312 PROCEEDINGS OP THE AMERICAN ACADEMY. 



whole process no stage occurs where there is any trace of an over- 

 growth, such as I have found in the pig, or of an external opening into 

 the cavity which lies between the two layers in question. Moreover, 

 this cavity in the mole is largely filled with a loosely arranged mass of 

 amoeboid cells, except in one case, where, as Heape ('83, Plate XXIX. 

 Fig. 25) says, their absence is due to mechanical injury in the process 

 of preparing the sections. Furthermore, it may be noted that this 

 bridge-like structure appears at a much earlier ontogenetic stage in 

 the mole than in the pig. In the former the entoderm covers the area 

 of the germinal disk only ; in the latter I find a complete didermic 

 vesicle when the first trace of the bridge cells appears. It is true, we 

 need not be surprised to find an inversion of the layers in other mam- 

 mals than those in which it has already been established, especially 

 in the light of Mall's ('93) paper on a human embryo of the second 

 week, m which he explains the conditions as the result of inversion ; 

 but, for the reasons I have given, I do not consider the bridge in the 

 pig as even a potential inversion, — the less so as the explanation 

 I am about to offer is to my mind a much more satisfactory inter- 

 pretation of the phenomena. If it should be urged that the bridge is 

 homologous with the roof-like structure over the germinal disk of the 

 hedgehog (Hubrecht '89), it would be necessary to assume the ex- 

 istence of a potential opening through this structure into the cavity 

 beneath it from the very beginning of its formation, but the facts, as 

 recorded for the hedgehog, seem to furnish no grounds for this assump- 

 tion. Moreover, in this animal the roof-like structure never comes in 

 contact with the ectoderm of the germinal disk, but contributes to the 

 formation of the placenta, a very different fate from that of the bridge 

 in the pig. 



I am inclined to compare this bridge with the overgrowth which 

 occurs in the development of Amphioxus just after gastrulation and 

 the elongation of the embryo have taken place. We knew through 

 the investigations of Kowalevsky ('67 and '76) and of Hatschek ('81) 

 that at this time there is a sinking of both ectoderm and entoderm 

 along one side of the gastrnla, the future dorsal or neural side of the 

 animal, which forms the so called medullary plate, and at about the 

 same time a proliferation of cells begins at the posterior margin of the 

 blastopore, and, growing forward over the blastopore and the medullary 

 plate, meets lateral elevations on either side, and fusing with them 

 forms a continuous roof over the dorsal depression, with an opening at 

 the anterior end, which persists for a considerable time as the neuro- 

 pore. The neural tube is later formed from the medullary plate, 



