The Morphology of Functional Activity in the (ianglion Colls etc. 508 
witli an unbroken series of sections so that no cell was unaccounted 
for, tliere were thirty ceUs of this main type in every case. There is 
little likeliliood of error in the total nuinber. In the further division 
into central and centro-peripheral cells, in three ganglia fonrteen cells 
were apportioned to the first and sixteen to tlie second in the initial 
aiTangenient before the size relations were calculated, wdiich is the fact 
that called attention to the point. In the other two, a consideration 
of the size volnmes made this the more satisfactory rearrangement, one 
or two cells being shifted. The subdivision, therefore, is not wdthout 
the same indication. Again, the existence of two cells and only tw^o of 
the central motor type in the supra-oesophageal ganglion already noted 
(Table I) is a more con\ancing proof of the constancy of numbers, amoiig 
certain types at least. 
From the results of two investigators there is evidence that nerve 
cells in the constancy of theh' niimber form the most conspicuous ex- 
ception to what appears at present the more general rule of variability 
in number for other cells, which determines differenccs in body and organ 
size. Gaule (1889) first concluded from a study of the spinal cord of 
the frog that the number of ganglion cells might be considered a constant 
factor. Donaldson (1895) thinks that there is a high degree of constancy 
in the dctermination of neuroblasts. 
The changes produced by activity exclusive of those of absolute and 
relative size. 
As the starting point for activity, the resting cells of the four groups 
are to be first described (Figs. 1, 8, 13 and 18). In common, they are 
distinguished by theü’ compact strueture, the absence of any edema, 
and a generaUy deeper and more distinctly blue color after toluidin blue 
as compared with the later stages. As akeady mentioned, the presence 
of the intracellular axone, though it is not so clear cut as in later stages, 
separates the presumably motor cells containing it (Figs. 1 and 8) from 
the sensory cells which do not. Both the types of motor resting cells 
have usually a more markedly irregulär coutour in their resting con- 
dition than the sensory cells, due to pseudopodial-like protoplasmic pro- 
cesses. That this has any significance does not appear. It is regarded 
only as an accommodation in arrangement with the supportive tissue, 
for with the state of activity the cell outlines smooth out. As also pre- 
viously stated, the main difference between the purely centrally con- 
tained motor neurones and those sending fibres peripherally is one of 
