38 W. B. SCOTT AND HENRY F. OSBORN. 



stated, it is much nearer to the upper pole than to the lower 

 and this gives at once the distinction of larger and smaller 

 blastomeres. The smaller blastomeres grow round the ovum 

 over the larger, and bear the same relation to them as they 

 do in the Frog. The segmentation cavity appears early, and 

 from the very first its roof is only one cell thick, just as in 

 the case of the Lamprey. As we shall see later the epiblast 

 is at first composed of one layer, and hence the roof of the 

 cavity is covered by epiblast only. In the Elasmobranch 

 Fishes the roof of the cavity is formed by lower layer cells 

 also, and this Mr. Balfour explains by the increase in the 

 quantity of food-yolk in the cells, compelling them to 

 creep up the sides of the cavity. Although there is propor- 

 tionately more food material in the Newt^s egg than in that 

 of the Frog the increase is relatively -small and does not 

 affect the position of the cells. The only difference between 

 the two at this stage consists in the fact that the roof of the 

 cavity in the Frog is two or more cells thick, and in the 

 Newt only one. In short, the ovum of the latter resembles the 

 morula of Amphioxus with a large amount of food material 

 stored away in its lower part. Judging from the descrip- 

 tions of Calberla, it is in no way different from the ovum of 

 Petromyzon of corresponding age. The floor of the segmen- 

 tation cavity, as in all ova which contain food-yolk, is formed 

 by the upper layer of yolk-cells from which, eventually, 

 the ventral epithelium of the alimentary canal is in part 

 derived. 



The next step in development is, as in the Batrachians, a 

 process of invagination, and, as in them, it is an unsym- 

 metrical invagination. The disturbing cause is in both cases 

 the presence of the food-yolk below. Owing to the fact that 

 the food to be made available must be placed upon the ventral 

 side of the body, the invagination must in this region take 

 place very slowly or not at all. By this simple considera- 

 tion Mr, Balfour explains the unsymmetrical gastrula of the 

 higher Vertebrates. 



At the period when our study of the two lower layers 

 proper begins, segmentation is complete ; the lips of the 

 blastopore are rapidly nearing each other ; the epiblast con- 

 sists of a single layer of partly columnar, partly wedge- 

 shaped, cells, and has already in great measure attained those 

 characters which persist throughout several of the following 

 stages. 



At the lip of the invagination (see Plate IV, fig. 2) there 

 is a decided swelling produced, in part by a lengthening, iu 

 part by a reduplication of the cells, a histological change 



