290 THE BRAIN OF THE TIGER SALAMANDER 



A survey of the commissures and decussations of all vertebrate 

 brains shows that some of them, like the posterior commissure and 

 optic chiasma, hold constant positions throughout the series, while 

 others vary widely in position and composition. It is evident that in 

 the latter cases the site of crossing of a particular system of fibers is 

 determined by the topographic arrangement of the parts to be con- 

 nected. These arrangements are widely diversified in the lower groups 

 of vertebrates and the crossing fibers tend to take the shortest avail- 

 able pathways. But when the course of a particular decussating tract 

 has been established in an ancestral species, the original site of the 

 decussation may be retained, despite great changes in the relative 

 sizes of parts of the brain in phylogenetic descendants of the ancestral 

 form, so that the tract may take a circuitous course to reach its decus- 

 sation, as illustrated by some components of the postoptic complex in 

 fishes. This conservatism may be accounted for by the fact that the 

 decussating fibers may have collateral connections at any part of the 

 course both before and after crossing. These considerations suggest 

 that great caution should be observed in using the actual sites of 

 crossing of the various systems of fibers as criteria of homology. 

 Some of them are very stable; in other instances fibers with similar 

 origins and terminations may cross in surprisingly different places, 

 as illustrated by the various courses taken by fibers of the hippo- 

 campal commissure. 



Primitively the roof plate of the brain, unlike the floor plate, was 

 membranous, and the locations of commissures which invade it are 

 determined by the functional requirements of the organs differenti- 

 ated below it in the various species of animals. In the telencephalon 

 the topographic arrangements of parts are extremely variable. In all 

 cases there is a wide interruption of the commissures at the site of 

 the stem-hemisphere fissure, paraphysis, dorsal sac, and their deriva- 

 tives. In the diencephalon the habenular commissure is separated by 

 the pineal recess from the commissura tecti, and the latter is con- 

 tinuous spinal ward with the posterior commissure and the com. tecti 

 of the mesencephalon. In cyclostomes the middle part of the mesen- 

 cephalic tectum is a membranous chorioid plexus, and the com. tecti 

 is here interrupted. 



At the posterior end of the tectum its commissure is continuous 

 with a thin sheet of crossed and uncrossed fibers in the anterior 

 medullary velum, and this, in turn, with the massive cerebellar com- 

 missures. Between the cerebellum and the calamus scriptorius the 



