VARIATION OF INSTINCT IN DEFINITE LINES. 223 



of confinement, of parrots, jays, jackdaws, and starlinp^s ; and 

 these facts are rendered more remarkable from the additional 

 fact that none of these birds have any proper son<^ of their 

 own, and might therefore be supposed not to have a developed 

 ear for bird-music. Still more remarkable, however, is the 

 fact that these birds are able correctly to imitate songs having 

 a proper musical notation, and that they both learn such songs 

 more readily and retain them better than even tliose singing- 

 birds which are most apt at learning tunes. For Bechstein 

 sa)^s that even the Bullfinch requires nine months of regular 

 and continued instruction to become firm in its performance, 

 and that very frequently all instruction is forgotten in moult- 

 ing. Couch, indeed, says that with all such birds " it is 

 with them as with the human race ; those which are quick 

 at attaining are also rapid in losing their acquirements," and 

 conversely; but clearly this statement is no more true of 

 birds than it is of " the human race." For of any of the 

 songless birds above named it would be a sign of unusual 

 dulness to require nine months of continuous instruction in 

 a single tune, and, on the other hand, they do not so readily 

 forget what they learn. But the most remarkable extension 

 of the power of vocal imitation whicli these birds display is 

 unquestionably that of uttering articulate words. This 

 subject will require to be considered more fully in my next 

 work. Meanwiiile it is enough merely to mention it with 

 reference to the w^onderful power and precision of the imita- 

 tion which is betokened in thus modifying the instinct of 

 uttering a caw or scream, into the singing of a definite tune 

 or the speaking of articulate words. 



The habit displayed by cats, and even young kittens, of 

 washing their faces might well be deemed instinctive, and so, 

 most probably, it is ; but that it may also be acquired by 

 imitation is proved by the fact that puppies when brought up 

 by a cat perform the same movements. This was first 

 observed by Audouin,* and has since been independently 

 corroborated by several observers, of whom I may mention 

 the following : — 



Dureau de la Malle gives the case of a terrier wliich 

 belonged to himself, and which from the time of its birth 

 was brought up with a kitten six montlis its senior. For two 

 years the terrier had no association with other dogs. Soon 



* Anns, des Sc. Nat., torn, xxii, p. 397. 



