The Morphology of Functional Activity in the Ganglion Cells etc. 493 
Of course, many cells and nuclei end within the depths of a section. It 
is possible, witli the help of the optical section to deterraine whether 
they barely enter, rim about half way or nearly through it. By laying 
a fixed value upon these estimations of two, three and four micra, un- 
doubtedly a closer approximation was reached. The accuracy of the 
whole procedure as applied to such work as this depends largely upon 
the uniform working and general trustworthiness of the microtome used. 
Two instrunients of different make were einployed, which in some measure 
is of value as a control, there being no appreciable difference in the re- 
sults from either. Outside of this there is no defense to be made save 
in the character of the results obtained. There are many things when 
all the points about a given cell under observation are considered which 
have served to inspire confidence. The uniformity of the results from 
different animals wiU speak for itself but at the moment attention is 
called to how commonly in the majority of cells in the tables submitted 
the third dimension closely or exactly coincides with one of the others, 
usually the transverse, since the longitudinal diameter, as stated, was 
given for the most part by direct measurement. This is particularly 
true of the nucleus which from its structural relations affords measure- 
ments obviously the more exact. For example, of the 115 sets of nuclear 
diameters in Table I\", in fifteen the third figure coincides with one or 
the other diameters, and in the remainder, it differs from one or the 
other of these by one micron in thirty, by two micra in thirty-nine and 
by three micra in twelve, leaving nineteen nuclei more irregulär. This 
illustration chosen at random concretely confirms the impression given 
by any of the series. 
In aU of the series after the preliminary work, no calcvdations were 
made until the measurements were completed. As subsequent correc- 
tions or changes introduce more of the factor of the personal equation, 
it is only fair to state that they were limited even where apparent dis- 
crepancies indicated a possible chance of error. A record of these was kept 
whether made on account of error or for justifiable reasons. In Table IV, 
there were six, in Table V, eight, in Table YI, none, in Table VII, 
one. This Statement does not apply to a comparatively small number 
of ceUs whose volumes showed that they did not belong to the series 
in hand and hence were either discarded or transferred to another in 
case there were several from the animal. Outside of the ordinary chances 
of error or Variation, which are frankly admitted to be great whether 
from the shape of the cell or the technic, no claim for exact accuracy 
of diagnosis is made. The differentiation of types itself is due largely 
