79 



b(xl\-, partly, also on the ilorsal sides PI. I, lis:,'. S, et„ PI. \'I1I, Ug. 2. 3, ;. These I have designated, sub-ectodermal testes; 

 they have no comiminiration with the true testes. The si>ermatox()a, develojjed in the.se testes, must pa.ss through the ecto- 

 derm to escape from the bod\-, \ide PI. VUl, tig. 3 a, h. c, d, f and fig. 2 h and 0. These sub-ectodermal testes have a 

 somewhat peculiar structure. I have not observed distinct cell-membranes, but a great many nuclei are often situated close 

 together, and thev h.ive, tiuis, freiiuentl)-, the appearance of a large cell with many nuclei, vide fig. 2 a. These sub-ecto- 

 dermal testes must be regarded, I think, as, originally, portions of the chief testes which h;i\e. probabi)-, in the earliest stages 

 of development. ,1 very extensive distribution through the whole body. When, however, the ovaries are de\eloped, they be- 

 come restricted to a more confined distribution, and during this process some portions arc isolated, and become, in the adult 

 specimens, sub-ectodermal testes, the spermatozo.a of which are obliged to force their own passage from the bod\-. Among 

 the sparmatozoic globules, jilent)- of undeveloped sperm.itocytes occur, in the vas.i dift-rentia, as well as in the vcsiculæ 

 seminales and ductus ejacuiatoni. The development of the penes v.aries vai\- much; in some species (e. g. M. giganteum 

 and M. graffi) the}' are \-er}- prominent, whilst in other species (M. cirriterum) the\- are scarcely \-isibk-. the ductus ejacul;iforii 

 opening into no prominence. 



Complemental males (Dwarf-males). 



I h;i\e found complemental males in M. giganteum, M. gigas, and M. carpenteri. The- structure in these males is 

 quite similar to that of the hermaphrodites, except that on the dorsal side, where the ovaries are situated, we have here, in 

 these, found testes whose cells are, however, little ileveloped, and the)- have, frequentl)-. thus, a partial resemblance to young 

 ovaries. Upon closer examination, however, it can be seen that they communicate with, and are parts of the more develo- 

 ped portion situated more to the centre of the body. The testes fill the entire body-cavit\-, so that luit little of it is \-isible. 

 A feature of great interest in these males is the fact, that the three oviducts of the hermaphrodites are present, in them. 

 The dorsal oviduct (uterus) is but little developed, but the lateral oviducts are well developed, and quite resemble those of the 

 hermaphrodite, except that the\- contain no ova. — In branches of the uterus are situated some organs, similar to those in the 

 hermaphrodites (PI. II, fig. 15, ovr.) and which are. probablj-, traces of ovaries. The presence of these oviducts can. in m>- opinion, 

 be explained in onl\- one of two ways viz. These males are, either sprung from hermaphrodites and these oviducts are. then, 

 traces of their origin ; or these males are only young hermaphrodites, and ovaries communicating with these oviducts are 

 subsequentlx- developed, whilst the testes are restricted in their distribution. I have not. as >-et, had a sufficient supph' of 

 material to enable me to determine these questions. It is af eature of interest, that in the single male specimen of M. car- 

 penteri found I have observed cells, quite similar to }'oung ova, situated on the dorsal side and, so far as I could discover, 

 situated between the testes and the connectue-tissue of the body-wall. These cells must be, either, remnants of hermaphro- 

 dites, or }-oung o\-aries which are subsequent!}- developed. 



The hermaphroditism of the Myzostomida. 



The hermaphrodites of the M}-zostomida are regarded b}- Beard as ha\ing a sccondar}- origin, and he regards the 

 diecious forms to lie primitive ones, or the starting point . I can not gi\'e ni}- assent to that view, tor the tollowing reasons: 

 i) The diecious species are the most parasitic forms, and it Heard's view, that the Myzostomidae have become hermaphrodites 

 because there is a tendenc}- in parasitic life to produce hermaphroditism, is correct, it ma}-, also, with some reason be con- 

 cluded, that the Myzostomida C}-sticola or most parasitic species should, especiall}-, be androg}-nous, but that is not the 

 case, for most of these species are diecious, whilst the most migrator}- ones, and, consequentl}-. least parasitic tbniis are, 

 especially, hermaphrodites; antl the little migr.-itor}- species .ire, on the other hand, provided with dwarf-males which accor- 

 ding to Beard's view are more primitive than the hermaphrodites, and are remnants of a diecious state. The most parasitic forms, 

 and the little migrator}- species w-ith dwarf males, cannot be assumed to be the starting point ol the iifigrator}- free-living 

 species. 2) The rudiments of testes visible in M. c}-sticolum are, to ni}- mimJ. more probahl}'. remnants of an androg\-nous 

 state') than a budding development of male organs. If such rudiments, without male generative apertures, occur, uniform!}-, in 

 all females, I cannot conceive how they can be a budding development, as the}- can have no opportumt}- of sexual function. 

 Their regular occurrence cannot be accounted for, solelv, by assuming a tendenc}- in the ovaries to develope spermatozoa. 

 1,) The males of the diecious species are smaller in relation to their females, th.iii the dwarf-males are, in ri'l;ition to their her- 



'; Under such circumstauces the male generative ai>ertures, usually, ilisajipear, as for instance in C'ymolhoiila. 



