1887. | BOTANICAL GAZETTE. 215 
The frequent occurrence of these combinations spoils the 
theory that the corpuscula enter the stigmatic chamber. 
That view can neither explain how they are formed nor how 
they escape destruction. 
hile in ordinary flowers an insect may be a useful vis- 
itor if it can reach the nectar, in Asclepias many other con- 
ditions influence the insect relations. (1) Of visitors whose 
tongues are suited to the nectaries, many are useless, because 
they do not light upon-the flowers (Sphingidea, geriade 
and Trochilus).’ (2) Others because their legs are not long 
enough to extract pollinia. Megachile is common on A. 
tuberosa, but never, so far as observed, carries pollinia. (3) 
Others have legs long enough, but rest their feet so lightly 
on the flowers that they seldom effect pollination; e. g., 
Diptera and small butterflies. (4) Still others are not strong 
enough to free their claws from the slits and break the reti- 
nacula; in all, seventeen species were found to be killed on 
this account. 
The table shows the number of species visiting the flow- 
ers, with the disposition of the corpuscula on them. 
too} 
ce ie ae || 
e— i . v 
eel spere 1s 
a1/2|}38)/2)3/6 
B o Pee) eS o 
mi ade7| a | ok 
Corpuscula on hairs, claws and tongue....... 14 3 16 
" MURAIS S00 CIS WE: u.seves0ccciseons 7 16 
: ae hairs and tongue 5 1 tos 3 > 
‘i “ hairs alone 6115] 1} 7) 2] 3} 54 
ee “ claws alone 3 3 
‘i “ tongue alone tie er ees a 3 
: none 13/12] 6|13] 5 49 
701351 7|25| 91 4 |150 
The most striking peculiarity of Hymenoptera is the fre- 
quent occurrence of pollinia on their tongues. Of twenty- 
eight species with pollinia in that situation, twenty are Hy- 
menoptera. Bembex which resembles certain Syrphida* in 
* The ruby-throated humming-bird visits A. incarnata, Sullivantii and purpurascens. 
*In the table the pulvilli are included under “ hairs.” 
* Packard, ‘Guide to Study of Insects,” 164. 
