BULLETIN OF THE UNITED STATES FISH COMMISSION. 295 



intervene between the liypoblastic and ei)il»lastic layers whieli cover 

 the yelk, sections throngh whole ova in varions stages of development 

 have thus far fiiiled to show its existence, except in the salmon, in which 

 it is quite evident in sections of advanced embryos. Even in the latter 

 I am not sure that it extends entirelj' over the yelk, ffillacher has ap- 

 parently understood the relations of the segmentation cavity much in the 

 same Avay as the writer, except as to the heart. 



COMPARISON OF THE TELEOSTEAN OVUM WITH THAT OF OTHER VER- 

 TEBRATES. 



A comparison of the different types of vertebrate ova will be useful 

 as leading to a clearer compreliension of the true nature of the yelk in 

 the teleostean egg. The eggs of the common frog {Eana) and Bomhi- 

 nator undergo total segmentation in the process of development. There 

 is no distinct vitellus or yelk, and the yelk of the fish egg is apparently 

 not homologous with any part of the amphibian ovum. There is, how- 

 ever, an almost complete homology between the germinal disk and blas- 

 toderm of the fish and the whole of the amphibian egg. The complete- 

 ness of the homology is impaired only by the peculiar way in which the 

 neurula or brain and spinal cord and the intestine are developed in the 

 fish. The fish Qgg may be regarded as the frog's ovum plus a large 

 store of food, which may be either homogeneous or heterogeneous, and 

 which at first takes absolutely no share in the process of segmentation. 

 If it were possible to place a frog's egg on a sphere of protoplasm sev- 

 eral times its own size and cause it to spread out and gradually grow 

 over the latter so as to comi)letely inclose it and yet develop perfectly, 

 the condition which obtains in the fish ovum would be very nearly at- 

 tained for the amphibian. The segmentation cavity, which appears in 

 the germinal disk at an early stage of develox)ment of the fish, is per- 

 fectly homologous with a similar cavity in the egg of the amphibian, 

 except that in the fish, instead of remaining a simple cavity it has been 

 so greatly modified by the peculiar way in which the disk of cells in 

 which it is contained is obliged to spread and grow over and around the 

 yelk that it is at first not easy to see a likeness between the two types. 

 The development of the tailed Batrachians and of the Lampreys is very 

 similar to that of the frog, and their ova undergo total segmentation. 

 The development of the ova of the genus Lepidosteus is probably not 

 essentially dilierent from those of the typical teleostean. The process 

 of spreading and inclosure of the yelk hj the blastoderm has not been 

 observed in the bony gar, but, as far as I am able to judge from the 

 account given by Balfour {Comp. Emhryol. ii, 91-98), a segmentation 

 cavity is probably formed, and the diflerentiation of the embryonic lay- 

 ers is apparently not essentially different from the same processes as 

 observed in teleosts by A^arious persons besides the writer. Fig. 58 in 

 the work just referred to, and relied upon by Balfour to shosv the seg- 

 mentation of the egg of Lepidosteus^ apiDears to me to be taken from 



