ZOOLOGY: E. P. FELT 
353 
the agency of birds, since we have records of birds feeding upon the 
galls of Cincticornia and there would appear to be no reason why those 
produced by the genus Caryomyia might not be equally acceptable. 
The same would be true to a certain extent of galls inhabited by Astero- 
myia and some species of Oligotrophus. It would be possible for 
midge larvae in firm, moderately thick-walled galls, to pass through 
the digestive tract of the bird without material injury, but this would 
be true of relatively few and owing to obvious Kmitations we can hardly 
consider this method of dissemination as one of great importance so 
far as the group as a whole is concerned. 
The flight of gall midges, especially such a one as Miastor, is so feeble 
as to render it improbable that the wings of themselves are important 
in distributing the species over extended areas, and yet these midges 
are surprisingly prevalent where breeding conditions are favorable. 
The adults issue in swarms from decaying logs and apparently drift 
with the wind. The probability of this method of spread being an 
important one is supported by the recent investigations of McCulloch,^ 
which show that Hessian flies may drift at least 2 miles with the wind. 
It is probable that this method of dissemination is very effective in 
the case of a number of gall midges. 
Extended drifting, some possibly may prefer the word flight, with 
the wiAd, is not impossible when we recall that the migratory locust, 
Schistocerca peregrina has been found five hundred miles east of South 
America and is credited with probably having crossed from South 
America to Africa. Such extended flights are rendered more likely by 
the record of Sphinx convolvulus being found 420 miles from land, 
while other moths and longicorn beetles have been reported out at sea 
at a distance of 230 miles. The extensive northern flight of the cotton 
moth is another case of the same kind. Recent investigations by 
ColHns^ have shown that young gipsy moth caterpillars may be carried 
20 to 25 miles by the wind — a distance which presumably could be 
greatly increased without endangering larval existence. 
With the above data in mind it seems probable that the carriage of 
gall midges by floating vegetation and drifting islands, or by the activities 
of birds, has occurred to only a very limited extent and can not begin to 
explain the general distribution of these fragile forms. Winds un- 
doubtedly carry these insects for short distances and might even, under 
exceptional conditions, transport them over seas, though this latter 
is by no means established and the present known distribution appears 
to be most easily explained by migrations (probably drifting with the 
wind in this group) over land areas about the same way as other mem- 
