1030 



If now we summarize the observations that have been made and 

 given above on these colonies of cells and the processes of the cells 

 that belong to them, it seems to be clearly shown that some at hast 

 of the apolar cells in these colonies grow out to new neurones during 

 the postembryonic growth of the animal. On the other hand It does 

 not seem to me so easy to decide how these colonies of cells arise. 

 The way is perhajDS that small cells from the capsule cells which 

 have been developed mitotically, or are at least situated within the 

 capsule, grow out into new ganglion cells, which are added to other 

 ganglion cells already existing within the same capsule. Might not a 

 relatively large ganglion cell, which in some respects is at an earlier 

 stage of development — for instance, apolar — increase in number 

 and become one of these colonies of cells by means of amitotic 

 divisions. 1 have not been able to decide with certainty whether 

 one or the other or both of these methods of formations occur, though, 

 as a matter of fact, there are signs in my preparations to support 

 the idea that both these methods of formation may occur. 



If, as seems to be shown above, a new formation of neurones in 

 the spinal ganglion i-eally occurs post-embryonally, one would and 

 might, of course, also expect to find, during post-embryonic life, 

 figures of growing axons in the peripheral nerves. I have examples 

 of such claviform figures, which are quite evident in sil\ er-impreg- 

 nated preparations of, for instance, the dorsal and ventral lOots of 

 young dogs. More details of this will, however, be given below. 



I consider that 1 have now shown that the cells. in the spinal 

 ganglia sufficiently explain the origin of the actually existing and 

 fairly considerable post-embryonic numeric growth of axons in the 

 dorsal roots of the spinal nerves. I shall now pass on to examine 

 to some extent in connection with those matters the 



Medulla spinalis. 

 There is but exceedingly scanty information about post-embryonic 

 divisions of the ganglion cells of the central nervous system to be 

 found in literature, and the existing accounts are not generally 

 admitted to be correct. These accounts, howexer, take two directions. 

 Some investigators maintain that the cells in this region divide by 

 means of mitoses, others say that the usual method of increase in 

 this case is that of amitotic cell division. 



Mitoses. 



Allen') states that in the spinal cord of an "albino rat" twelve 



1) Allen, Ezra. The cessation of mitosis in the central nervous system of the 

 Albino rat; J. Gomp. Neurol. Vol. 22, pp. 547—568, 1912. 



