EXPERIMENTS ON PROTECTIVE COLORATION 465 



evaluating the results. In both cases, however, the lateral color 

 is practically the same as the dorsal, while in that of Xyphidium 

 the light straw colored wings folded over the back tended on 

 both backgrounds to enhance the protective effect; for on the 

 straw background they closely matched the color of the straw, 

 and on the grass they gave the appearance of blades of dry grass 

 lying among the green. 



Experiment I. The insects were placed on the backgrounds at 7.26 

 a.m., and the box opened. Previous to 7.45 a.m. the dead insects 

 were not eaten, although the birds passed them several times in hop- 

 ping about the floor of the cage, and one live and active specimen, 

 which had been left in the cage by mistake, was taken from the grass. 

 From 7.45 to 8.30 the birds were not observed. During this time two 

 Melanoplus were taken from the straw and one from the grass. Be- 

 tween 8.30 and 8.45 the three Melanoplus remaining on the straw 

 and two of the four on grass, and one of five Xyphidium^ on straw 

 were taken. Twice during this period a bird passed by the grass with- 

 out feeding. Summed up the result of this experiment is : 8 Melano- 

 plus (5 from straw, 3 from grass), and 1 Xyphidium from straw (prob- 

 ably) eaten. 



Some uncertainty was introduced in this experiment by the presence 

 of active specimens in the cage, which, dropping upon the backgrounds 

 among the dead insects, attracted the attention of the birds. 



ExTperiment 2. In order to ascertain which species, if either, the 

 martins preferred, I placed five of each on a white cloth where both 

 kinds were plainly visible. If any difference in contrast existed, that 

 between Melanoplus and the white cloth was the greater. One bird 

 took five Xyphidium and one Melanoplus and would probably have 

 eaten the remaining Melanoplus if I had not interrupted it at this 

 point. They were eaten shortly after. This shows that for this indi- 

 vidual at this time at least there was no preference for Melanoplus 

 over Xyphidium. 



Experiment S. In this experiment one bird passed by the grass and 

 picked a Melanoplus from the straw. Then a bird^'' flew over both 

 the grass and the straw and picked two Melanoplus from the latter. 

 I now left three of each species on the straw and five of each on the' 



^ Some uncertainty exists regarding this. At 8.45 there were five Xyphidium 

 remaining on the grass. From its position, it is possible, though not probable, 

 that one of these had been thrown from the straw to the grass by the movements 

 of the birds, in which case the record would be one Xyphidium, from the grass 

 and none from straw taken. It was of course impossible to watch the birds 

 from a distance and keep individual specimens of the grasshoppers in view. 



1'' The identity of each individual of the two birds I could not distinguish. 



