The Morphology of Functional Activity in the Ganglion cells etc. 491 
two minutes in 1% erythrosin, heated to 40°. After washing thoroughly, 
they were transferred to a saturated aqneous solution of toluidin blue 
for usually seven minutes, then differentiated in 95% alcohol, passed 
througli absolute alcohol and xylol, and mounted in xylol balsam. 
In addition, various mixtures of chromic and osmic acids have been 
used, followed by staining with iron-hematoxylin or combinations of 
acid and basic coal-tar dyes. As these afford no better preparations 
for the purpose pf this study and add nothing pertinent to the data pre- 
sented, they need not be discussed in detail. 
The technic of size computation and comparison. 
The technic of cell measurement. — As the process differs 
materiaUy from that employed in previous work, it seems desirable to 
state the reasons for the adequacy of each in detail. 
The measurements for cell body and nucleus in the plane of section 
were made by means of caraera lucida projection, the extent of the 
greatest major and minor axis of each element being marked upon the 
projected Image. These diameters were measured by a U. S. Standard 
rule graduated to one-half mülimetres. It was, however, soon discovered, 
all sections being in serial, that the two diameters very commonly attained 
theh füllest extent in different sections or that the cell attained its 
fuUest size in one and the nucleus in another. The only consistent pro- 
cedure which would insure uniformity would obviously be to choose 
invariably the greatest diameter. The practice was, therefore, after 
measuring the apparently largest cell body and nucleus, to control these 
measurements by superimposing the projected Image of the adjacent 
sections and correcting any measurement which feU short. In the case 
of the smallest types, the füllest size feU usually within one section, though 
it was found essential not to neglect the control. In the larger types, 
whüe it was possibly more usual for the füllest size to fall obviously 
within two sections, at times as many as five or six had to be com- 
pared to insure that all diameters were of their greatest extent. 
Over three hundred preliminary measurements of crayfish ceUs 
which were made according to this procedure and from which the volumes 
were calculated, including the majority of aU ceUs from the first abdo- 
minal gangha of four animals and a varying number from others, showed 
that it was deficient and unsatisfactory. This was the procedure used 
in the several thousand measurements of Purkinje cells (1910), the un- 
known third dimension being taken to be equal to the known shorter 
