1886. | BOTANICAL GAZETTE. 209 
was directed to this plant by seeing in one of its flowers a large 
dipterous insect vainly trying to climb up its walls. After failing 
in this it turned to the hairy style, up which it climbed with ease, 
and over the stigma and took flight. The interpretation of these 
movements were simple enough, if they were shown to be con- 
stant. An insect visiting the flower for the -nectar in the bot- 
tom of the corolla and not being able to escape by reason of the 
smooth corolla, except by climbing the style and over the stigma, 
and the plant being proterandous, cross-fertilization becomes 
almost inevitable. To test the matter further, especially as re- 
garded the slipperiness of the corolla tube, the following ex- 
periments were made: A number of insects of different shapes 
and species were put into the flowers and their actions carefully 
observed. The first one tried was a small cricket, which was 
potently against the corolla tube until wearied out. It was then 
left alone for an hour, when being touched with a straw it began 
again its struggles, with no better success than at first. After 
that it was turned around so as to face the style, when it easily 
climbed up and escaped. 
A spider was the next subject and proved itself just as unable 
to climb the corolla as the others preceeding. It succee ed in 
getting out at last by resting one of its long legs against the style 
and working another one over the rim of the corolla and so draw- 
ing itself to the top of the cup. : : 
I next made use of a small grasshopper, which also entirely 
failed to take any hold upon the corolla wall, and after imprison- 
ment for some minutes succeeded in making its escape by jump- 
Ing out. ‘ : 
Finally a house-fly was caught, and having had both wings 
carefully clipped with scissors was dropped into the flower. The 
result in this case was beyond expectation of those who witnessed 
