BEHAVIOR AND COLOR CHANGES OF TREE FROGS 63 



temperature, we shall have the maximum expressions of the 

 appropriate reaction. To quote directly Parker's statement, 

 " Light causes a distal migration of the pigment granules of 

 these cells; its absence a proximal one. A high temperature 

 causes a proximal migration of pigment; a low temperature a 

 distal one." He believes there are no real exceptions to these 

 rules. I have not experimented with lizards, but from what 

 has already been presented it will be seen that there are among 

 tree-frogs very striking and positive exceptions, and accounts 

 now to be cited will make this quite certain. 



In my temperature experiments two chief methods were 

 employed. First, and I think the more satisfactory, was the 

 simple transfer of the cage from the room temperature to that 

 of the outside air, which was easily done by raising the window 

 and quietly lifting the jar from the table and placing it upon the 

 outside window shelf; making accurate records of the tem- 

 perature in each case. This feature could also be supplemented 

 by utilizing the presence of a temperature oven which was 

 within easy reach of table or window. 



The second method, and this proved preferable at Naples, 

 consisted in placing specimens in temperature baths, i. e., by 

 means of ice or heat it was quite possible to have a vessel of 

 water of any desired temperature, into which a specimen could 

 be put and left the desired time without in any serious manner 

 injuring it or interfering with its behavior. 



Experiment. A jar containing two specimens was placed on 

 the window shelf where the thermometer registered 14 C. At 

 the end of one and one-half hours both specimens had changed 

 from a dark brownish to very light grey. They were then placed 

 on the inner window shelf where temperature was 21 C. and at 

 the end of half an hour both had resumed the original color. 

 In the room on the same table a control jar with three specimens 

 had remained without change during the same time. The ex- 

 periment was now reversed, that is, the jar was placed over a 

 radiator shelf where the temperature gradually rose to 35 C. 

 Bath specimens showed signs of discomfort, moving about, 

 shifting the body variously, scratching the nose with fore feet, 

 etc., and in the meantime both had become markedly lighter 

 in color, though not equally so. After a half hour the jar was 

 transferred to the outer window shelf where in a half hour the 



