1 34 NATURE STUDIES. 



taught to do, at the risk of offending his master or 

 mistress by such refusal, makes it absolutely certain 

 that he had clearly recognised the object which was 

 to be attained by ringing the bell. 



It may be objected that in cases such, as those we 

 considered last, the animal has merely imitated an 

 action which it has seen performed by others, and has 

 subsequently -learned to associate the action with its 

 ordinary consequence. Apart from the consideration, 

 however, that although in any single case such an 

 interpretation might possibly be correct, it would 

 be most improbable that it should explain all cases in 

 which cats or dogs have used knockers or rung bells 

 in the usual way : cases may be cited in which animals 

 have devised a way of their own for producing such 

 signals. Thus Mr. E. L. La-yard, of the British con- 

 sulate, Noumea, relates the following case in which a 

 cat acted in a way which can hardly be explained, 

 save by assuming that she reasoned : " Many years 

 ngo," he says, " we lived in Cambridge, in Emmanuel 

 House, at the back of Emmanuel College. The pre- 

 mises were partly cut off from the road by a high 

 wall; the body of the house stood back some little 

 distance. A high trellis, dividing off the garden, ran 

 from the entrance door to the wall, in which was 

 another door, or grate. A portion of the house, a 



gable, faced the trellis We were, after some 



time of residence, extremely troubled by runaway 

 rings, generally most prevalent at night, and in rainy, 

 bad, or cold weather, which was a great annoyance to 



