498 R. T. YOUNG 



lieve played a relatively small part in determining the results. 

 That it may have been of some importance however is indicated 

 in Experiment 6, Series IX. 



It may be argued from these results that, since the birds find 

 their prey chiefly through the movement of the latter, it is of 

 no consequence in the struggle for existence whether or not an 

 animal resembles its surroundings, provided only that it re- 

 mains quiet or hidden when exposed to attack. Unfortunately 

 we have as yet no means of knowing the precise role which move- 

 ment plays in animal fatalities. It is obvious even without ex- 

 perimental proof that a moving animal is more liable to attack 

 than a quiet one, but as to what extent moving animals are killed 

 by their enemies we have no definite data. Granting however 

 which is probably true, that moving animals are more frequently 

 killed than motionless ones; it is possible that protective resem- 

 blance would have sufficient selective value to become permanent. 

 This of course leads up to the great question as to the efficiency 

 of selection in fixing minor variations upon a race of organisms; 

 a question with which the present paper has nothing to do, the 

 purpose of the experiments being, not to determine this latter 

 point, but rather to ascertain whether or not protective resem- 

 blance in the case of motionless animals is really an efficient 

 means of protection to them. 



This latter question has I believe been answered in the affirma- 

 tive by these experiments. I have endeavored in them to put to 

 an experimental test a hitherto practically untested hypothesis. 

 They suggest further lines of desirable observation and experi- 

 ment as follows: 



1. Will close protective resemblances, such as those shown 

 for example by many of the underwing moths, deceive animals 

 of deliberate approach, such as crows, grackles, etc.? 



2. Is protective resemblance efficient in the case of moving 

 animals? 



3. Is the sight of wild birds as keen as is frequently assumed? 



4. Is protective resemblance as efficient with wild as with 

 captive animals? 



