l'.W7.] NATURAL SCIENCES OF PHILADKLPHIA. 339 



plcx of the posterior lip, and tlnis represents the final position of the 

 l)lastt)pore too near the posterior pole, but it represents fairly well the 

 relative movements of the different organ bases during the first four 

 stages sliown (1-4). Reference to the first and second cleavage furrows 

 l)etween the macromeres, shown in figs. 25-2.S, will show that the 

 posterior margin of the blastoderm remains relatively fixed during this 

 period, while at the same time the blastoderm extends forward and 

 laterally ; still later the blastoderm extends posteriorly also, and finally 

 the blastopore closes at the vegetal pole (fig. 30). 



2. APICAL INVAGINATION. 



The first structure which appears in the course of development is an 

 invagination of ectodermal cells just anterior to the animal pole and in 

 the region of the anterior arm of the cross. This invagination is pre- 

 ceded by a broad depression of the blastoderm (fig. 18), and then the 

 center of this depression becomes deeply cup-shaped (fig. 19), and finally 

 tubular in form (fig. 20). The axis of this tubular invagination at first 

 runs forward under the ectoderm; then it becomes perpendicular to the 

 surface of the blastoderm ; and subseciuently it runs backward under the 

 ectoderm, its opening being at its anterior end (fig. 20). The inner, 

 blind end of this invagination lies near the anterior end of the polar 

 furrow, and in some instances, though not in all, a few cells are separ- 

 ated from the invagination at this point. The anlagen of the cerebral 

 ganglia are formed on the right and left of this invagination, but they 

 do not appear to be derived from it. Similarly a few mesoderm cells, 

 which probaljly correspond to the "larval mesoderm" of other mol- 

 lusks, lie on each side of the invagination, though they are not derived 

 from it (see the black nuclei in figs. 19 et seq.). 



Subsequently this conspicuous and definite structure completely 

 flattens out and disappears (figs. 21 et seq.), leaving not a trace be- 

 hind. This fact seems so remarkable and the significance of the 

 invagination is so problematical that I have devoted considerable time 

 and effort to the study of it. 



It is evident that this invagination is not merely the result of the 

 sinking in of the blastoderm over a cavity in the yolk. There is 

 frequently a cavity between the macromeres, but this invagination 

 invariably lies anterior to this cavity, and in many cases it actually 

 forces its way into the substance of the yolk. In some cases also a 

 prominent evagination is formed instead of an invagination, showing 

 that there is very active growth in this region of the blastoderm. 



It seems to me probable that this invagination is a mechanical 



