6o 
Psyche 
[September 
last females close the nest at about the time it becomes difficult for a 
human to observe them without artificial illumination. So far as I 
know this is unique among digger wasps. 
(2) The females hunt flies which are at rest in herbaceous vege- 
tation. The wasps fly in amongst the plants and presumably snatch 
the flies from the leaves and branches. This is in marked contrast to 
other Bembix, which pursue flies visiting flowers or flying about ani- 
mals, dung, etc. The Hies taken by u-scripta are exceedingly diverse 
and generally rather small for the size of the wasp. 
( 3 ) The mound of earth at the nest entrance is leveled in a man- 
ner unlike that of any other North American Bembix , though bearing 
some resemblance to that of spinolae and several other species. 
(4) The nest is multicellular (apparently usually bicellular or 
tricellular) , much like that of nubilipennis. 
(5) The nest entrance is left open during provisioning but is at 
all other times closed from the inside (somewhat as in sayi). 
(6) The egg is laid on the first fly (as in all but a few of the more 
specialized species of Bembix.) 
(7) The cell is of simple structure and is not cleaned by the 
female. 
Of these seven characteristics, the first three are unique, the fourth 
and fifth unusual although not unique, the last two common to all the 
more generalized species of Bembix (several more characters of this 
nature could easily be tabulated). Thus the total picture is not unlike 
that obtained from a study of structure: several unusual features built 
into a basically primitive Bembix. The most interesting of these 
features pertain to the unusual time of flight of provisioning females 
and the fact that they take Diptera which are at rest in vegetation. 
The well developed ocelli probably represent part of the morphological 
component of this adaptive complex. Since the ancestral Bembix 
doubtless had ocelli of nearly normal sphecid type, it seems probable 
that u-scripta split off from this stock at a very early date, before the 
loss of ocelli became complete. This would also account for the fact 
that certain features of wing venation and male secondary sexual 
characters are more Stic tie Ha-\ike than Bembix- like. 1 he crepuscular 
provisioning behavior of u-scripta clearly enables the species to exploit 
a slightly different source of food from that of other species of the 
genus: a miscellany of mostly rather small flies which are snatched 
from their resting positions in vegetation. So far as can be judged at 
present, 'u-scripta is not strongly restricted ecologically ; perhaps the 
