42 ILLINOIS BIOLOGICAL MONOGRAPHS [42 



which usually supports the anterior end of the telencephalon, is the only 

 connection between the two capsules. In Cryptobranchus, Amphiuma, 

 and Necturus, the planum basale of the larval stages is slightly convex, 

 due to the origin of the verticale from its median line; while on the other 

 hand the planum basale of Salamandra and Amblystoma are broadly 

 concave, being in no way associated with the development of the verticale. 

 The size of the basale is variable in the different groups; being very small 

 in Cryptobranchus, Spelerpes and the larval Amphiuma, it increases in 

 Plethodon and Salamandra, and reaches its greatest development in the 

 larvae of Amblystoma where the fused basale and verticale form a thick 

 anterior wall to the cavum cranii. Individuals approaching the end of 

 metamorphosis show a reduction in the size of the planum basale, and in 

 the adult stage it is greatly reduced, correlated with the development of 

 other parts of the capsule and the increase in size of the sensory parts. In 

 both larvae and adults of Triton and Diemictylus the planum basale is 

 very small, and in the latter, is the only connection between the two cap- 

 sules. 



The planum verticale arises in different ways throughout the order. 

 In Cryptobranchus, it arises from the median line of the basale as a small 

 ridge, and later, uniting to the median margins of the ethmoidal columns, 

 it forms a complete wall separating the internasal space from the cavity of 

 the forebrain. The same is true in Amphiuma, although evidence is lack- 

 ing as to the origin of the verticale, yet the later stages would suggest its 

 development from the planum basale as in Cryptobranchus. The early 

 larvae of Necturus have a small median ridge on the basale, probably a 

 vestigial verticale; but it never develops and is less prominent in the later 

 stage, a further evidence of the neotenic condition of this genus. 



In the remaining groups, the verticale does not arise from the basal 

 plate, but is developed by medial growths from the columnae ethmoidales 

 which subsequently unite to the basale, so that the results are identical. 

 In my material of Spelerpes, evidence of the formation of the verticale is 

 lacking; but judging from the relation of the early columna to the cornu, 

 and by the flat dorsal surface of the basale, it would appear that here, as in 

 Amblystoma, the verticale arises from the medial growths of the ethmoidal 

 columns. In a 25 mm. Salamandra larva, there is no evidence of a median 

 verticale plate. The cornua and columnae are united just anterior to the 

 planum basale, but internasal space and cavum cranii are continuous with 

 each other. In the later stage, medial growths from the anterior parts of 

 the columnae unite just above the anterior margin of the basale, and form 

 a pons ethmoidalis, separated from the ventral plate by a circular gap, the 

 fenestra ethmoidalis. This gap is only temporary in Salamandra, for by 

 the more ventral chondrification of the pons, a complete verticale is formed 

 and the internasal space and cavum cranii are completely cut off from each 



