103 ME XT A L EVOLUTION IN MAN 



any other kind would have served as well as written words ; 

 for it clearly would be absurd to suppose that the dog could 

 read the letters, so as mentally to construct them into the 

 equivalent of a spoken word, in any such way as a child would 

 spell b-o-n-e, bone. But, all the same, these experiments 

 are of great interest as showing that it falls within the 

 mental capacity of the more intelligent animals to appreciate 

 the use of signs so conventional as those which constitute a 

 stage of writing above the drawing of pictures, and below the 

 employment of an alphabet. 



Enough has now been said to prove incontestably that 

 animals present what I have called the germ of the sign- 

 making faculty. As the main object of these chapters is to 

 estimate the probability of human language having arisen by 

 v»-ay of a continuous development from this germ, we may 

 next turn to take a general survey of human language in its 

 largest sense, or as comprising all the manifestations of the 

 sign-making faculty. 



Referring again to the schema (page 88), it is needless to 

 consider cases i and 2, for evidently these are on a psycho- 

 logical level in man and animals. Case 3, also, especially in the 

 direction of its branch 4, is to a large extent psychologically 

 equivalent in men and animals: so far as there is any difference 

 it depends on the higher psychical nature of man being much 

 more rich in ideas which find their natural expression in 

 gestures or tones, and which, therefore, are impossible in brutes. 

 But it will be conceded that here there is nothing to explain. 

 The fact that man has a mind more richly endowed with 

 ideas carries with it, as a matter of course, the fact that 

 their natural expression is more multiplex. 



The case, however, is different when we arrive at con- 

 ventional signs ; for these attain so enormous a development 

 in man as compared with animals, that the question whether 

 they do not really depend on some additional mental faculty, 

 distinct in kind, becomes fully admissible. 



The first thing, then, we have to notice with regard to con- 



