190 
A. W. LINDSEY 
there is nothing to show the fate of the cubitals except several 
degrees of stalking of Cu^ and M3, the complete loss of one branch 
in THckoptilus, and the complete or partial loss, by independent 
reduction, of the remaining branch in T. pygmaeus Wlsm. 
The radius of the primaries is the one o-ther vein whose history 
is complex. The fact that its fifth branch runs to the apex of the 
wing in Agdistis but ends behind the apex in all of the cleft- 
winged genera where it is indubitably present, suggests funda- 
mental separation of these lines, as has long been the writer’s 
view. A significant fact regarding this vein is that no genus re- 
taining five branches of the radius has the fifth stalked with an- 
other, although the third and fourth are frequently stalked, (and 
the second with them in some specimens of Pselnophorus heU 
fragei Fish, with only four branches remaining) . In the genera 
retaining only four branches of the radius, several things may be 
noted. First, when the posterior branches are stalked, the most 
posterior ends in the apex of the wing, both conditions suggest- 
ing that it is R4 or a complex including this vein. Second, when 
the most posterior branch of the four is not stalked, it ends be- 
hind the apex, and excepting Oidaematophorus , is preceded by 
stalked branches, both conditions suggesting that it is R5. Third, 
when R5 as thus interpreted is present, it appears that the pro- 
gressive anastomosis of Ro, Rr. and R^, actually visible in a series 
of P. belfragei Fish, has resulted in the coincidence of two 
branches. These would naturally be the two showing the earliest 
and most prevalent tendency to unite, viz., Rg and R4, and the vein 
ending in the apex of the primaries must be Rg+i- This is. a 
condition such as seen in the wings of Pterophorus periscelidac- 
tylus Fitch. (Fig. 2 ) 
In T. pygmaeus Wlsm. three radial branches at the most are 
present, and the first two show partial reduction without stalk- 
ing. Apparently then both R4 and Ro may be lost in this way. 
The writer’s previous association of the genus, phylogenetically, 
with Pterophorus (Oxyptilus) cannot stand, since in the latter 
Ro is invariably stalked. Meyrick’s diagnosis of Trichoptilus 
states, however, that Ro may be stalked. This would indicate 
that some of the exotic species, not available to the writer for ex- 
amination, represent a different line of evolution which may 
properly be derived from Pterophorus (Oxyptilus) . 
The condition of the veins in some of the more specialized 
genera, of which Aciptilia is the most extreme, is one which can- 
