L. DO^’CASTKK, A\D .1. (UtAV. 
4S4 
in 19 1 1 consisted of segmenting eggs of these three species, 
and of the hybrids esculentns ? x acntus (^ , acutus ? 
X escnlentus (5' , and acntns ? x milaris , preserved 
at intervals of from half an honr to five lioni'S after fertili- 
sation. 
In 1912, besides a few additional batches of the same 
c losses, we obtained eggs of esculentns ? x mill aids 
S iiii d m i 1 i a r i s $ x e s c n 1 e n t n s and acntns ^ . The 
great majority of the egg’s sent to ns were in stages from 
shortl}" before the conjugation of the pronunclei up to 2- or 
4-cell stages ; very few had reached the third segmentation 
division. Several samples of eggs were from batches the 
remainder of which were reared to plutei. 
The preservatives used were sublimate-acetic, sublimate- 
nitric, Flemming’.s, Hermann’s and Perenyi’s fluids. The 
preservation varied in different cases ; in general, sublimate- 
acetic and Flemming gave the best results. The eggs were 
sectioned (the sections we made ourselves were 5 or 7 ju in 
thickness), and stained on the slide with Heidenhain’s iron- 
hannatoxylin. In counting chromosomes we have found it 
absolutely necessary to draw every spindle; repeated 
attempts to count by eye have shown that by this method a 
number smaller than the true one is usually recorded. 
Hefore desctibing the hybrid eggs, a few words are needed 
on those of the pure species. The greater part of the work 
has dealt with crosses between acutus and esculentns ; of 
pure miliaris comparatively little material was available, 
and an account of it will be postponed until the miliaris 
hybrids are dealt with. Of acutus and esculentns we 
had a considerable supply of material from the early stages 
of fertilisation up to the second segmentation division. No 
account is needed of the stages of fertilisation nor of details 
of the segmentation mitoses except those relating to the 
chromosomes. We have determined the chromosome number 
in both species as 38; in early anaphase groups seen in 
face this number can frequently be counted with confidence 
(I’l. 28, figs. 1, 2, 3j, and where a lower number is found it is 
