[Illustration: FIG. 54.--_A._ After Harrison. Union of two tadpoles by posterior ends. Two days after operation. The line to the left of plane of union indicates where the two were cut apart. _B._ Tail of right-hand tadpole in _A_. Five days after cutting apart. _C._ Same. Nine days after cutting apart. _D._ Same. Ninety-five days after cutting apart. _E._ After Born. Combination of _Rana esculenta_ (anterior) and _Rana arvalis_ (posterior). Thirteen days after the operation.] species, even when they belonged to different genera. It is found, however, that some of these combinations can be more easily made than others, but it is not clear whether the difference depends upon differences in the sizes of the pieces, or the rate of growth of the ectoderm over the cut-surfaces, or to a deeper-lying lack of affinity between the tissues. A combination of _Rana esculenta_ (anterior) with _Bombinator igneus_ (posterior) was made. The combination lived for ten days, and then showing pathological changes, it was killed. Another combination is shown in Fig. 54, _E_, in which the anterior part of _Rana esculenta_ was united to the posterior part of _Rana arvalis_.[88] The blood of the posterior component was driven through the vessels by the action of the heart of the anterior component. The animal lived for seventeen days. In all these combinations between different species, each developing part retains its specific characters, and, although in several cases one part received its nourishment from the other through the common circulation, yet no influence of one component on the other could be observed. Harrison has succeeded in keeping an individual made up of two species, _Rana virescens_ and _Rana palustris_, for a much longer time,--until, in fact, the transformation of a tadpole into a frog had taken place. Each half retained the characteristic features of the species to which it belongs. The absence of regeneration after the union of the pieces may be attributed, in several cases, to the absence of this power in the region through which the cut has been made; but in other experiments this cannot be the explanation, since the power to regenerate can be shown to exist in the part. This is the case in an experiment carried out by Harrison and repeated later by myself. If the tips of the tail of two tadpoles are cut off and interchanged (Fig. 55, _A_, _B_), a perfect union takes place between the two parts, and a single tail develops. Each of the cut-surfaces has the power to regenerate, but the union of the parts has suppressed the regeneration. If, however, like parts are not brought in contact, regeneration may take place in the region of union (Fig. 55, _D_).