The statement that like poles repel and unlike poles attract is, I believe, a conclusion that goes beyond the evidence. The experiments show that like poles do often unite in plants, and this has been abundantly shown to be the case in the lower animals, and even in forms as high as the earthworm and the tadpole. Even if when like poles are united subsequent changes take place, that in some cases, although apparently not in animals, lead to the death of the graft, it by no means follows that this has anything to do with the attraction or repulsion of the parts, but rather with some difficulty in obtaining food, or with the transportation of substances through the plant. In the lower animals we have seen that when like poles are united there is sometimes a stronger tendency to produce new organs at or near the place of union than when unlike poles are united, but it would be going too far, I think, to state that this is due to repulsion of the parts, especially in the sense in which the like poles of a magnet repel each other. It seems to be due rather to the two parts failing to unite into a whole organization, each retaining the same structural basis that it had before grafting, but this is a very different principle from that of an attraction and repulsion of the parts, and the question of the union of the parts appears also to be a different question from that of the organization of the parts themselves. In the mammals, and in general in all forms in which there is a dependence of the parts on each other, it is impossible to carry out grafting-experiments on the same scale as those described in the preceding pages. The principal difficulties are to make the parts unite, and to supply nourishment and oxygen to the graft. Owing to the dependence of the parts of the body on each other for a constant supply of oxygen and food derived from the blood, as well as for the removal of the waste products, the parts cannot remain alive, or even in good condition, while new connections are being established. For this reason, as well as for others, it would not be possible, for instance, to graft the arm of a man upon another man. The tissue may have the power of uniting even in this case, as is seen when the bone is broken and subsequently reunited, but the difficulty would be in supplying the grafted arm with nourishment, etc., during the long time required for the union to take place. Smaller parts of the body may be successfully grafted, and there are several recorded cases in which parts of a finger, or of the nose, are said to have been cut off and to have reunited after being quickly put back in place. Pieces of human skin may be grafted without great difficulty upon an exposed surface, and it has been said that small pieces succeed better than larger ones, owing, most probably, to their being able to absorb sufficient oxygen, etc., and keep alive until new blood vessels have grown into the grafted piece.