The nectar glands of these hypothetical outer stamens, although situated behind the others, would not be disqualified in the least by such a position from continuing to perform their function. On the contrary, it might well happen that, as they no longer produced pollen, they would secrete all the more nectar in consequence, and thus relieve the inner stamens of so much of their work. No longer required, the glandular appendages of the latter would be reduced to rudiments (as in mahonias), or entirely disappear, as we have seen in the case of the true barberries.
Do the above considerations help us to any better understanding of how the irritability of the stamens came to be developed? In our previous consideration of this remarkable property, we saw reason to believe that such a peculiar manifestation of protoplasmic activity could only be satisfactorily explained as having resulted from a rare combination of favoring circumstances. Although, in the discussion of such a matter, we are confessedly treading upon uncertain ground, still, it may be worth while to inquire whether, supjoosing the barberry flower to have been evolved essentially after the manner indicated, there have not been thus happily combined the very factors we should deem necessary and adequate to produce this result. It should be remembered that we are not endeavoring to account for that fundamental property of protoplasm known as contractility, but only for its being in the stamens of the barberry so much more strikingly exhibited than in other organs of the plant, and in the great majority of other plants.
In the first place, in order that this or any other property of protoplasm should be especially well shown in any organ, it would seem to be a prerequisite that the organ should be unusually rich in protoplasm—a supposition which is confirmed by the comparative study of motile organs. Such a very considerable reduction of parts as we believe to have taken place in the barberry flower might well be connected with the enrichment of the remaining tissues.
Secondly, a mechanical stimulus applied repeatedly for innumerable generations, at a very definite part of the stamen, would seem to be also necessary in order to account for the fact that movement of the organ occurs in response to a touch only when applied to the front of the filament and near its base. From the position which the glands came to occupy in the flower, just such a stimulus was afforded by the proboscis of every insect that sipped the nectar.
Whether we are at liberty to suppose that the direct effects of such a repetition of stimuli may be accumulated through inheritance, or whether we must assume only the inheritance of fortuitous variations, is of comparatively small consequence in this
