MATURATION OF UNFERTILISED EGG IN TENTHREDINIDE. 577 
dicular to the edge of the egg. They give rise to four 
vesicular nuclei, which resemble those of P. luteolum in 
their arrangement, the two outer, i.e. the halves of the first 
polar nucleus, are close together near the edge of the egg, 
and are separated from the second polar nucleus by a wider 
interval than in N. ribesii. The outer nucleus degenerates 
rapidly, and is soon followed by the second; the inner or 
egg-nucleus sinks in and begins to divide as in other species 
(fig. 20,a and b). Meanwhile the second polar nucleus breaks 
up into chromosomes, the number in this case as in others 
being eight or about eight (fig. 21). This group of chromo- 
somes appears to persist without change for some time, but 
at a stage two hours later (probably about five hours after 
the egg is laid) they have disappeared, and nothing is found 
but nuclei in the yolk derived from the egg-nucleus. As in 
N. ribesii centrosomes are visible in the yolk-nuclei mitoses, 
but I have not been able to distinguish them in the polar 
spindles. 
This species, therefore, seems to resemble C. varus in 
the course of its development; in each the two outer polar 
nuclei degenerate, and the inner one breaks up into chromo- 
somes, which may each split into two, but which finally 
disappear. It resembles P. luteolum in the fact that no 
conjugation of polar nuclei takes place, but in the latter 
species all three nuclei usually disappear without resolution 
into chromosomes. 
9, CHROMOSOMES AND CENTROSOME. 
In all the species of which satisfactory sections across a 
mitotic spindle were obtained, the number of chromosomes 
going to each end of the spindle appeared to be eight. In 
some cases only seven, or even six, were observed, but more 
than eight were never found except when chromosomes 
already split were seen all together in the equatorial plate 
(i.e. fig. 25, a and b, which were superposed in the same 
section), 
