EXPERIMENTS ON PROTECTIVE COLORATION 467 



one Melaiiopliis from the grass. After removal of straw four Melano- 

 plus and no Xyphidiiim were taken from the grass. 



A comparison of the results of this experiment with those of Experi- 

 ments 1 and 3 shows a marked difference, due unquestionably to the 

 change in position of the insects and the consequent change in color con- 

 trast. In Experiments 1 and 3 the Melanoplus — straw combination was 

 the one mostly chosen by the birds, being that one in which the color 

 contrast was greatest ; while in 4 the Melanoplus — straw and Xyphid- 

 ium — straw combinations were about equally selected, the color con- 

 trast being very little different in either case. It is probable that the 

 Melanoplus — grass combination would have been oftener selected, 

 had the birds more frequently happened to alight on the side of the 

 cage nearest to the grass. 



Experiment 5.^* In this and the following experiment I employed 

 backgrounds of mud and sand instead of grass and straw, placing at 

 the outset of the experiment five Melanoplus on each background, those 

 on the sand being (to my eye) the more conspicuous. In arranging 

 these, I placed those on the mud background last, so as to attract, 

 if possible, the birds' attention to these, rather than to those on the 

 sand. A bird passed first to the sand background and took three in- 

 sects. During the taking of these three the bird left the sand once and 

 passed by the mud ; and at another time it inspected the mud closely, 

 but apparently without seeing the insects on it, as it did not take them 

 at that time. Observation was then discontinued for a time and when 

 renewed it was found that all of the remaining insects had been eaten. 



Experiment 6. In this experiment five Xyphidium were placed 

 on mud and five on sand. One bird flew to the mud past the sand, 

 but without taking any insects from it. It then returned to the sand 

 and ate all five placed on it. In this experiment the comparative re- 

 semblance and difference between the insects and their background 

 was not as marked (to my eye) as in Experiment 5, but was still suffi- 

 cient to influence decidedly the results. 



Experiment 7. In this experiment five of each species of insects 

 were placed on grass and five of each on straw. The birds passed the 

 grass (one walking directly over it), and then flew up to their perch. 

 Returning to the floor of the cage, they took two Melanoplus from 

 the straw. Three of each species were now left on straw and five of 

 each on grass. A bird came from the side of the cage nearest the grass 

 and took three Melanoplus therefrom. In this latter case, as in Experi- 

 ment 4, the element of chance evidently decidedly influenced the re- 

 sults, the bird feeding from that background to which it first came. It 

 is to be noted however that in this last case that species, Melanoplus, 

 was taken which presented the greater contrast to the background 

 (grass) . 



^* In this experiment but one bird was employed, instead of two as heretofore. 



