SKIN TRANSPLANTATION IN FROG TADPOLES 399 



rounding skin, but that a 'homoeotransplant' of the same kind 

 loses its pigment. Seehg ('13) performed the reverse operation 

 and reported that both autotransplants and 'homoeotransplants' 

 of white skin on a black region became pigmented. No explana- 

 tion of the process is given. Recently Dawson ('20) made grafts 

 of this sort on Necturus. Pieces of pigmented skin from the tail 

 and of white skin from the venter were interchanged. Although 

 it is not definitely stated, it is believed that all of his grafts were 

 autoplastic. He found that the white grafts became partly pig- 

 mented. The center of each graft remained clear. The single 

 graft of black skin on the venter began to lose its pigment ten 

 weeks after the operation. At the end of his observations a 

 large irregular central area was free from melanophores. This is 

 contrary to Sale's grafts, which preserved their pigment and ex- 

 tended it into the surrounding skin. Only in homoiotransplants 

 was the pigment lost. 



All of these experiments agree in showing that unpigmented 

 skin grafted into a pigmented region of the same animal becomes 

 pigmented. Dawson is the only one to observe that the pig- 

 mentation is chiefly epidermal, as it is in frog tadpoles. The 

 two explanations that have been offered to account for the proc- 

 ess are, first, that the epidermal melanophores have migrated 

 (either actively, or included in the migrating epidermis as a whole) 

 from the host skin into the graft. Active migration of the 

 melanophores independently of other epidermal cells has not 

 been observed in my transplants, and that method will not be 

 considered here. Secondly, the melanophores have arisen in situ 

 from epidermal cells. Observations of the surface of homoio- 

 transplants in frog tadpoles leads to the belief that the melano- 

 phores are carried along by the migrating epidermis, over the graft. 

 The parallel arrangement of the melanophores and the gradual 

 centripetal spreading seem to support this belief. The possible 

 methods of such epidermal migration are four: 1) the graft may 

 be 'overgrown' by a thin epidermal layer; 2) the epidermis of the 

 graft may be 'undergrown' by the invading layer of epidermal 

 cells; 3) the epidermis of the graft may be gradually replaced by 

 an advancing layer of host epidermis; 4) the migrating host 



THE JOURNAL OF EXPERIMENTAL ZOOLOGY, VOL. 35, NO. 4 



