466 



The complete memoir will be illustrated by many exact figures, 

 and in it the literature will receive fuller mention than in the following 

 abstract. In it, also, the finds made in the various embryos will be 

 described in detail, and each of the important questions, which have 

 arisen out of the work, will receive separate consideration. 



A brief narrative of the results in connected form, beginning at 

 the earliest period to which the germ-cells of Raja have yet been 

 traced, will be given here. 



General Survey of the History of the Germ-Cells. 



Eigenmann ('92) has concluded from the sizes of the germ-cells 

 of Cymatogaster = Micrometrus aggregatus, that they probably re- 

 present cells of the fifth division of the egg. At the moment, although 

 the germ-cells of Raja can be definitely traced in preparations to the 

 close of the segmentation, or to the period immediately preceding the 

 first formation of the embryo, their actual presence in earlier phases 

 has not yet been established by observation. He found them in an 

 embryonic foundation, in which no mesoblastic somites were yet pre- 

 sent; my observations have revealed them still earlier, before there 

 is any real trace of the future embryo. In this respect, 

 apart from others, the present work may claim to be an advance. 



Whenever the segmentation of the egg of Raja shall have been 

 described and figured after the manner of Rückert's work ('99) on 

 Pristiurus and Torpedo, it will be a matter of comparative ease to 

 determine the particular period of the egg-cleavage, at which they 

 are separated off. But it may be surmised, that in Raja they go 

 back, as in Macrometrus and Cyclops, to a very early period. From 

 their average size no certain conclusion can be drawn in the absence 

 of information as to these early cleavage phases. 



But in embryo no. 454 theii^ number was found to be 512, in 

 embryo no. 448 there were 490 germ-cells without counting a con- 

 tingent of about two dozen others in an impossible situation, and 

 none of the observations in earlier embryos of less than 32 mm 

 (no. 454 measured 32 mm and no. 448 26 mm) are at variance with 

 this. Obviously 512 represents either 8 or 9 divisions. 



RtJCKERT ('99) has shown, that with the 10th synchronous division 

 the segmentation-cavity is fonned. Beyond this point our knowledge 

 is indefinite, but, judging by the numbers and sizes of the cells at 

 the close of the cleavage in the skate, it appears to me very unlikely, 

 indeed, that the majority of the cells divide more than twice or thrice 

 after the formation of the segmentation-cavity, and before the period 

 represented by embryo no. 434, where the formation of the embyro 



