THE EMBRYOLOGY OF THE OPOSSUM 69 



origin of the nephric system should prove to be the correct 

 one, as the present writer believes it will, the last substantial 

 evidence for either the annelid or arthropod theory of the evo- 

 lution of the chordates will have been removed. For every 

 other structure in the chordate body which exhibits any degree 

 of metamerism can be shown to arise in the first instance as a 

 non-metameric structure. The medullary plate, the neural 

 crest, the mesoderm and its coelom, the notochord, the chon- 

 drostyle, and other structures to be mentioned below, all 

 acquire whatever degree of metamerism they possess secon- 

 darily. Even the coelomic pouches of Amphioxus arise 

 from a longitudinal groove in the enteron. All this seems 

 to mean that the chordates arose from unsegmented ancestors 

 and acquired a certain degree of metamerism independently 

 during their own evolution. 



The first branchial pouches. At stages 14 and 15 (second 

 normal stage plate) ectoderm and endoderm are in immediate 

 contact throughout the vesicle. At stage 16 the first meso- 

 dermal cells begin to separate them. By stage 20 the mesoderm 

 completely underlies the medullary plate and is beginning to 

 extend beyond it posteriorly. In stages 21, 22 and 23 a wide 

 vascular area is formed and the sheet of mesoderm is un- 

 interrupted throughout this region except in the midsagittal 

 line where the notochordal plate in the endoderm is in im- 

 mediate contact with the keel of the medullary plate (fig. 

 18, B). Now at stage 24 migration or resorption of meso- 

 derm at six different points allows ectoderm and endoderm 

 to make secondary contact with each other. 



Four of the six points of secondary contact constitute the 

 closing plates of the first two pairs of gill pouches (McCrady, 

 '36). These are best illustrated in figure 31 where under the 

 caption, first branchial pouches, stage 24, ectoderm and endo- 

 derm are seen to form a broad plate of contact medial to the 

 hearts and lateral to the medullary plate. I do not know of 

 any other animal in which branchial pouches have been recog- 

 nized before the pharynx has acquired a floor, but that these 

 are actually the branchial pouches is obvious enough from the 



