nerve-tubes spring) are for the motoric (centrifugal) nerve-tubes, the 
difference is that the former are situated outside the central nervous 
system and have got. a simple shape without many prosesses, pro- 
bably because they can easily get their nutrition without special 
organs; whilst the latter are situated in the central nervous system 
and need generally many and long processes to reach the blood- 
vessels or the periphery of the nervous organs and there obtain 
sufficient nutrition for the cell and its nervous process. 
From the recent works by my much estimated friend Adolf 
D. Onodi (1. c. 1884, 1885, 1886) (cf also the previous papers by 
Marshall, 1. c. 1877; Balfour, 1. c. 1876 & 1877; Hensen, 1. c. 
1876; SCHENK, 1. c. 1876) we have learnt that the spinal ganglia 
spring from the spinal cord. We are thus entitled to say there is 
no principal difference between the ganglion cells of the nervous 
system and those of the spinal ganglia, they have the same origin 
and the same function but they have got a different situation and 
differ therefore also in shape, some are unipolar whilst other are hi- 
polar, and even miUtipolar, ivhich, hotvever, does not make any diffe- 
rence in their function. 
I have not yet had opportunity to examine in invertebrates how 
the relations are as to the nutritive centres of those nerve-tubes 
which spring from the dotted substance by a union of slender tubes. 
Whether the ganglion cells which I have found in the peripheral 
nerves of various animals (cf my paper on Mysostoma 1885 p. 31 
& 74 and this paper p. 141 & 143, note 2) are of importance in 
this respect I can not decide at present, it is, at all events, not 
impossible. In the peripheral nerves of Patella I have, as mentioned 
p. 141 found ganglion cells which had even protopla^mic processes 
(cf fig. 20, gc). 
Whether the ganglion cells of the central nerve-system have 
any functions besides being nutritive centres is, of course, extremely 
difficult to decide. 
It is not imposisble that they may be the seat of memory. 
A small part of each irritation producing a reflex action, may on 
its way through the dotted substance be absorbed by some branches 
of the nervous processes of the ganglion cells, and can possibly in 
one way or another be stored up in the latter. Howsoever that 
may be, and whatever the function of the ganglion cells is, this 
new theory of the combination between the centripetal and the 
centrifugal nerve-tubes gives, if approved, a quite new view of 
the importance of the dotted substance (or the interlacing of nerv- 
