304 
Psyche 
[September 
They are similar in size and both are yellowish with a darker meta- 
soma; in both there is a metallic sheen to the anterior part of the 
body, but the pompilid differs in having a black spot on the vertex 
and one on the pronotum. The pompilid has very long, slender, 
smooth legs; there is a constriction in the middle of the mesosoma 
exactly paralleling that in the Iridomyrmex worker; and the wings 
are so small as to escape immediate notice. The resemblance between 
the two is indeed striking. 
The second species (like the preceding known from one specimen) 
was collected by myself in a sandy blow-out in South Australia. 
This was an area in which a somewhat larger species of Iridomyr- 
mex , viridiaeneus Viehmeyer, 2 was abundant, and these ants were 
conspicuous because of their metallic bluish bodies. When I first 
observed the pompilid (described below as Iridomimus violaceus ) 
I had no suspicion that it was not simply another worker ant of 
this species. Something about its gait warned me that it was not, 
and close inspection showed that it had minute, stalked wings and 
that it lacked a nodose petiole. In size, color, and general form of 
the mesosoma it was an almost perfect copy of a worker Iridomyrmex 
viridiaeneus. 
Why should female pompilids with a well-developed and pre- 
sumably painful sting mimic ants? Iridomyrmex purpureus and cer- 
tain other species of this genus are known to produce methylheptenone 
and iridodial from anal glands; it is believed that the former serves 
as a defensive secretion while the latter may be an adherent, retarding 
the loss of more volatile components when discharged on a predator 
(Pavan and Trave, 1958; Cavill and Robertson, 1965). Thus these 
relatively abundant and conspicuous ants may be avoided by many 
or all predators. Presumably the pompilids hunt and nest in situa- 
tions where one particular species of Iridomyrmex is prevalent, and 
it proved selectively advantageous for them to assume a common 
Miillerian mimetic pattern with the ant. There may, of course, 
have been selection for flightlessness first (we do not know what 
kind of spider these pompilids prey on), and if so it would be par- 
ticularly advantageous for them to compensate for their inability to 
escape by assuming an ant-like form. It is curious that these wasps 
2 Dr. R. W. Taylor found my specimens to compare favorably with a 
syntype. This form has been considered a variety of the meat ant, 
Iridomyrmex purpureus Smith, but Taylor believes it is probably distinct. 
He states that it is probable that I. viridiaeneus and rubriceps are fairly 
closely related within Iridomyrmex. 
