THE EMBRYONIC DEVELOPMENT OF THE SALPIDAE. 445 



cavity appears within each of the developing muscle-hoops and these 

 cavities have been compared by SALENSKY with those in the primitive 

 muscle-plates of the Vertebrates (cavities of the primitive segments). 



The rhtrnhlnxt (Fig. 224, eh) seems to be derived irom cells of the 

 mesenehyme. 



The development of the other species of Salpa that are provided 

 with the enveloping fold (S. a/ricana, S.fusiformis and S. punctata) 

 seems to follow essentially the same course as that of S. pinnata. 

 Judging from the very fragmentary statements as to the ontogeny 

 of these forms (BAKROis, No. 87, SALENSKY, No. 104), however, there 

 appears to be considerable variation in points of detail [see HEIDEB 

 (X. XIII.) and KOROTNEFF (Nos. XX., XXa., and XXIa.)]. 



General Considerations on the Embryonic Development of 

 the Salpidae. 



[The embryonic development of the Salpidae is still, in spite of the recent 

 investigations of BROOKS, HEIDBR and KOROTNEFF, anything but satisfactorily 

 understood. The confusion arising from the immigration of the kalymmo- 

 cytes and the difficulty of discriminating between these cells and the smaller 

 formative blastomeres of HEIDER and KOROTNEFF, still requires to be swept 

 away before we can make any satisfactory comparisons of the development of 

 Sdljxi with that of other Tunicates. When, further, we compare the mono- 

 graphs of the two last authors with that of BROOKS, we must f 3el that additional 

 confirmation of one or other of the views put forward by these writers is- 

 necessary before one of them is finally accepted. It is obvious, however, that 

 we are dealing with a highly specialised form of development which has- 

 possibly arisen in connection with the viviparous habit of Salpa and has been 

 further complicated in connection with the peculiar life-history of this form.], 



The embryo undergoes direct development in accordance with its 

 retention during the whole embryonic and larval periods within the 

 body of the parent. We have already seen a similar omission of 

 metamorphosis among the Ascidiacea in the Molgvlidae. The Salpidae 

 in this respect show a more specialised condition in so far as the- 

 embryo becomes closely connected with the tissues of the parent, 

 and a placenta develops for its nourishment. Since we may, with 

 some probability, derive the Salpidae from the attached Ascidia-like 

 forms, we might be tempted to trace back the fixation of the embryo' 

 on the wall of the respiratory cavity of the parent to this original* 

 attached manner of life. 



[Considerable stress was formerly laid upon the ontogeiietic differences- 

 described by SALENSKY as occurring in the various species of Salpa, especially 

 on the presence or absence of covering folds and on the structure of the- 

 placenta. According to SALENSKY, the covering folds were completely wanting. 



