In a dense forest of growing trees the smaller, side branches of the main stem, as well as those of the larger branches, are continually dying off. But the tree rids itself of these dead branches by forming a collar of tissue from the cambium layer around the base of the branch, which presses more tightly as layer after layer of living wood is added, until the branch finally falls off and the hole which is left is grown over in a short time. This process is known as natural pruning. But the process which we wish to consider is very different from this, and we desire to distinguish it by the term, self-pruning. In this case the living branches are cut off or else the cutting-off process is the cause of the death of the branch. A special adaptation is provided to accomplish the result and the process is one whose purpose is the shedding of the branches rather than the attempt to accommodate the plant to conditions of injury brought about by other causes. In a number of species perfectly formed winter buds were developed on the branches which were shed, and so far as our observations go, the twigs are cast in the fall and winter. Although the shedding of branches is well known, especially in the conifers, not as much notice has been taken of it as we think it deserves. We have been taking observations for several years and have been partly anticipated by Dr. Bessey in a note in _Science_ 12. 650, 1900,--Botanical Notes--The Annual Shedding of Cottonwood Twigs. Bessey describes the shedding of the twigs of Populus deltoides as occurring about the middle of October, and after giving the details of the process, concludes as follows: “It is an interesting fact that the Tamarisks (Tamarix sp.) which are held by some botanists to be closely related to the Poplars, shed their twigs by exactly the same device as that described above. In the Tamarisks the shedding of the twigs is a part of the annual process of defoliation, their leaves being so small that it appears to be less trouble and expense to drop twig and all than to separate every individual leaf. Possibly in the Cottonwoods, with their large leaves, we have a survival of the Tamarisk twig-shedding habit long after its original significance has disappeared.” Dr. Bessey, however, we believe, will not insist on this supposition when he considers that the same thing occurs in species of Prunus, Quercus, and other widely separated genera. [Illustration: FIG. 1--Twig of Populus alba, showing large basal joint and scars where smaller twigs have been detached. FIG. 2--View of basal joint of same twig as Fig. 1. FIG. 3--Twig of Salix nigra, showing position of the brittle zone (_a_). FIG. 4--Green twig of Ampelopsis cordata, showing joints (_a_) at the leaf nodes.]