importance. For it is easy enough to say, as many authors do, that if a species (as e. g. S. paradoxa) presents some striking characteristic, it must be set apart as the type of a new genus, but frequently we have no guarantee that such a feature is really of sufficient importance. We meet a similar difficulty where several 83 species have the distinct negative character of wanting trunk-legs and caudal stylets in both sexes, for as there are species which, though wanting these organs in one sex, not in the other, or through the male having trunk-legs, but no caudal stylets, form a transition to the species in which both sexes possess trunk-legs and caudal stylets, one cannot very well set apart the first mentioned species as a separate genus. Jf I can find no leading principle to guide me in carrying out the division of a large genus, the elements of which seem to be heterogeneous, and if I am not obliged to undertake a division in order to bring about an equivalence with previously established acceptable genera, I prefer putting off the division till the discovery of new forms has thrown new light on the question. If I had had to subdivide Spheronella, the result would have been, not two or three, but six or seven genera (some of which would have consisted of only one or a couple of species), in order to establish a pretty correct equivalence, but these new genera would not have been tolerably equivalent with such types as Homoeoscelis, Mysidion ete. c. Characters of the Family. An examination of the genera will show very clearly that, in spite of several diffe- rences, they are all very closely related and belong to the same family. We will here attempt to give a summary of all its more important characters, some of which separate it from one, some from another, of the rather numerous families of parasitic Copepoda, for it would be impossible to give a condensed characteristique with merely exclusive features, our knowledge of several points in the organisation and development of other families being too defective. The adult Females are ovate or sub-globular. The head occupies only a smaller or a minute part of the greatly swollen, unsegmented body; the abdomen is comparatively rather small and unsegmented, or mostly altogether wanting. Antennule 1-3-jointed; antennz small or wan- ting; rostrum good-sized, comparatively stout with cup- or funnel-shaped mouth provided with a border formed by a membrane which is interrupted only in front and supported outside by free hairs; maxillule consisting of a basal part almost entirely fused with the rostrum, and of two or three usually setiform branches; maxillz short and powerful prehensile limbs consisting of a stout basal joint and a slender, 1- or 2-jointed, somewhat claw-shaped, distal part; maxillipeds rarely wanting, mostly appearing as good-sized grasping appendages, consisting of a long, rather stout basal joint and a shorter, slender, 2- or 3-jointed distal part. We often find two pairs of rather small or minute trunk-legs, each of which consists of one single joint or sometimes of a peduncle with one or two unjointed branches; the legs are wanting in not a few species. Caudal stylets present or wanting. Some species, at least, can hinge themselves by an adhesive plate or a frontal thread. Spermatophores (found in many species) consisting of a globular or oval vesicle on a rather long thread-shaped stalk. like 84 The Males are several or many times smaller than the females, oblong or sub-globular. The head forms a little more or a little less than half of the unsegmented body. Abdomen nearly as in the females. Antennule, antenne, rostrum, maxillule and maxillipeds nearly similar to those of the females. Frequently, though far from always, we find too pairs of trunk-legs, which are often good-sized and two-branched, with long terminal sete, but very rarely jointed. Caudal stylets frequent. They hinge themselves by a rather long, or very long frontal thread. Development. The eggs are deposed in one or two (rarely three) free, irregular lumps, or most frequently, in several (at least four or five) or many (up to twenty-eight) ovisacs, which, as a rule, are free, though sometimes hinged on the lips of the genital apertures. The Nauplius stage is passed through in the egg; the forthcoming larva is in the first Cyclops stage, with an oval, somewhat depressed cephalothorax, which is divided far back by one articulation, and a 3-jointed abdomen with caudal stylets, each with a very long terminal seta. Cephalothorax with 2- or 3-jointed antennule, provided with a very long olfactory seta, 2-, 3. or 4-jointed antennz; rostrum in the main as in the adults, maxillulee with — as a rule — indistinct basal part and (one) two, three or four setiform branches; 2- or 3- jointed maxille and 4jointed maxillipeds, both pairs chiefly constructed as in the adults; finally, two pairs of natatory legs, each with two 1-jointed branches. Out of this larva, which hinges itself by a frontal adhesive plate, the males not unfrequently, the females sometimes, appear directly, without passing through any intermediate stage. In other species the larva develops into a pupa, out of which the male proceeds. In most species the same meta- morphosis is gone through by the female; in one case the female passes through at least one additional intermediate stage. Where a pupa is found, it is always hinged; besides it is nearly always provided with a mouth and increases considerably in size. After hatching the males grow comparatively rather little, and the same is the case with the head of the females, whereas the trunk of this sex swells excessively. A distinctive mark of this family is the above (p. 27—28) described mouth, which appears, not only in the female and in the male, but — as far as its most important features are concerned — also in the larva, and nearly always in the pupa. d. Place of the Family in the System. During the last thirty years and more, the parasitic Copepoda have been very little studied, and not a single really leading work has appeared about this subject. Several authors have established a series of genera, some of which might easily be ranged in the old families, while others stand rather isolated. Some smaller families have also been instituted. If, however, we try to get a general view of our present knowledge, we find that several families are badly defined, and others so imperfectly known, that we cannot form a definite opinion of their place in the system: whether they belong to one of the established families, or must be taken as types of new families. The reasons 85 of this uncertainty are partly, that the males of too few forms are known, partly — and particularly — that the metamorphosis of numerous genera among the old families and of the more abnormal forms is entirely unknown, and that the structure of the mouth in the adults as well as in the larve is often badly studied, etc. A revision of the classification of the parasitic Copepoda would be most desirable and ought to be based upon a thorough study of the external structure of both sexes, and upon numerous new data which throw light ou the post-embryonic development; that a representation of the internal structure of numerous types would be excellent, goes without saying, but even without undertaking this gigantic work such a revision as the above-mentioned would be exceedingly useful. However, as such a work does not exist, I do not see that it can be of much use to discuss the relation- ship of the Choniostomatide and their place in the system more in detail, so I will content myself with some few remarks. The last detailed systematic arrangement of the parasitic Copepoda was undertaken by A. GeRSTAECKER in »Bronn’s Klassen und Ordn. des Thier-Reichs, fiinfter Band, erste Abth.« p. 721—729, and this part was published about 1870. Perhaps we might also mention the more condensed grouping in »C. Craus: Grundziige der Zoologie, B. I, 1880, p. 554—58,« as it is set up by the author who has also gained great distinction in this domain of carcinology. By studying these treatments and several papers on special groups, I have found out that the family Choniostomatide stands far apart from all hitherto established families, except Lernzeopodide, from which, however, it also differs considerably. If Salensky in his often mentioned paper means that Spheronella comes nearest to Lerneide, because he thinks that in the structure of the mouth and in the form and position of the maxille and the maxillipeds, it resembles Lernea branchialis in the pairing stage, we admit indeed that the resemblance in the structure of the mouth is doubtless very striking, but in other respects the various larval stages of Lernea and Pennella differ widely from the larve and pup of Choniostomatide, and the subsequent development of the two genera of Lernzxidz, as we know, differs so thoroughly from that of the Choniostomatide, moreover, the structure and egg-laying of the female of Lernea is so exceedingly different from these features in our family, that a closer relationship is entirely out of the question: in my opinion Lerneidz and Choniostomatide stand very far from each other. But undeniably it stands even farther apart from Herpyllobiide, though Giard and Bonnier have attempted to unite it with this most remarkable family, which differs widely from all other parasitic Copepoda. They do so by establishing a new family: Spheronellide, which they subdivide into Choniostomatin and Herpyllobiine. This peculiar classification I have criticised at length in my general historical view (p. 15—21), to which I refer. The same two authors, in their earlier work, published in 1889, say that Choniostomatidz comes nearest to Chondracanthidze, Lernzeopodidz and Ascomyzontide. The first and the last of the families in several respects — e. g. in the structure of the mouth — deviate so much from Choniostomatide, that any closer relationship is out of the question; indeed our family stands widely apart from both, 86 but as for Lernzeopodide, there is a considerable resemblance in various points, e.g. in the structure of the male and the female, and especially in that of the larve; at the same time, there are numerous and important differences. However, as the family Lernzopodide is comparatively well known‘), I do not think it necessary to repeat and compare all the characteristics of the two families, but will content myself with stating my opinion that the Choniostomatidz, though coming much closer to the Lernzopodide than to any other form of parasitic Copepoda, yet differ very much from them in the way they lay their eggs, in their development after the first larval stage, in several peculiarities in the internal and external structure of the male (e. g. in that of the mouth), and most conspicuously, in the structure of the mouth, the antennz and the maxillipeds of the female. A comparison of the figures in: W. Kurz: Studien wiber die Familie der Lerneopodiden (Zeitschr. fiir wiss. Zool., B. XXIX, 1877)« with my present work, will give the best idea of the resemblances and the differences between the adults of these two families. In elucidation of the matter I will add, that Kurz concludes from the development that the pair of limbs which in the females of Lernzopodide are fused together into one long arm that serves as organ of fixation, are the »first pair of maxillipeds.« If this be correct — which is quite possible — this appendage would correspond to what I term the maxille. 1) In the above-mentioned ,Grundziige* (p. 557—58), Claus enumerates most of its characteristics and refers to the most important accounts of its structure and development. IV. DESCRIPTION OF GENERA AND SPECIES. Conspectus of the Genera, based on the Females. A. Abdomen is found, it protrudes from the trunk, is comparatively rather small, unseg- mented and not set off by an articulation. Eggs are laid in one free lump or in a two (or three) lumps without distinct shape. (Live in the marsupium of Stenothoide, a family@oteAriphipoda)\ ewe ee usine a cmonet a ceed cate rnnne I. Stenothocheres n. gen. B. Abdomen is wanting, though caudal stylets are frequently found. Eggs are laid in a smaller or greater number of ovisacs. a. Genital area is found; genital apertures close together and surrounded by a common plate, ring or semi-circle, which is more solidly chitinised than the remainder of the skin. The entrances into the two receptacula seminis are situated within the genital area. The ovisacs are deposed freely. «, Trunk-legs and caudal stylets apparently consist of a very small, short basal part which tapers into a comparatively rather long and very narrow conical branch. Live in the branchial cavity of Cumacea. . II. Homoeoscelis n. gen. 6. Trunk-legs and caudal stylets not unfrequently wanting; if found, they are very small, generally sub-cylindrical and terminating in two — the caudal stylets sometimes in one or three — sete. Live in marsupia. §. Maxillipeds good-sized and at least always longer than the maxille. Live in the marsupium of Amphipoda, Cumacea and Isopoda. III. Spheronella Sal. §S. Maxillipeds quite rudimentary, several or many times shorter than the maxilla. Live in the branchial cavity of Hippolyte Leach, a genus of OGY Ws EimhGi S ce th couture nen fare aterm betes IV. Choniostoma H. J. H. b. Genital area wanting; the genital apertures often situated at a considerable distance from each other, each having its own arch or ring of solid chitine. The odd recep- taculum seminis is far from the genital apertures. The ovisacs hinged on the lips of the genital apertures. 88 a. Maxillipeds good-sized. Genital apertures very far from each other, placed very near the line where the hind margin and the lateral-margin meet. Live in the marsupium of the genera Hrythrops G. O. 8S. and Parerythrops G. O. Sars, belonging to Mysidacea......-.........- V. Mysidion n. gen. 6. Maxillipeds wanting. Genital apertures closer together, somewhat up on the dorsal surface. Live attached outside on the back and on the sides of the body or on the eye-stalks of the genus Hrythrops G. O. Sars, belonging to Mysidacedices 25). a een ee VI. Aspidoecia Giard and Bonn. ~ I. Stenothocheres n. gen. FEMALE. The body somewhat longer than broad; naked all over. The head which is comparatively good-sized, is not marked out from the trunk; it has neither frontal nor lateral borders. Antennulz: comparatively long, without distinct articulation and with few sete. Antennze middle-sized. The mouth-border with very short hairs. Maxillule without additional branch. Maxillz robust, without hairs. Mavxillipeds of scarcely medium size, second and third joints coalescent, all joints without hairs, spines or processes. Sub-median skeleton very feebly developed, consisting only of a plate in front of each maxilliped. The trunk-legs are situated rather far from the lateral margin; they are comparatively of considerable size and consist of a basal part with two unjointed branches, each as a rule ending in a strong or spiniform seta, besides, on the outer branch of the first pair of legs, is found a smaller seta. Abdomen pretty well developed, consisting of a robust basal part and a narrower distal part, which passes without articulation into two short and broad caudal stylets, each of which bears four sete of unequal length. Genital area wanting. Genital apertures situated partly on the lateral surface, partly on the ventral side of the basal part of the abdomen, more or less close to its lateral margin. Two large receptacula seminis, the entrances of which have not been found. — Spermatophores have not been noticed. MALE. Agrees with the female in most features, so that only some characters need be pointed out. The head is not marked out, it occupies about half of the body. The trunk-legs are situated at a short distance inside the lateral margins of the trunk; both pairs are comparatively very considerable, with a thick basal-part. A little behind and outside the second pair of trunk-legs proceed a pair of spines (pl. I, fig. 21, x), which are possibly the rudiments of a third pair of legs. Only the narrower, distal part of the abdomen can be distinguished from the trunk. The body is naked, but on its dorsal side, and especially on its front part, it is furnished with moderately small, irregular protuberant knots, and on 89 the ventral side, particularly outside a line between the base of the antennule and the first pair of trunk-legs, with peculiar, irregular stripes or with keels and knots. OVISACS. Real ovisacs are not found; the eggs are laid in one free lump or in two (or three) lumps. LARVA. Is only known of one species. Antenne about the length of the anten- nul. The maxillule have a short but distinct basal part and two stout branches, the anterior of which is very long. The maxillipeds are situated closely behind the 3-jointed maxille; all joints of both pairs are smooth. The peduncle of the natatory legs is very broad. The third segment of the abdomen, together with the broad caudal stylets, which are not set off by an articulation, is almost as broad as and nearly double the length of the second segment. The longest seta of the caudal stylets is only a little longer than the abdomen and not nearly half the length of the cephalothorax. POST-LARVAL DEVELOPMENT. Unknown. HABITAT. The two species known live each in a species of the genera Metopa Boeck and Stenothoé Dana. Hitherto found only in Norway and at the western coast of Greenland. REMARKS. My material of the Amphipod family of Stenothoide, which contains a great multitude of species, being somewhat limited, because I have only seen a few spe- cimens of most Greenlandish and Danish species, it may be expected that researches made in a large material will lead to the discovery of a number of new species of this inter- esting genus. Conspectus of the Species. The basal joint of the maxillipeds in both sexes conspicuously longer than that of the maxilla. The female without median frontal process. The male elongated. 1. St. egregius nu. sp. The basal joint of the maxillipeds in both sexes almost shorter than that of the maxilla. The female has a median frontal process. The male is short and broad. 2. St. Sarsii n. sp. |. Stenothocheres egregius n. sp. (Pl. I, fig. 1a—11) FEMALE. A very large specimen is ‘63 mm. long. The specimen represented (fig. 1a and fig. 1b) is ab. 59 mm. long and -46 mm. broad. The body as a rule a little longer than broad, apart from the abdomen evenly rounded; seen laterally (fig. 1 b), the ventral side is rather flat, the back strongly convex. No median frontal process between the bases of the antennule. The antenne distinctly jointed; the terminal joint furnished with two short sete of unequal length. The basal joint of the maxillipeds conspicuously longer than that of the 12 90 maxillz. Each maxilliped proceeds from a chitinous list (fig. 1 e,h), projecting between the bases of the maxilla. The proximal part of the abdomen two to two and a half times broader than the distal part. Each of the triangular caudal stylets (fig. 1g, t) has four setz, the foremost of which is short, the apical one moderately long. MALE. A well developed specimen is 196 mm. long and -11 mm. broad — thus good-sized in proportion to the female (fig. 1c¢: fig. 1a). So the body is a good deal longer than broad (fig. 11). On the ventral side, stretching from the base of the antennulzx outside the maxilla, the maxillipeds and the legs, backward towards the abdomen, and from the appendages towards the lateral outline, are found a comparatively small number of irregular stripes or grooves. On the basal part of the first pair of legs we see some irregular pro- jections and taps; similar though blunter taps or knots are spread more scantily over the dorsal side of the animal, whereas the frontal part is closely covered with larger knots. Antennulz shorter than in the following species. The basal joint of the maxilJipeds longer than that of the maxilla and more robust than in the following species. In the second pair of legs the apical spine on the inner branch is frequently somewhat curved, but not hooked. Of the setz on the caudal stylets, the apical one is thick and longer than the others. — A frontal thread was found in a few cases; it was about as long as the animal (fig. 1 ¢), simple and somewhat dilated towards the distal end. HGGS. They are very large (fig. 1d compared with fig. 1a), and are deposed in a large, loose, irregular lump, or in two (or very rarely three) lumps; the greatest number found is forty-two, the usual number is about thirty. LARVA (fig. 11). Length of the body (except caudal seta) 22 mm., which shows that it is longer than the male, though its volume is somewhat smaller. Cephalothorax oval, somewhat longer than broad. The front has a transverse band which curves backward, ending at a short distance from the base of the antennule. Antennule 3-jointed; olfactory seta at least double their length, reaching a little behind the middle of the cephalothorax. Antenne of medium length, 3-jointed; basal joint broad and longer than broad, about the same length as the second joint; third joint short, terminating in two or three sete, one of which is stout and as long as the second and third joints together. Second and third joints of the maxillipeds of about equal length. The longest seta at the hindmost angle of the first abdominal segment a little longer than the second segment. Hach caudal stylet fur- nished with five sete, one of them a little longer than the abdomen, two of the others a little longer than the last segment plus the caudal stylets. (See besides the diagnosis of the genus). POST-LARVAL DEVELOPMENT. Unknown. HABITAT. On Metopa Bruzelii (Goés) from two localities near the western coast of Greenland. In a glass labelled: »Godthaab, deep water {probably 40—60 fathoms], in Sertularia, Holboll«, were found numerous specimens of Met. Bruzelii (Goés) and of M. sinuata G. O. Sars, as well as a number of specimens of M. longicornis Boeck, M. longimana Boeck 91 and MW. neglecta H.J.H., and an examination of all these only exhibited parasites on eight adult females of MW. Bruzelii; in five specimens this Stenothocheres was found, in three others Spheronella Metope un. sp. (s. later on). The second locality is: »lat. 66° 30’ N., long. 54° 50’ W., forty fathoms, stones with many Balani, 5. VIII. 1886, Th. Holm«; here was taken a very great number of Met. Bruzelii and of M. sinuata, but whereas the latter is free from parasites, Stenothocheres occurs frequently in the first mentioned species, while no specimen of Spher. Metope was found. Unfortunately the abundant material from this locality was somewhat roughly handled, as the animals while still alive had been put into too strong spirit. I investigated and put down statistics on the contents of the marsupia of twenty-three infested females, which, added to those from the former locality, makes a total of twenty-eight. In one case neither females nor males were found, but at least twelve larve and a lump of six eggs without larve. In another case only one not half-developed female was found; in a third marsupium two females, but neither males nor eggs. In twenty- five cases a female was found, and often a male besides, in one case even two males (once I also found a normal male and the larger part of the skin of a dead male), and finally, I frequently met with eggs or recently hatched larve. Concerning the eggs, I refer to the description given on p. 44. The female was always seen in the foremost part of the mar- supium, the male and the eggs behind. Only in a few cases a frontal thread was found in the male. Subsequently more material of Met. Bruzelii was examined (adult females as well as young specimens), in order to find — if possible — stages of development. Several finds in adult females corresponded to the above stated results, but in one young female with scarcely half-developed marsupium I succeeded in finding two larve which had evidently swum in beneath the body of the animal, and in a young specimen without marsupium I found a single larva. From all these data it may be concluded that, at least as a rule, the female is infested before the marsupium is quite developed. At least one of the last-men- tioned larve had hinged itself by the usual adhesive frontal plate, but about the subsequent development I learned nothing. 2. Stenothocheres Sarsii n. sp. (Pl. I, fig. 2a—21.) FEMALE. The largest specimen (fig. 2a) is swollen to such an extent that the body is vaulted beyond the abdomen, so that this part does not add to its length or breadth which are respectively “80 mm. and ‘69 mm. The specimen exhibited in fig. 2d and fig. 2e is only ‘67mm. in length. Seen from below, the body (apart from the abdomen) is a short, at the ends rather flattened oval; seen laterally (fig. 2), the ventral surface is rather flat, the back considerably vaulted. Between the base of the antennule is found an odd, blunt, horizontal process of considerable size. The antenne weak, with indistinct articulation, the distal 12* 92 joint terminating in a short spine. The maxillipeds comparatively short and slender, their basal joint almost shorter than that of the maxilla. The proximal section of the abdomen more than three times broader than the distal section (fig. 21); each of the oblong caudal stylets furnished with four setae, the foremost of which is the longest. — Fig. 2a shows that this specimen possessed two long frontal threads (s), the proximal parts of which are united into one single thread. MALE. The largest specimen (fig. 2k and 21) measures to the extremity of the caudal stylets -27 mm. in length; breadth 24 mm.; a rather considerable size compared with the female (fig. 2b: fig. 2a). So, the body is proportionally only a little longer than broad and somewhat depressed. The ventral surface outside the limbs from the base of the antennule to the abdomen, the posterior part of the sides, the hindmost part of the back, the stout basal part of the trunk-legs and the inner branch of the first pair, are closely covered with peculiar, very irregular eminences, which are partly shaped like keels, partly like knots or short, acute taps. The median part between the maxillipeds and the trunk- legs shows fewer keels and stripes. The dorsal side has very few knots, whereas the front part of the head is covered with numerous blunt knots. The maxillipeds like those of the female. The terminal spine on the inner branch of the posterior legs is strongly curved and hooked. Of the set of the caudal stylets, the two apical ones are stout and of sub- equal length. — The frontal thread (fig. 2k, s) a little shorter than the animal, simple, slender, yet somewhat thickened towards the distal end. EGGS. Much smaller, but also much more numerous than in the preceding species. In one female were found two somewhat oblong lumps of eggs of about equal size and of irregular shape; they did not show any trace of larvee. One of these lumps is exhibited in fig. 2c, enlarged to the same scale as the largest female, fig. 2a, and a comparison of these figures with the male (fig. 2b) will show the relative size. LARVA and POST-LARVAL DEVELOPMENT. Unknown. HABITAT. On Stenothoé marina (Sp. Bate) from Norway. The locality cannot be precisely indicated, but, according to Sars, the host occurs along the southern and western coast of Norway up to the Nordland coast (Tjot6). I have only seen two infested specimens lent me by Prof. G. O. Sars, after whom I have named this remarkable form discovered by him. One of the hosts was a female with marsupium, in which were found an adult female, a very small female and the two afore-mentioned lumps of eggs, one of which, probably by chance, was adhering to the abdomen of the female, whereas the other was free and situated more to the front; the female had attached one of its above-mentioned frontal threads to it, which I think shows that the animal must have moved after deposing this lump. The other host was also a female; its marsupial plates were somewhat smaller, though they appeared to be quite developed, being furnished with marginal setz, and in this specimen were found an adult female, a male and an empty skin of a somewhat smaller male. 93 II. Homoeoscelis un. gen. FEMALE. The head small and distinctly defined from the circular, rather depressed trunk. Frontal and marginal borders and sub-median skeleton well developed. Antennule of scarcely medium length, 2-jointed. Antenne wanting. Mouth rather small, the mouth-border being somewhat narrow, with well developed hairs. Maxillule without additional branch. Maxillee about middle-sized, smooth. Mazxillipeds of medium size, second and third joints coalescent. Body entirely naked in the adults, at the utmost a few hairs on the sides behind the head; in the recently hatched and younger specimens the hair- covering resembles that of the male. The trunk-legs are situated on the lateral margins; they are small, each apparently consisting of a rather thick, short basal part which is jointed without articulation to an elongated, very narrow conical, sometimes partly hairy branch, at the base of which the basal part bears either a few set or just a vestige of another branch. Abdomen wanting. Well developed genital area with a transverse chitinous arch, opening towards the front and surrounding the genital apertures, which are situated close together, and the entrances of receptacula seminis. Rather close behind the genital area are the two cau- dal stylets which are a little thicker and longer than the legs. — Spermatophores frequently seen. MALE. Body seen from below elongated ovate, from nearly twice to a little more than twice as long as broad. Length of the head somewhat exceeding a third of the total length. (The antennule 2- or 3-jointed, and very small antennz are perhaps found in one species). In other respects all the other organs of the head, as well as the trunk-legs and caudal stylets, mostly agree with those of the female. The trunk, except the anterior part of the ventral surface, is covered rather closely all over with comparatively short hairs. OVISACS. Are deposed freely, and are of moderate or rather large size, containing few, six to eighteen, eggs which are very or exceedingly large. The number of sacs laid by one female can amount to eight. LARVA. Is known of both species. Antenne much longer than in any other genus, more than twice the length of the antennule, owing particularly to the fact that the penultimate joint is very much elongated; the terminal seta is very long. Maxillule I have been unable to discover. Maxille only 2-jointed, the second and third joints being fused and forming one curved joint, which, moreover, along the larger part of both margins is furnished with extremely fine and short, setiform processes. Maxillipeds far behind the maxillz; all joints smooth; second joint half or scarcely half as long as the third one. Peduncle of the natatory legs rather slender. Third segment of the abdomen together with the small caudal stylets distinctly articulated to the segment, much smaller than the second segment. The longest seta of the caudal stylets sometimes shorter, sometimes considerably longer than half of the cephalothorax. POST-LARVAL DEVELOPMENT. Observed only in one species and described above, p. 5d. 94 HABITAT. The animals live in the branchial cavity of Cumacea, causing a gradual swelling of the carapace above the place which is occupied by the parasite and its ovisacs. A parasite with several ovisacs may be found on immature specimens of both sexes as well as on adult females. Two infested adult males have also been found. he larve infest not only immature specimens, but frequently also females with marsupium. The two species here described come respectively from Denmark and from Messina (and a deep-sea species was found on a Diastylis brought home by the »Ingolf« expedition)*). REMARKS. The genus is distinguished partly by the shape of the trunk-legs, partly by the similarity of both pairs in the male as well as in the two sexes mutually, and by their resemblance to the caudal stylets. In giving the genus its name, I have tried to allude to this conformity in the appendages. The females are very small, more so than in any other genus, which harmonises well with the scanty room left for them in the branchial cavity of their rather small hosts. The males, on the contrary, are uncommonly large in proportion to the females: in the large species about half the length, in the small one even longer than a middle-sized adult female. (In H. mediterranea the antennul of the male are decidedly 2-jointed; what in fig. 1f on pl. XIII appears to be a short basal joint, is an angular excrescence proceeding from the head. On the other hand, the antennule of H. minuta sometimes appear to be 3-jomted, as the two last joints, though coalescent, are separated by a distinct line, which, however, is too strongiy marked in the drawing.) Conspectus of the Species. In the female the basal joint of the maxillipeds is comparatively more slender, a good deal longer than half the breadth of the head at its base. Trunk-legs and caudal stylets of the male are long, longer than half the breadth of the body, and furnished with hairs aboutwalll over sth esc stele] tesa eee ee 1. H. minuta vn. sp. In the female the basal joint of the maxillipeds comparatively stout, scarcely longer than half of the head at its base. Trunk-legs and caudal stylets of the male shorter, not nearly half the breadth of the body, and with very few or no hairs. 2. H. mediterranea n. sp. |. Homoeoscelis minuta n. sp. (PLI, fig. 3a—3b; pl. Il, fig. 1a—11). FEMALE. The largest specimen (fig. 1b), which had not begun laying eggs, is ‘52mm. in length and -49mm. in breadth. A female which has nearly finished laying eggs (fig. 1c) is only 35mm. long and 39mm. broad. The frontal margin has seven small incisions *) J. Bonnier, in his above-mentioned treatise, published probably in Febr. 1897, under the name of Spheronella sedentaria Bonn. described a. species belonging to this genus. He found it in the branchial cavity of Cyclaspis longicaudata G.O.Sars, taken in a depth of 960 metres in “Le Golfe de Gascogne”. 95 (fig. 1h) and very short hairs. Basal joint of the maxillipeds rather slender and a good deal longer than half of the breadth of the head at its base. The genital area (fig. 3a) more than double as broad as long; between, behind and obliquely outside the genital apertures, as well as behind the caudal stylets, are a number of very fine hairs. MALE. A normal specimen (fig. li and fig. 1k) is -20mm. in length and ‘09 mm. in breadth, or about half the length of a middle-sized adult female (comp. fig. 1d with fig. 1b and fig. 1c). The body between scarcely double and a little more than double as long as broad. The frontal margin seems to be like that of the female; the incisions are extremely difficult to see (the hair-covering in fig. 1i is too long). Basal joint of the maxillipeds of a shape similar to that of the female. Trunk-legs and caudal stylets long, longer than half the breadth of the body, and very distinctly furnished with hairs about all over the distal half. OVISACS. Of medium or rather large size, globular or shortly ovate (fig. 1e and fig. 1f). As a rule there are fourteen to eighteen very large eggs in each sac. It is a not common occurrence to find eight ovisacs with one female, and a greater number has never been observed. LARVA. A free specimen (fig. 1g) is 15 mm. in length. Its cephalothorax is nearly double as long as broad. Having only one such specimen in hand, I prepared some larve out of their egg-membranes and examined them more closely; one of them is seen in fig. 1]. We notice that its cephalothorax is still somewhat shorter and broader than that of the freely swimming larva, and the animal is only 14mm. in length. The olfactory seta of the antennulz turned backward reaches beyond the posterior extremity of the cephalothorax. Distance between the maxillz and the maxillipeds about as long as the basal joint of the latter. Second abdominal segment as long as the first. The longest seta of the caudal stylets considerably longer than half the length of the cephalothorax, about half the length of the body in the free specimen. POST-LARVAL DEVELOPMENT. Described in detail above, on p. 5d. HABITAT. In the branchial cavity of Diastylis lucifera (Kyr.) from Denmark. The parasite I have found in sixty-six specimens, the special locality of fifty-seven of these are unknown to me; five specimens were taken at Hellebek (four by Dr. Joh. Petersen, one by the author), and four in the following four stations of the expeditions of »Hauch«: Stat. 25 (110 fathoms), Stat. 368 (13 fath.), Stat. 370 (15 fath.) and Stat. 383 (14 fath.)'). Hither the right or the left side is infested; in no specimen have [ found both sides infested. Only one female and generally also one male, rather seldom two males, are found in the same branchial cavity. Where the parasite has laid several ovisacs, the carapace of the host is very considerably swollen, and frequently this swelling rises somewhat above the median dorsal line. 1) Details about the exact localities of these stations, the description of the bottom ete. is found in: »U. G. Joh, Petersen: Det videnskabelige Udbytte af Kanonbaaden ,Hauch*s Togter i de danske Have indenfor Skagen i Aarene 1883—86*, p. 1—33, 1893. In my descriptions of several of the following species, other stations from these cruises will be quoted and may be looked for in the afore-mentioned work. 96 An adult parasite, especially when it has laid several ovisacs, is easily seen through the carapace, but if we want to find out if a specimen without swelling is infested with larve, pup or recently hatched specimens, we must examine it carefully under a good dissecting microscope. I will try to give special statistics of my material, but unfortunately, at the beginning of my investigation some years ago, I omitted to put down sufficient notes about a few of my specimens, so I cannot give as perfect statements as I should like. Of 433 specimens of Diast. lucifera, 66 were infested, 367 were not. Of the latter, 182 were females with eges or young ones in the marsupium, 74 females with empty marsupium, 28 adult males, 7 young males (before the last moulting) and 76 young females (some of them may have been males without rudiments of abdominal appendages). Of the 66 infested specimens 13 were females with young ones in the marsupium, (there was not one with eggs), 31 females with well-developed though empty marsupium, no adult males, 3 young males and 15 young females; concerning 4 specimens sufficient notes are wanting, at all events none of them was an adult male. Of the 62 specimens, 33 contained an adult female with one or more ovisacs, and also, of course, a male, and of these 33 hosts, 4 were females with young ones in the marsupium, 13 were females with empty marsupium, 13 young females and 3 young males. In 29 of the 62 specimens was found either a moderately large female without eggs, or one (or two) males, or in many cases recently hatched specimens or pup, and in at least one case, only one recently entered larva. Of these 29 hosts, 9 were females with young ones in the marsupium, 18 were females with well-developed, empty marsupium, and 2 young females (without marsupium). No help is needed to draw various conclusions from these figures; I will only observe that the number of males collected is too small to allow us to conclude that they are never infested (s. the following species). REMARKS. ‘The three infested specimens first observed were discovered by the Inspector, Dr. F. Meinert, whom I had asked to look out for eventual parasites in determining the Cumacea from the cruises of the »Hauch«. With respect to the figures it may be observed that in fig. 1a I haven given a drawing (in the same enlargement as fig. 1b and 1¢ etc.) of a rather young female, whose limbs are remarkably long, as in the male (fig. 14d), and which has already four spermatophores attached to its genital area, though it is far from being old enough to begin laying eggs. Fig. 3a (pl. I) exhibits two spermatophores (s) fixed at the entrances to the receptacula seminis (r), which are seen through the skin and are indicated by dotted lines. 2. Homoeoscelis mediterranea n. sp. (Pl. XIII, fig. 1a—1h). FEMALE. ‘The specimen represented (fig. 1a) is 28 mm. in length, -23 mm. in breadth; the largest specimen taken out measures -31 mm. in length, ‘32mm. in breadth. 97 The frontal margin uninterrupted, with scarcely any hairs. The basal joint of the maxillipeds moderately stout (fig. 1d), scarcely longer than half the breadth of the head. The genital area (fig. le) not nearly twice as long as broad; the whole area between the genital apertures and the caudal stylets, a narrow part behind the latter, and an area outside the soft mem- brane which borders the genital apertures, are covered with rather short hairs. The remainder as in the preceding species. MALE. The well-developed specimen illustrated (fig. 1f and fig. 1g) is 17mm. long and ‘09mm. broad, thus a little more than half the length of the largest female. The body more clumsy than in the preceding species, not twice as long as broad. Frontal margin and maxillipeds as in the female. Trunk-legs and caudal stylets shorter, not nearly half as long as the breadth of the body, and with very few or no hairs. OVISACS. Rather large (fig. 1c), shortly ovate or sub-globular. There may be five to twelve, but generally we find six to ten comparatively extremely large eges in each ovisac, and as many as eight ovisacs have been found with one female. LARVA. In one ovisac were found larve nearly on the point of swimming out; one of these is figured (fig. 1h); its body is ‘15mm. in length, and the cephalothorax is very elongated. The olfactory seta of the antennule reaches the posterior extremity of the cephalothorax. Distance between the maxille and the maxillipeds considerably shorter than the basal joint of the latter. Second abdominal segment scarcely the length of the first. The long setze of the caudal stylets considerably shorter than in the preceding species, not half the length of the cephalothorax and frequently much shorter. POST-LARVAL DEVELOPMENT. Unknown. HABITAT. In the branchial cavity of Iphinoé trispinosa (Goods.), at Messina. In May and in the beginning of June 1893 I caught fifty-eight specimens in all ages of this species in the harbour of Messina, in a depth of ten to twenty fathoms, and seven out of these were infested with the parasite. It was only found in specimens which were either much more than half-grown or full-grown, so that no parasite appeared on a single one of the numerous specimens which were only half-grown or still younger. A female with the marsupium containing Spheronella marginata (s. later on), had on its right hand side a considerable swelling, in which were two adult females, two males and six ovisacs. In a female with less than halfdeveloped marsupium, the left branchial cavity contained a male; the right hand side of the carapace, especially its posterior part, bulged very much, and under the hindmost part of it were found an adult female, in front of it a male and eight ovisacs, the foremost of which was evidently newly laid, whereas another, which was lying close up to the female, contained full-grown young ones; this arrangement showed clearly that the female had changed place. An adult male contained in its left branchial cavity one half-grown female. Another adult male contained, also in its left hand side, one adult female, one young female, one male and two ovisacs. The three remaining hosts, viz. a male before its last moult and two not quite adult females, have not been dissected, but it 13 98 could be observed through the carapace that in one of the specimens (a female), the parasitic (no doubt adult) female had not begun laying eggs, whereas the two other specimens lodged not a few ovisacs beneath their carapace. REMARKS. In this small species the female is smaller and the eggs comparatively larger than in any other form of this family hitherto found. It is closely allied to Homoe- oscelis minuta, though the male in particular is easily distinguished from this species by its shorter legs. It may be observed that an examination of numerous specimens of [phinoé trispinosa from Denmark gave a negative result. III. Spheronella satensky (1868). FEMALE. Head small, generally, though not always, defined from the trunk, which is ovate or globular, sometimes even a little broader than long. Mazxillipeds large or rather large and at least always longer than the maxille. Trunk-legs sometimes wanting; if found, they always consist of one minute cylindrical joint with a couple of terminal setz, or they are reduced to small eminences. Genital area is found and always well developed, so that the genital apertures — which are rather or very close together — and the entrances to the two receptacula seminis, which are situated close in front of them, are surrounded or at least bordered posteriorly and at the sides by common rather solid chitine. Caudal stylets are sometimes wanting; if found, they are shaped somewhat like the trunk-legs and terminate in one, two or three sete. — Spermatophores frequently observed. MALE. This sex is known in a little more than two thirds (twenty-five) of the species. The body, seen from below, is ovate or sub-globular (seen laterally, the back is much vaulted, the ventral side rather flat, sometimes even concave in the middle). The head always furnished with frontal and lateral borders. The trunk is covered with hairs on the larger part of the ventral surface, as a rule, also on its sides and on the whole or part of the back. Trunk-legs and caudal stylets sometimes wanting, but generally found, and in this case differing much from each other. The trunk-legs always deviating very much from those of the females. OVISACS. Always deposed freely. LARVA. Observed in scarcely half of the species. Antenne at most a little longer than the antennule, sometimes very short. Maxillipeds situated close behind the maxilla. Peduncle of the natatory legs of medium breadth or narrow. Posterior abdominal segment together with the caudal stylets nearly always smaller than the penultimate segment. Long seta of the caudal stylets longer than half the length of the cephalothorax. POST-LARVAL DEVELOPMENT. Known or partly known in several species (s. above). 99 HABITAT. The marsupium of Amphipoda, Cumacea and Isopoda. Younger females and animals in various stages of development of some species are also found on the ventral surface of the thorax in Amphipoda. Found in Denmark, Norway, Greenland, in the Kara Sea, the Mediterranean, off Cuba, and finally on the East-coast of Asia between about lat. 22° and 51°N. REMARKS. ‘This genus is very large. Subsequently thirty-four species will be described, besides S. Leuckartii Sal., which I have not seen (and S. diadema G. and B. which has not been described). The above given diagnosis of the genus is rather meagre, and it is easy to see that some of the characteristics are qualified by an »either .. . ore. The obvious reason is that many of the species in several respects vary considerably among themselves. Above, on p. 82—83 I have already given a general view of the most important of these differences, stating my reasons for not feeling justified — in spite of these differences — to divide the genus into several genera. It is impossible, for two reasons, to give complete analytical keys of the two sexes in all species, firstly, because the male is unknown in nine species, secondly, because the differences between the females and between the males in one small division are too vague to be represented with sufficient preciseness in such a conspectus. However, in order to procure a kind of general view, I will divide all species into three groups according to the orders of their hosts, subsequently giving as good a conspectus as possible of each of the two sexes in the species of each group. In this place I will only give the analytical keys of the first division; the other keys will be found immediately preceding the divisions to which they belong. a. Parasites on Amphipoda. It may be observed that the males of this group always have well developed trunk- legs and caudal stylets. In the larve the front is never furnished with processes or lists in the sub-median part (there are one or two rather small lists near the basis of each antennula), the basal joint of the maxilla is smooth, without combs, and the terminal joint not serrated. 1. Conspectus of the Females. The figures preceding the names of the species indicate their number in the subsequent representation. In this conspectus all species are included except S. abyssi n. sp., of which my knowledge is too fragmentary (s. the description below). 1. Head with distinct frontal border and distinct prominent lateral borders. Genital area road ersthanilOne a. pln. don See ea tes ee eae ae PCR eek eee On aes) ease a RO es 2 on 100 Head without distinct frontal border; lateral borders wanting. Genital area longer hae TOA Ss sec fxs ek ymca aes edet se o> AULA C RE aes 26. S. microcephala G. and B. Antenne quite rudimentary. A tuft of hair near the base of each maxillula. Group: S. Leuckartii Sal. including the following species: . . . 1. S. elegantulan. sp., 2. S. Atylin. sp., 3. S. danica n. sp., 4 S. vestita n.sp., 5. S. Leptocheiri n. sp., 6. S. messinensis n. sp., 7. S. chinensis n. sp., 8. S. antillensis n. sp. . Antenne pretty well developed. No hair-tufts near the base of the maxillule ... 3 Krontal margin! without expansion yin the middle 22.9.) snen ne iene eee 4 Frontal margin with a flatly cup-shaped expansion in the middle . . 25. S. frontalis n. sp. Trunk covered with short hairs, 2- or 3-branched at their basis............ 5 Trunkenaked om with rather tewa Siniple) hairs msm ogra een men mate ane 6 Caudal stylets situated between the genital apertures. The trunk thickly covered swt, aIPSs 3.1.4 Se x, gh Rene pres cleo ees ches de pode Yea tee Peco Sate bed 9. S. Calliopii n. sp. Caudal stylets situated behind the genital apertures. Hair-covering less thick. 10. S. irregularis n. sp. Trunk fastened to the host by a very short ventral thread. . . 17. S. paradoxa n. sp. Troink-never attached. to, the: lost: 4:...<