MATURATION OF UNFERTILISED EGG IN TENTHREDINID®. 575 
normal nucleus, and seems to be enclosed in a faint mem- 
brane ; in others it is plainly a group of separate chromosomes 
(fig. 19). It seems to be quite irregular in its occurrence, 
since it may be present in one egg and absent from another 
lying by its side in the same leaf. All the eggs of this batch 
are well advanced, with numerous nuclei scattered through 
the yolk, so that I have no means of determining the origin 
of this nucleus with certainty, but it seems probable either 
that the mass of chromosomes which sometimes arises from 
the polar nuclei has persisted, or that exceptionally a conju- 
gation of nuclei has taken place, which has given rise to the 
structure described. 
The mitotic figures of P. luteolum are more clear and 
distinct than in the other species observed. The chromosome 
number is plainly seen to be eight both in the maturation 
mitoses (figs. 27, 29), and in those found later during seg- 
mentation and in the blastoderm (figs. 28, 30), so that, as in 
N. ribesii, no doubling of the number takes place in the 
somatic nuclei. Centrosomes are visible in the segmentation 
mitoses, but are smaller and less conspicuous than in N. 
ribesii, and only appear in rather deeply-stained figures 
(fig. 31). I have not been able to find them in the maturation 
spindles. 
7. CR@SUS VARUS. 
C. varus also belongs to the class in which virgin eggs yield 
females only, and the male is not certainly known. The eggs 
are laid in rows on the veins on the under side of the leaf of the 
alder. I obtained only a small amount of material, and failed 
to get the flies to lay in confinement, so that it was necessary 
to put them in muslin bags on bushes. For this reason I 
could not determine the age of the eggs with complete 
accuracy, and my material was further reduced by the fact 
that the eggs of this species are very difficult to embed, and 
many were spoiled in cutting. 
The earliest stage observed is shortly after the second 
