248 
Psyche 
[December 
of the meconium (which is voluminous), then entrance either to the 
so-called pronymphal stage and succeeding sixth moult, or to dia- 
pause, remaining an eonymph. Nor have I been able to influence the 
numbers of moults that surviving larvae undergo by attempting to 
starve them from the second instar on. Five larval moults, therefore, 
appear to be the normal number. On enquiry, Prof. O. W. Rich- 
ards informs me that it has been clearly shown that Vespa orientalis 
Linn, also has five larval instars. It is thtis possible that du Buysson 
( 1 903 ) erred in his determination of the larval moults of Vespa , 
and that possession of five larval instars is both primitive and quite 
general among Vespidae just as it seems to be for many other acu- 
leates, both bee (Bertholf 1925) and wasp (DeBach and Schlinger 
1964), as well as for the order Hymenoptera as a whole. Because 
no moult separates the fifth instar from the clearly defined prepupa, 
Morris’s (1937) suggestion that an instar has been lost in develop- 
ment of the Ichneumonidae (as compared with the “lower phyto- 
phagous Hymenoptera”) would seem to apply with equal force to 
eumenids and perhaps all other aculeates. 
The prepupa, or fully-fed, quiescent fifth instar larva, or eonymph, 
tends regularly to enter diapause in the case of our northeastern 
Symmorphuy cristatus there being but one brood each year. Yet 
exceptions occur, for in one trap nest the sibs in all cells save one 
entered diapause as usual ; the remaining individual transformed to a 
pupa which eclosed as a normal adult early that same fall. In prin- 
ciple then, S. cristatus is capable of being bivoltine. I have also had 
several prepupal Ancistrocerus antilope and Rygchium megaera re- 
main flaccid and dormant as eonymphal prepupae through two years, 
even though all of their sibs broke diapause more or less together 
and transformed to adults in the usual time. When placed out of 
doors in their third winter, these blocked prepupae transformed to 
normal adults in the following spring. They were at that time not 
less than three years old. Meade-Waldo (1913) has had similar 
experience with prolonged prepupal states (of 2-3 years) in Raphi- 
glossa flavo-ornata (P. Cam.) as has Williams (1919) with Tiphia 
ashmeadi Crawf. Morris (1937) comments that in Sweden nearly 
two thirds of the overwintering, diapaused, prepupal ichneumon fly 
Exenterus abruptorius Thb. emerged in the spring of 1935, and the 
remainder again over-wintered to emerge in the spring of 1936. 
Such instances indicate the possibility that recombinant genomes of 
these hymenopterans need not always be selectively tested or passed 
along within one generation, or even within immediately consecutive 
generations as may happen with regularly bivoltine species. 
There have been many observers who have commented on mason 
