THE WAYS OF LIFE 245 



results were obtained when the food was put in a cylinder 

 of another colour. 



Bouvier was able to prove that wasps of the genus Bembex 

 associated a certain stone, for instance, with the way to 

 their burrow. It has been shown that the American 

 crawfish, the crab, and the hermit-crab can be taught to 

 take the more advantageous or the easier of two alternative 

 paths. Anna Drzewina gave hermit-crabs which had been 

 deprived of their shells a number of top-shells [Trochus) 

 with the openings closed. The hermit-crabs spent futile 

 days and nights tryiag to use the closed shells, but after 

 six to eight days gave it up. Even when a shell with a 

 paper lid was given them, they would not so much as try. 

 They associated the form of the shell with failure. But 

 when other closed shells of a difEerent shape were given to 

 them, they began eagerly again their futile attempts to 

 win a way in. 



Trial and Error. — In illustration of another experi- 

 mental method, we may refer to Professor Thorndike's 

 iavestigation of the learning powers of cats and dogs. 

 He contrived cages with doors which could be opened by 

 the manipulation of more or less intricate combinations of 

 bolts and levers. Hungry cats and dogs were shut ra 

 and were tempted, by food placed just outside, to solve 

 the problem of their prison-doors. In similar circumstances, 

 we should probably do a little thinking, make one trial, 

 and be free. But this was not what the cats and dogs did. 

 They got out by the ' trial and error ' method ; that is to 

 say, they made one experiment after another until they 

 hit upon the fit and proper way of working the mechanism. 



The experiment was repeated over and over again, and 

 the curves recording the times taken to escape showed a 



