1887. ] BOTANICAL GAZETTE. ae 
The great number of hive-bees killed on the flowers of 
this plant and of A. Cornuti, besides being a matter of curi- 
osity, and, indeed, of economic importance, is interesting in 
the study of the insect relations of the different species. 
Dead hive-bees are found on the flowers of A. Sullivantii 
much more frequently than on A. Cornuti. From the flowers 
of a patch which bore fifty-two follicles, I picked 147 dead 
hive-bees, from which it seems that the flowers are bet- 
ter adapted to kill hive-bees than to produce fruit through 
their aid. On seventeen days between July 2 and 27, 1885, 
I visited a patch to collect the insect visitors, and picked 671 
from the flowers. I have often found four, and, in one case, 
aster to hive-bees. 
<i pollinia, fixed to some part of the tongue, and these 
the claws being blunted by the corpuscula. 
esides hive-bees, species of Megachile, Halictus, Astata, 
Lucilia,® Trichius, Pamphila, and Scepsis were found dead 
on the flowers 
poe A eS ae 
5 foe R. Bickford Ni ist, ii, 665. J. Kirkpatrick says: “‘ When the claws are 
pelingettered, th nap ajenchagaar ge a ag vorty & llect honey, and is soon €x- 
be sé limb upon the combs nor colle . ¢ 
tmogyeem the hive and must die. The watetwered bees tumble them out with little cer 
: m. Nat., iii, 109. 
i i e by Dr. S. W. 
wines Dip tera mentioned in this paper were kindly determined for me by 
