558 



more important than their position over the pericardium. In Triton 

 the postbrauchial body is found behind the fifth branchial 

 cleft, exactly in the place M'here one would repect a sixth cleft. 



Maurer tells us further that the postbrauchial body, which is 

 paired in the Anura, develops only on one side — the left 

 — in the Urodela. 



The following brief account of the development of the thyroid 

 and suprapericardial bodies in Necturus maculosus calls attention 

 to the fact that their development in Necturus differs from that 

 in Triton and Siredon in one or two points on which Maurer 

 lays stress. His description is therefore not applicable, as he supposed, 

 to the Urodela as a whole. 



The Thyroid Gland. 

 When the head of the embryo is folded off from the yolk, the 

 anterior extremity of the alimentary canal closes first, then the ec- 

 toderm surrounds it, being separated laterally from the endoderm by 

 the mesoderm^) of the mandibular arch, which extends backwards 

 almost horizontally, and first fuses with the mesodem of the opposite 

 side at the point where the mandibular mesoderm unites ventrally 

 with that of the hyoid arch. Thus two pairs of mesodermic arches 

 arise from the same ventral space ; the hyoid extending dorso-laterally, 

 while the mandibular extend in an autero-lateral direction. Between 

 the hyoid and mandibular arches the ectoderm and endoderm meet in 

 the oblique lines of the hyomandibular clefts, while between the 

 mandibular arches, the ectoderm and endoderm fuse in a median, 

 axially elongated, space, of which the relative length and width is 

 shown in fig. 1 by the broken line in continuity with the posterior 

 part of the T-shaped oral fusion which the figure represents. The 

 lateral projections which are outhned, appear later. From the posterior 

 extremity of this median ventral fusion, the floor of the alimentary 

 canal rises dorsally over the median mesoderm, which, as just de- 

 scribed, unites the four anterior branchial arches, and then turns to 

 a horizontal plane. Before taking a horizontal direction, however, the 

 floor of the alimentary canal forms the anterior wall of the ventral 



1) While retaining the expression "mesectoderm" for median tissues 

 which I know to be of ectodermic origin, I return to the familiar term 

 "mesoderm" in place of "mesendoderm'', in designating the remaining 

 median tissues for reasons to be given in a subsequent paper on the de- 

 velopment of cartilage. These reasons, however, do not lessen the 

 evidence against regarding the mesoderm as a germ layer. 



