1966] 
Cooper — Eumenine bVasps 
247 
nearly impossible without abundant material. Yet it would require 
an unusual effort to collect just a few “wild” nests of Ancistrocerus 
antilope, for example, even though the wasp abounds in the region, 
and extraordinary luck were any such nests to prove recently closed 
by the mother and to contain unhatched eggs. In fourteen years I 
have found but one natural nest of this wasp, and that was con- 
structed in a vacated nest of Sceliphron and contained but two cells. 
Any who wish to check or extend my results, or to investigate other 
aspects of the immature stages of Ancistrocerus and its burrow-nest- 
ing allies, will find the use of “trap nests” very helpful (Cooper 
1953 ). 
Although hitherto not known to occur in Hymenoptera (van 
Emden 1925, 1946), it is likely that ruptor ovi will be found to be 
widespread among the eumenine wasps at the very least. A second 
species of eumenine nesting in my traps, which regrettably I did not 
rear, also has ruptor ovi developed in the very same sites on the 
thorax and on the first three abdominal segments as those of An- 
cistrocerus antilope ; they have, however, a distinctly different patern 
of grouping and fusion of their cuticular spines. Ruptor ovi may 
thus prove serviceable in generic or specific identification of first 
instar larvae of eumenine wasps. It is also probable that ruptor ovi 
will be found elsewhere among the aculeates, even though the few 
descriptions and figures which I have found of wasp larvae at instar- 1 
fail to suggest their presence; for example, the cases of Sapyga (Soika 
1832), Crabro (Hachfeld 1948), Omalus , Chrysis, Astata, Stizus, 
Sceliphron, Philanthus, and Tachysphex (Grandi 1961), and so on. 
By analogy with the polyphagous Coleoptera, however, where a dif- 
ferent thpraco-abdominal distribution of egg bursters is common 
even though species of many genera and perhaps families are without 
them (van Emden 1946), it is quite possible that closely allied acu- 
leates do differ widely in their means of hatching, and ruptor ovi 
are not universal among them. 
Although five appears to be the primitive, most frequent, and upper 
number of larval instars found in Hymenoptera (see Bischoff 1927, 
DeBach and Schlinger 1964), it might be argued that, as in Tineola 
and some dermestid beetles, the number of moults may in principle 
be indeterminate. Certainly excessive feeding does not lead to a 
sixth instar larva. When a fifth instar larva is provided continuously 
with food, it may eat prodigiously (Cooper 1957), but there comes 
a point at which it ceases feeding. This is not, however, followed 
by a moult that gives rise to an additional larval instar. The ex- 
cessively corpulent larva simply follows the normal routine that 
marks the close of the fifth instar: cell cleaning, cocooning, passage 
