152 Wilson . — On Spore Formation and Nuclear 
over estimated. In the above description of the premeiotic prophases it 
was suggested that the increase of granularity in the nuclear reticulum was 
frequently a result of improper fixation due to the use of acetic alcohol. 
It is probable that in many cases the similar appearances obtained at this 
stage by other investigators are partly due to use of this or of other fixing 
reagents. Flemming’s Mixtures gave much better results at this stage, and 
this was especially the case when the dehydration was performed by the 
glycerine method (see p. 143). Too rapid changes in the concentration of 
the liquids used both at this and at later stages are quite sufficient to pro- 
duce an apparently early prophase. These difficulties can only be avoided 
by the use of great care during the transference of the material from water 
to alcohol and vice versa, and these remarks will also apply in a less degree 
to the change from alcohol to xylol, cedarwood oil, or chloroform. 
The use of acetic alcohol produces preparations which are character- 
ized by a ragged appearance, and possibly this is due to solution of a 
portion of the material composing the cell. As above noted, this reagent 
precipitates some of the proteids present in the plant in a form soluble in 
water, and these would be lost during the subsequent staining operations. 
Although well fixed preparations of the chromosomes are obtained by the 
use of acetic alcohol, this mixture is not a reliable fixing reagent, especially 
for the earlier stages of division. Farmer ( 10 , p. 490) has already noted 
that at the reduction divisions the spore-mother-cells of the Hepaticae can 
only be fixed with great difficulty. This is especially the case in the 
Muscineae. Methods which resulted in perfect fixation of the premeiotic 
stages, gave very inferior results at the reduction divisions, and at this 
period acetic alcohol is valuable on account of its penetrating power. 
A large amount of discussion has taken place with regard to the 
formation of the bivalent chromosomes at the meiosis. It is generally 
admitted that at the heterotype division the equivalents of somatic chromo- 
somes are separated, but there is considerable divergence of opinion con- 
cerning the details of formation of these bodies. 
Farmer and Moore (8) find that a longitudinal fission of the spireme 
takes place during or shortly subsequent to the first contraction ; this 
fission temporarily closes during the second contraction. Approximation 
of the thread into more or less parallel lines now goes on, whether by 
looping or otherwise. At emergence from the second contraction traces 
of the original longitudinal fission can be found in the spireme, and these 
persist in the individual chromosomes. The latter are bivalent, and con- 
sist of two somatic chromosomes fused end to end. At the metaphase 
transverse division goes on, and during the anaphase the original longi- 
tudinal fission may again be seen in the chromosomes. 
Gregoire ( 14 ) considers that lateral approximation of the thin spireme 
takes place at or immediately before the first contraction. The parallel 
