QsLStnila-phase of Development in Mollusca. 459 



egg-cell leads to the formation of a mulberry mass {Morula 

 of Hackel) ; at one point the cells forming this mass become 

 invaginated ; the cavity of invagination is the primitive ali- 

 mentary canal, the invaginated cells constitute its walls. 

 The orifice of invagination closes up, and the pedicle formed 

 by that portion of the primitive alimentary cavity which is 

 continuous with the cells of the outer or body- wall, the pedicle 

 of invagination as I term it, becomes the rectum. ' In the 

 cases above cited, with the exception of Paludina, the rectum 

 is thus for a long time blind. In Paludina, however, the 

 orifice of invagination does not close up until a very late 

 stage, if at all, and is ciliated. The mouth and cesophagus 

 eat their way into the primitive alimentary sac subsequently 

 in all these cases. In a recent paper on Lymnceus I have 

 shown that the cells which are invaginated to form the primi- 

 tive gastric sac of the Planula or Gastrula undergo very 

 remarkable modifications before the ultimate form of the ali- 

 mentary canal is developed (see 'Quarterly Journal of Micro- 

 scopical Science for October 1874). 



I do not hold that the formation of a double-walled sac by 

 invagination of a primitively single-walled sac is the essential 

 feature which constitutes a Gastrula or Planula. The endo- 

 derm of this developmental stage may take its origin by 

 delamination — that is, by direct cell-division from the primi- 

 tive single series of cells constituting the wall of a hollow 

 Morula. We have accordingly to distinguish these two very 

 different modes of origin of the Diploblastic Planula or Gas- 

 trula. Facts must be accumulated to enable us to decide 

 which is the original mode of formation of this developmental 

 form, and to understand the steps by which the one process 

 was substituted for the other. The origin of structures by 

 invagination, when looked at broadly in a large series of 

 animal forms and in the case of many organs, points to the 

 conclusion that invagination is an economy of material — a 

 mode of rapidly filling in the outline of an organ in the em- 

 bryo, whilst leaving the organ in a hollow condition for sub- 

 sequent completion. This is seen in the contrasted modes of 

 development (by delamination on the one hand and by in- 

 vagination on the other) of the nerve-chord in annelids, in 

 osseous fishes, and in higher Vertebrata, also in the cases of 

 the otocysts of Gasteropods and of Cephalopods. At the same 

 time I do not know that at present we have any strong reason 

 for supposing that the delaminate mode of origin of the Gas- 

 2rw/a-endoderm preceded the invaginate. That the difference 

 between these two modes of origin is not a fundamental one 

 appears from the fact that in closely allied genera we find 

 either the one or the other occurring indifferently. As I have 



