90 
Dr. T. A. il?rij)mau on Collo'plirys avis. 
together, and the one C. avis I have mounted differs in 
several points from these. 
The proximal ventral projection (saccus) is in G. avis 
broader but with straighter sides, and therefore sharper 
point, than it is in C. ruin. The dorsal hooks are at their 
bases suddenly bent rather than curved and are thicker, 
especially towards their ends; their to.pering extremities 
being shorter in C. avis than in G. rubi. The exserted 
portion of the shaft of the aedoeagus is rather more slender. 
The dense portion of the clasps is very distinctly more long 
and slender in G. avis (as 7 to 6). 
In the medio-dorsal line the proximal margin is a sharp 
angle in G. avis, an open arch in G. ruin, whilst the distal 
margin is narrower in G. avis and the denser chitinous 
elements more solidly knitted together. 
The preparations I have made of the female abdomen 
(Plate XXT, fig. 2 ; Plate XXII) are far from being as satis¬ 
factory as I should like, but one or two points of difference 
between the two species, G. avis and G. rubi, are evident 
enough. 
The long chitinous tube in the eighth segment (corre¬ 
sponding to the very long aedoeagus of the male Theclid), 
is narrower in G. avis (two specimens examined) especially 
towards its lower end, so that in G. avis it gradually in¬ 
creases in width from its lower end forwards. In G. rubi 
the narrowest part is about the middle, thence enlarging 
towards either end. The tube is also distinctly longer in 
G. rubi than in G. avis, though I have a specimen of rubi 
that approaches G. avis more nearly than the one photo¬ 
graphed. The two remarkable spinous organs in the wall 
of the bursa seem to be nearly alike, viz. each with two 
long sharp spines; in G. riibi there seems to be much 
variation in the length of the second spine; in G. avis they 
are both equal, but an examination of further specimens 
may (or may not) show them to be as variable as in G. rubi. 
The reason for supposing they are not so variable in G. avis 
is, of course, that in other characters in which the two 
species differ, G. rubi is always the more variable. 
I took my first specimen in the little range of hills 
between Hyeres and Carquieranne; of these hills the 
best known is the Gollinc, des Oiscanx, near Costcbelle, and 
so, assuming this fly to be one of the birds for which that 
summit is famous, I have given it the specific name of 
Avis. 
