446 Gregory. — Spore- Formation in Leptosporangiate Ferns . 
Stevens 1 , using Scolopendrium vulgar e , Cystopteris fragilis and Pteris 
aquilina , described the division of the chromosomes by two longitudinal 
fissions — the first provided for by the longitudinal fission of the spireme 
thread in the spore-mother-cell, the second provided for by an exactly 
similar fission of the chromatin thread of the daughter-nuclei. Strasburger 2 
suggested that the process in these Ferns is the same as that in Osmunda , 
where tetrad-like bodies occur ; and recognized Stevens’s mistake in passing 
over the early appearance of the longitudinal fission which provides for the 
second (homotype) mitosis. In the essential point his view agrees with 
that of Stevens in indicating the absence of a qualitative reduction-division 
of the chromosomes. 
2. Methods. 
Fresh material was teased and examined after treatment with either 
acetic iodine green, acetic methyl green or acetic carmine. 
The fixing reagents used in the earlier part of the work were: (i) 
absolute alcohol ; (2) Hermann’s platino-acetic-osmic ; (3) Flemming’s 
weak solution (chrom-osmic-acetic), used both cold and hot; (4) Merkel’s 
chromo-platinic solution ; (5) 1 to 2 % of chromic acid in water. Of these, 
the last proved the most satisfactory. During the later part of the work 
a mixture was used, consisting of two parts of absolute alcohol to one part 
glacial acetic add. Small pieces of material placed in this for from 15 to 
20 minutes gave excellent results. 
From absolute alcohol the material was transferred to a mixture of 
alcohol and xylol, and thence, after a few minutes, to pure xylol, in which 
it was allowed to remain for periods varying from 15 minutes to 24 hours. 
Two changes of paraffin were used, the time allowed for penetration ranging 
from i\ hours up to 2 days. 
The sections varied in thickness from 5 to 20 \x. 
The stains found most useful were : (1) Heidenhain’s iron-alum- 
haematoxylin ; (2) carbolic fuchsin and Licht Grim ; and (3) Flemming’s 
triple stain (aniline-safranin, gentian violet, orange G). 
The best results were obtained from material preserved in acetic- 
absolute, washed for about \ hour in each of four changes of redistilled 
absolute alcohol, transferred to xylol for 15 minutes and allowed to remain 
for from to 4 hours in melted paraffin 3 . 
The outlines of the nuclear structures were much sharper in material 
prepared in this way and stained with Heidenhain’s haematoxylin, than in 
material which had been subjected to the prolonged process of embedding. 
1 Stevens, ’98. 2 Strasburger, ’00, p. 79 . 
3 I should like to take this opportunity of acknowledging my indebtedness to Professor Farmer 
for the advice as to methods which he has so kindly given me. 
