198 KIXGSLEY AND CONN 



al ones. At a later stage this lower layer of protoplasm becomes segmented. Aside 

 from this difference, which is to be explained by the presence of quantities of yolk gran- 

 ules, the external features pf segmentation present no important differences from our 

 results. 



The eggs of Merlucius although containing large numbers of deutoplastic globules in 

 the germinal portion undergo a complete segmentation of the protoplasm as in that of 

 the Cunner. There was to be noticed, however a very marked irregularity in the pro- 

 cess of segmentation even from the very first. The blastomeres varied widely in size and 

 the segmentation furrows progressed at varying rates and times in different portions of 

 the germinal area. 



Though we found it impossible to obtain any satisfactory sections of our eggs, their 

 perfect transparency enabled us to see, clearly and plainly, that even the first segmenta- 

 tion furrow extended clown to the vitelline globe and that the first two cells as well as 

 the subsequent ones were entirely separated from each other. At no time was there an 

 unsegmented basal portion of protoplasm except that presented by the intermediary layer 

 and peripheral cushion of Van Bambeke. 



Kupffer mentions some marked irregularities in the eggs which he studied : in some 

 the second segmentation furrow sometimes was eccentric or occasionally was even parallel 

 to the first. We have seen nothing of this sort in the eggs which we studied. 



The theoretical segmentation of an egg is first two meridional furrows and then an 

 equatorial one, but frequently this regularity is interrupted, in fishes noticeably so. If we 

 interpret Haeckel aright, the equatorial furrow is the fourth to appear in the eggs of 

 ?Motella as he figures (fig. 58) a section of a blastoderm of sixteen cells and in it 

 lower layer cells are seen. These figures however show in every line that they are 

 wholly diagrammatic and could not have been drawn from either actual or optical sec- 

 tions. In the eggs of the fresh water fishes it is at a somewhat later stage that the equa- 

 torial furrow is formed and the lower layer cells produced, but even in eggs of the same 

 species there does not appear to be much regularity. In our eggs lower layer cells did not 

 appear until the blastoderm was composed of about a hundred blastomeres and even then 

 they did not appear simultaneously in all parts. 



III. Formation of the Germikal Layers. 



In this section we have to consider the extension of the blastoderm over the yolk from 

 the time of the appearance of the lower layer cells until the formation of the notochord 

 and the neural canal. In it also will be discussed the differentiation of epiblast, mesoblast 

 and hypoblast, and also the phenomena of invagination. Certain of the features here to 

 be described belong in part to the next section, but from the fact that they are first 

 noticed before the formation of the notochord they are best treated here. 



Until after the blastoderm has acquired a stage with about a hundred cells it consists, as 

 before mentioned, of but a single layer, thus offering a marked contrast from the «ggs of 

 most fishes on which observations have been published. This simple condition of the 

 blastoderm at this time was conclusively shown by optical sections in which the outlines of 

 each cell could be readily traced with a power of a hundred and fifty diameters. 



