C. Dobell and M. W. Jepps 
823 
or 0-75 ju,, between the two measurements. One of us (M. W. J.) recorded 
the diameters of 18 of these 29 cysts as half a division smaller, and 11 
as half a division larger, than the other (C. D.). The two series of 
measurements when recorded graphically gave curves of closely similar 
form (Fig. 1). We concluded therefore that for the majority of cysts 
measured our findings would be identical, whilst in the case of the 
remainder the individual error would not be greater than one unit of the 
scale employed: and furthermore that this error would probably not 
influence the character of our curves. We decided to measure a large 
number of cysts (500) in every infection studied, in order to obtain an 
approximately true average value for the sizes of the cysts; and to 
measure half (250) of this number each in every 500, so as to eliminate 
as far as possible the personal error. Each series of measurements here 
recorded was obtained in this way. 
As we recorded all our measurements in units of the micrometer 
scale we found it most convenient to plot our curves in these terms. The 
conversion of the units into microns, as shown in the charts, was made 
directly on the charts themselves. For this reason the points on the 
curves do not always coincide with the integral values in microns on 
the abscissae. 
B. Determination of the differences in the dimensions of cysts when ex¬ 
amined alive and after fixing, staining, and mounting: ivith an inquiry 
into the factors which determine these differences. 
It is well known that the dimensions of cysts after fixation, staining, 
and mounting are not the same as those of the living cysts. As, however, 
it was not always possible to measure large numbers of living cysts at the 
time when they were obtained, it became necessary to rely upon fixed 
and stained material; and it therefore became necessary also to ascertain 
what effect this would have on our measurements. To determine this 
point we adopted two different methods: first, the measurement of a 
large number of cysts (from the same infection) both living and when 
fixed, stained, and mounted—the mean dimensions so obtained being 
then compared; and secondly, we devised a method for observing and 
measuring an individual cyst during the various processes of fixation, 
staining, and mounting in balsam. Throughout this paper the expression 
“fixed, stained and mounted in balsam” signifies—unless the contrary 
is stated—that fixation was effected with a modified form of Schaudinn’s 
fluid (saturated corrosive sublimate in water, two parts; absolute 
