Morphological Studies. 761 



are already being given off into the ganglion to form the lateral 

 ganglion of the olfactory i. e. those ganglionic cells which are derived 

 from the sensory thickening, as in the case of the branchial sense 

 organs connected with the other cranial segmental nerves, such as the 

 trigeminus, vagus &c. (36, 37)., This figure also shows that, like the 

 sensory epithelium of the lateral line or of any cranial segmental 

 nerve, the olfactory neuroepithelium is very early differentiated from 

 the rest of the epiblast, and grows by cell divisions within itself, 

 not by the addition of new epiblast cells. The growth is, so to speak, 

 interstitial. The upper and lower limits are marked off at a, b, and 

 beyond this point on each side lies the indifferent epiblast. 



In the next figure (fig. 21) this marked boundary between neuro- 

 epithelium and indifferent epiblast is very obvious at a and b. This 

 section is from a slightly older embryo, and shows that many cells 

 have already been given off into the ganglionic foundation , and that 

 others are about to pass into if. The ganglion is not yet connected 

 with the brain, but the connection takes place very soon after this. 

 It is hardly necessary to discuss here the question as to how this 

 connection of the ganglion of a sensory nerve with the brain takes place, 

 as I have considered it at some length in a recent publication (37, 

 pp. 171 — 173). But I believe that what holds for other cranial nerves 

 in this respect also obtains for the olfactory. 



Most certainly the final attachment to the brain is the only one 

 that ever exists, and there is no shifting of the root in Marshall's 

 sense (53, p. 11). In Elasmobranchs the whole or nearly the whole 

 of these ganglion cells, both neural and lateral ganglia, are used up 

 in the formation of the olfactory nerve. This is a curious fact, and 

 probably stands in connection with two things. 



Firstly, the olfactory organ has tremendous central ganglionic 

 centres in the much later developed (Marshall, 35) olfactory lobes, 

 and, though a large sense organ in these fishes, it is hardly likely to 

 be a very delicate one. In any case, though ganglionic cells are cer- 

 tainly formed from the sensory epithelium in Elasmobranchs, — a 

 fact, which satisfies all my contentions, — there are very few peri- 

 pheral ganglionic cells in the olfactory epithelium of Elasmobranchs. 

 Generally speaking, the number of ganglionic cells used up in the 

 formation of the Vertebrate olfactory nerve is something astonishing, 

 and the cellular nature of this nerve, — even in man — has 

 been admitted by Kölluler (21) who otherwise agrees with 

 His (47 and elsewhere) in rejecting the cellular nature of em- 



