81-1 PROCEEDINGS OF THE AMERICAN ACADEMY. 



There are several objections to this theory which suggest them- 

 selves at once. In the first place, as to the structure of the two 

 overgrowths ; that in Amphioxus is figured as consisting of but one 

 layer of ectodermal cells ; the bridge in the pig, on the contrary, is 

 from two to three layers of cells thick. This seems to me but a 

 minor point, however, readily explained by the morphological differ- 

 ences between the larva of Amphioxus and the blastodermic vesicle 

 of the Mammalia. In the former case there is no part which may not 

 be said to develop directly into some important tissue or organ of the 

 adult animal ; in the latter, it is essentially the germinal disk only of 

 which this is true. Here, then, we have a germinal and an extra- 

 germinal region, and in the process of rapidly increasing the number 

 of cells in the disk the extra-germinal cells of the ectoderm, as well 

 as the cells of the disk, apparently contribute to the development 

 of the bridge, thus producing an overgrowth two or three cells 

 thick. Again, it may be urged that I have not shown a blastopore 

 around which the overgrowth should take place. I have found in my 

 youngest embryos no trace of an opening through the germinal disk 

 which could be compared with wnat Hertwig ('93) says may pos- 

 sibly be a blastopore, as figured by Heape ('83), Selenka ('86-'87)j 

 and Keibel ('89). But the blastopore in Amphioxus becomes the 

 neurenteric canal, which leads from the posterior end of the cavity 

 beneath the overgrowth into the gastrula cavity. It is at just this 

 point in several of my embryos that I find a canal leading from 

 the cavity beneath the bridge to the cavity located at the margin 

 of the disk, between ectoderm and entoderm. To be a true neuren- 

 teric canal, it should be continued into the gastrula cavity, i. e. the 

 cavity within the entoderm ; but on account of the loose connection 

 of the entodermal cells, it is impossible here to trace any such pas- 

 sage. I do not, however, think we should be justified in asserting 

 on this account that it does not exist. There ought, however, to 

 be at least a fusion of the two primary germ layers in this region, 

 such as always occurs, I believe, in the case of a true neurenteric 

 canal. 



Still another objection may be raised to this theory, and that is 

 the extremely remote relationship existing between Amphioxus and 

 the Mammalia. If the bridge of the pig is really a reappearance of 

 a structure which occurs as far back in phylogeny as the lowest 

 representative of the Vertebrata, why may we not justly look for its 

 presence in intermediate groups ? The question is certainly a fair 

 one, and I can only say that, so far as I am aware, no structure 



