BEHAVIOR AND COLOR CHANGES OF TREE FROGS • 65 



an earlier part of this section that these frogs fail to sustain 

 the theoretical rules. 



Other experiments. In connection with the preceding obser- 

 vations and tests it was noticed in several instances that the 

 mere handling of specimens often had the effect of involving 

 color changes quite as marked as any of those already described. 

 For example, in several instances it was noted that in carrying 

 a cage of specimens from one room to another a perfectly evi- 

 dent color change was induced. To test this careful attention 

 was given to color patterns of specimens in perfect quiet by 

 two or three observers. Then a cage would be taken up and 

 turned about, lifted up and then returned, etc., etc. Then 

 another careful scrutiny would be made; and it soon became 

 evident that almost any such disturbance might act as a stim- 

 ulus. I even tried the effect of merely close scrutiny without 

 any jostling or moving of the jars, and found that where a 

 specimen could be brought to observe that it was observed, 

 there was often a color change quite as evident as the others. 



The following experiment will show that a mere stimulus 

 is not of itself always sufficient to arouse a color reaction. 

 I introduced the electrodes of a battery into a cage in such 

 manner as to enable me to apply them to a specimen almost 

 at pleasure. I was able thus to compel the specimen to leap 

 vigorously about the cage for some time. But this was ap- 

 parently without direct effect on coloration. This was the more 

 surprising as other experiments of mechanical irritation with 

 needle or forceps had seemed to act to induce a darkening of the 

 skin. But similar experiments on H. arbor ea were frequently 

 negative, hence it may be regarded as doubtful whether such 

 stimuli serve to influence to any appreciable degree the color- 

 ation. 



This does not, it seems to me, discredit the suggestion as 

 to the fact of a nervous, or emotional factor as involved in 

 these phases of behavior. So far to the contrary I am dis- 

 posed to believe that we have here one of the important ele- 

 ments, possibly the most directly important, involved in this 

 class of behavior. With this in mind one may seriously question 

 whether the marked color changes almost always associated 

 with the capture and carrying of specimens to the laboratory 

 be not an expression of the emotional condition of the animal, 



