476 C. JUDSON HERRICK 



Since the details of this application in no way modify the general 

 conclusions, I shall not take them up in this contribution, save for 

 one further point, in which the very recent literature is in some con- 

 confusion, growing out of a failure to effect a correct analysis of 

 the most rostral part of the diencephalon. 



Johnston in a series of very important papers (see especially 

 '06, p. 308, and '09, p. 522) has described the primordial hippo- 

 campus (his dorsal part of the epistriatum in fishes) as directly 

 continuous with a structure in the unevaginated telencephalon 

 medium which he terms the caudal part of the epistriatum. We 

 have seen that the latter structure in amphibians and reptiles 

 borders the taenia thalami behind the velum transversum and is 

 therefore wholly diencephalic, and that two and sometimes three 

 parts of the diencephalon are represented within it. Its most 

 ventral part is the rostral end of the pars ventralis thalami, or 

 eminentia thalami, bounded above by the sulcus medius. Its 

 dorsal part may be formed by the pars dorsalis thalami in 

 mammals, or by the epithalamus (fig. 18) or by both of these 

 structures (rig. 49). We have seen further that the massive con- 

 nection of the hippocampal formation with these structures as 

 seen in Amphibia and some Reptilia is secondarily acquired and 

 is not of great morphological significance. 



The mammalian structure corresponding to Johnston's caudal 

 part of the epistriatum was first clearly described for foetal Orni- 

 thorhynchus by Elliot Smith ('96) under the name paraphysis. 

 A study of Elliot Smith's descriptions and figures in comparison 

 with embryos of eutherian mammals makes it evident that this 

 structure is nothing other than the rostral border of the dienceph- 

 alon adjacent to the taenia thalami, and this has since been ex- 

 pressly stated to be the case by Professor Smith himself ('08, p. 

 535). Figures 9, 10, 11 and 12 of the original memoir ('96) give 

 relations of the pars ventralis and pars dorsalis thalami, sulcus 

 medius and sulcus ventralis, almost identical with those of embry- 

 onic reptiles (cf. my fig. 49). There is, therefore, no direct mor- 

 phological relationship between the " paraphysis" of Elliot Smith's 

 earlier description and the hippocampal formation. 



Now returning to the amphibian brain, let us examine a section 



