90 BULLETIN OF THE 



in addition its cells are not in form greatly unlike those of the interme- 

 diate cell-mass, has been looked on by many as the source of the cells in 

 question. No one, it is true, has seen cells in such a position as to war- 

 rant the conclusion that they were actually detaching themselves from 

 the lower wall of the gut and entering into the intermediate mass, but 

 the gut after flattening is not sharply enough marked off" to preclude such 

 a possibility. 



Such are the conclusions one may reach in the study of the advanced 

 stages ; but let us again examine the younger stages, now with regard 

 to the endoderm. 



The single row of endoderm cells that is present in the stages shown 

 in Figures 13 to 15 is clearly distinct from the intermediate cell-mass. 

 In fact these endoderm colls are, from the earliest stages until the gut is 

 completely closed, so sharply defined that it is difficult to conceive how 

 the intermediate cell-mass could arise from that source. The later stages, 

 as I have said, present a condition which would make such an origin 

 conceivable, but the proof of the passage of individual cells from the gut 

 to the intermediate cell-niass even at this stage is wanting, while from 

 the younger stages it is evident that the cells do not as a mass arise 

 from the endoderm. 



3. Lateral Plates. — As regards the possibility of an origin from the 

 lateral plates of the mesoderm, there is not much to be said, since the 

 evidence in favor of this theory is even less satisfactory than that for 

 the pi-eceding views. Balfour ('73) and Rlickert ('88) have maintained 

 that the lateral plates furnish some of these cells, but I can find no one 

 who supports the theory of their exclusive origin from this source. 



It is true that, when in the cod the approaching ends of the lateral 

 plates are thickening and the cells of the intermediate mass are arranged 

 in rows parallel to the ends of the plates, the relation is such as to render 

 possible a connection between the plates and the intermediate cell-mass. 

 Such conditions may be seen in Figures 6 and 7. But these are stages 

 long after the formation of a large part at least of the intermediate cell- 

 mass, and since a careful examination of the lateral plates, from the ear- 

 liest stages through the entire development, shows them in this region 

 to be clearly defined, the argument drawn from this condition would be 

 without value. It is true, that in a posterior region the lateral plates 

 are not so clearly defined in any of the stages. 



In fact, Boyer ('92, Plate IV. Figs. 25-29) supports, for fundulus, 

 the theory that, in parts of the embryo posterior to the fourth proto- 

 vertebra, some of the cells of the intermediate cell-mass come from the 



