214 BOTANICAL GAZETTE. [ September, 
The open mouths are also of advantage in making the flow- 
ers more conspicuous, but are to some extent a disadvantage, 
since they make the nectar more accessible to many insects 
which areuseless. The horn partly offsets this disadvantage. 
But for its presence small insects could crawl bodily into the 
hoods of the larger flowers. The horn has also the effect of 
making the nectar more or less double, notably in A. pur- 
purascens ; and I have observed in A. Cornuti and Sullivantii 
bumble-bees insert their tongues regularly on each side 
of it 
Since a small gynostegium can catch more processes on 
an insect’s leg, and so can fasten more corpuscula directly 
upon the insect, the habit of forming combinations of corpus- 
cula’ is less important to flowers having it. I have found no 
combinations of corpuscula of A. tuberosa, and few small 
ones of verticillata. (1) One advantage of their formation, 
and a ready explanation of their frequency in certain species, 
as A. incarnata, is to be found in the fact that the broken 
retinacula are often more easily caught by the wings than 
the hairs. Often the hairs are so short that they do not easily 
enter the slits; but whena corpusculum comes to be fastened 
ple, a specimen of Apathus elatus has six corpuscula on its 
tongue, all in one combination, illustrating the fact that it 1s 
often easier for A. incarnata to fasten a combination to a hair 
that is once caught than to catch another of the same length. 
every available “process the carrying’ capacity of the leg is 
still indefinitely increased. This is a ranareatt in the large 
pride i Species that I do not believe they could have been 
flow oped until this habit had become fixed in the ‘smaller 
He the pulvillus of a hive-bee’s foot I found a com- 
I i. But for the 
, at the have required every foot to fae 
cula, : ou 
room for attachment. Bea oe ae 
ee es 
° For 7 
Vol. 11 a ag of these combinations, see Corry’s paper, Trans. Linn. Soc., Ser: % Bot 
