38 GEORGE BIDDER. 
spherical increasing the breadth in proportion to the height. 
In Series C the measurements give no evidence of contraction 
in this dimension ; but the cells are spherical and even orange- 
shaped, showing that the absence of change in anatomical 
dimensions is no guarantee against the most profound cellular 
distortion. 
MertuHobps. 
The main practical conclusion was that cell-form tends to be 
profoundly modified in the passage between alcohol and paraflin 
solvents, and that this may unfortunately be the case even in 
the process of staining on the slide. It seemed likely that the 
dangers of the embedding process are modified by very gradual 
dialysis from alcohol into benzol, and largely guarded against 
by super-hardening in 1 per cent. osmic acid and in absolute 
alcohol. For osmic acid even the sponge tissue requires to be 
cut in the smallest practicable pieces and repeatedly shaken, 
otherwise the inner chambers are not thoroughly hardened ; 
the exposure used was one hour in the dark. Dialysis from 
water into absolute alcohol, or from alcohol into benzol, each 
took from six to twelve hours; they were left up to fifteen 
hours with good results. The best preparation (Series D) was 
stained in bulk with equal parts of Grenacher’s hematoxylin 
and 70 per cent. alcohol, being brought into this solution from 
40 per cent. alcohol by four equal changes of strength; no acid 
was used, and the result was a very valuable overstaining of 
the collars and iris membranes. The sections were fixed with 
water, the paraffin cleared in chloroform. It will be found 
convenient to have in a pipette a thin solution of balsam in 
chloroform, so that it can be squirted instantly on the sections 
after removal from the chloroform, to prevent drying before 
the thicker balsam has time to spread. 
It will be seen that I am greatly indebted to the methods of 
Vosmaer and Pekelharing (19), which were closely followed 
up to the stage of embedding in paraffin; but I am convinced 
that staining on the slide is highly destructive of cell-form, 
unless the transference from benzol to alcohol be effected with 
