400 WILLIAM H. COLE 



epidermis may intermingle with the graft epidermis. In each 

 method the melanophores would be carried along by the migrat- 

 ing host cells. From the surface appearance and from the fact 

 that the new pigmented layer may be peeled off, it seemed likely 

 that the overgrowth was the method realized in the case of some 

 grafts. However, an examination of sections, both of early and 

 late stages of the process, failed to give conclusive proof. In the 

 case of the twenty-four-hour process, there were indications in 

 some sections of two distinct layers of epidermis, only the outer 

 containing melanophores. In other sections, no such condition 

 could be found, although the epidermis was slightly thicker than 

 normally. In still other cases the epidermis of the graft over a 

 region about 2 mm. from the line of union was considerably 

 thicker than it was in the central region, indicating that the 

 advancing epidermis had intermingled with the graft epidermis. 

 No evidence in favor of the second and third possibilities was 

 found. The only safe conclusion that could be drawn from the 

 sectioned material was that migration from the host epidermis 

 into that of the graft had taken place. Which of the four methods 

 prevailed remains undetermined, although overgrowth seems 

 most likely. 



Evidence of the opposite kind came from the grafts LB 35, 

 LB 36, and others like them. Their history did not favor the 

 migration theory at all. The presence of so many pigment gran- 

 ules and irregular pigment masses pointed to a formation in situ. 

 There was no indication of overgrowth or intermingling in any 

 of the sections. It will be remembered that these animals (LB 

 35 and 36) were undoubtedly close relatives. Therefore, the dif- 

 ference in their specific protoplasms was not very great, and the 

 reaction caused by the grafting was weak. Equilibrium was 

 quickly established, and the slow appearance of pigmentation 

 was probably the result of formation in situ. Assuming that all 

 epidermal cells are potential melanophores and that the sur- 

 rounding skin exerts a stimulus upon the graft, such formation 

 may be explained. In the twenty-four-hour process (LB 32) 

 relationship between the host and the graft was remote, causing 

 a violent reaction. Actual union was established only b}^ the 



