Metamorphosis of Skin in Amphibians. 89 



The first one (Fig. 1) shows a normally fed animal, still being 

 larval; the skin is then reddish or yellowish brown without any 

 particular patterns (N 15). The next stage is brought about by 

 the development of a greenish or yellowish network on the brown 

 background. Color and definiteness of this network depend upon 

 the kind of food and on the amount of light. Animals kept in 

 bright light and fed on thymus show the greatest definiteness of 

 this stage (Fig. 2). When the skin develops the network, the 

 animals are still larvae no matter how they have been treated, 

 though rare exceptions do occur. The next stage is the separation 

 of this network into single green or yellow spots. Fig. 3, p. 2, 

 shows this stage in an animal which was kept in daylight and fed 

 on thymus; such animals always reach this stage while they are 

 still larvae. But animals fed on worms or kept in darkness usually 

 do not work out their first spots before they have left the water. 

 In this case another characteristic which appears independently 

 from the stage of separation in animals reaching this stage while they 

 are larvae appears simultaneously with the separation stage ; namely, 

 soon after the animals leave the water, the background changes 

 its color to a dark reddish or sometimes greenish brown and the 

 skin appears leathery. All colors become more dim and faint. 

 We called this stage "Cinnamon." It is shown in Fig. 4, N 3. 

 The animal then becomes darker and darker until it is finally 

 black. The yellow spots, which have become reduced in number 

 and size lighten up and are finally bright yellow and shining. 

 But the number of the spots as well as the shade of the yellow are 

 subjected to great variations which of course cannot be discussed 

 here. 



I have here a number of formalin specimens which represent 

 different stages of the skin colorations and which may perhaps 

 illustrate these stages better than I could explain it. 



The method was then as follows: 



For each experiment three larvae of A. punctatum of about the 

 same age were used. From A one piece of skin including one eye 

 was grafted to X while the other half of the head's skin including 

 the other eye was grafted to Y. 23 pairs were operated on in this 

 way. The skin grafts in each pair were continuously compared 

 with each other as well as with their respective hosts. 



