Experiments in Grafting 293 



When two cut aboral ends are united (Fig. 9), two new foot 

 processes are formed (Fig. 10), and the two pieces subsequently 

 pinch apart to form each a separate individual. The explana- 

 tion is the same as in the last case. 



These results can be somewhat better understood when taken 

 in connection with the experiments of hydras partially split length- 

 wise. If the anterior end is split for only a short distance, each 

 half will round up and produce a separate head (Fig. n). 

 Subsequently the two heads may slowly fuse into a single one. 

 The result is similar to the fusion of two bodies in the grafted 

 hydra with one head. If the split extends farther into the 

 hydra, as in Fig. 12, the two parts slowly draw apart until 

 they finally pinch off at the foot region and produce two 

 hydras. Why in one case the two separated parts unite and 

 in the other case separate farther can possibly be ex- 

 plained on the view that I have tentatively suggested above. 

 It will be noted that in the case of slightly separated heads 

 the influence of the single body prevails, and reunites the 

 parts from behind forward by drawing them together and es- 

 tablishing the normal tension relations ; and in the case of the 

 more separated anterior ends, these ends prevail and cause 

 the single trunk to pinch farther apart, causing, as it were, the 

 halves of the single trunk to draw up into their respective 

 anterior ends. 



The results of uniting the squarely cut posterior end of one 

 individual with a cut surface in the side of another individual 

 (Fig. 13) are similar in many ways to the last. The united 

 pieces usually adjust themselves in such a way as to share the 

 common trunk (Fig. 14). Subsequently the two pieces split 

 apart farther and separate in the region of the foot. If the 

 grafted piece does not succeed in halving the trunk, i.e. if it 

 does not turn anteriorly, a foot develops at the line of union 

 (Fig. 15), and the graft pinches off without passing down to the 

 base of the stock. What conditions lead to this difference in 

 behavior have not been sufficiently made out, but it is probable 

 that the kind of union, or the relative sizes of the graft and 



