to the stigma of another, the result of such cross-pollination being greater vigor in the offspring. Thus Prof. Halsted[1] found that barberry flowers from which insects had been carefully excluded produced no fruit (others uncovered on neighboring branches fruiting abundantly), and this in spite of the fact that through jarring or as a result of age the stamens had curved inward as far as they ever could. Microscopical examination showed a considerable quantity of pollen to have been deposited among the viscid hairs which form a ring about the top of the pistil (see Fig. 3, H), but none whatever upon the cushion-like summit which was found to be the only part that served as stigma.
Barberry blossoms are great favorites among the insects. Few of our June flowers gather about them a larger number of bees, hornets, flies, butterflies, and beetles. The smaller bees and certain flies are especially abundant.
There is some reason to believe that the intense color of the glands may serve as a guide to the insect, directing it at once without loss of time to the nectar which collects in little hollows between the bases of the filaments and the glands, where it is held by capillary attraction. An insect in thrusting its proboscis into a nectar cavity must touch the base of two filaments, whereupon both stamens suddenly bend inward and strike the insect's head. Now, Müller[2] calls attention to the fact that while large insects such as bumblebees pay no attention to this, but continue to make the circuit of the flower, smaller ones, like the hive bee, appear to be somewhat startled by this performance and fly away at once to another barberry flower. But the insect carries with it some pollen upon one side of its head, and if in the next flower this comes into the same relative position as before, more pollen will be added on the same part; but if, on the other hand, the flower is approached from the other side, then the pollen already collected will be deposited upon the stigma, while at the same time a new supply of pollen is being received which may in turn be carried to still another flower. As the smaller insects are the more common visitors, cross-pollination, which is so much the best for offspring, must therefore be the most usual result.
This sensitiveness of the stamens is exhibited by all the species of Berberis so far as known, but is not found in other members of the family, although a somewhat similar irritability of stamens has been observed in certain of the Portulacaceæ, Tiliaceæ, Cistaceæ, and Compositæ. The strikingly animal-like nature of the movement is well shown by the following facts: A chemical stimulus, such as ammonia gas, will induce contraction as effectually as a mechanical stimulus. The presence of oxygen
