SELECTION 163 



what the fittest is, but we can discover by 

 observation and experiment whether or no 

 protective coloration has selective value. A 

 case in point is given by Mr. A. P. di Cesnola 

 in a short but highly interesting paper in 

 "Biometrika" for 1904. 



"It is well known that the * praying 

 Mantis,' Mantis religiosa, occurs in Italy in 

 a green and a brown form. The former is 

 usually to be found on green grass, the latter 

 on herbage browned by the sun. Mr. Cesnola 

 tied down among green herbage twenty green 

 Mantis, and among withered grass a similar 

 number of brown individuals. After seven- 

 teen days they were all alive. He also 

 tethered twenty-five green Mantis among 

 brown herbage, and they were all dead after 

 eleven days. The converse experiment was 

 also made, forty-five brown Mantis being 

 exposed on green grass, and of these only ten 

 survived at the end of seventeen days. Most 

 of the Mantis were killed by birds; five of the 

 green ones were killed by ants. Here, then, is 

 a proof, quite conclusive though the numbers 

 are small, of the selective value of the pro- 

 tective coloration of both races of Mantis. 

 If green Mantis and brown Mantis be exposed 

 on green grass, the green ones will survive 

 rather than the brown, the death-rate will be 

 F 2 



