REASON AND LANGUAGE. 147 



to primitive articulation ? When once any one has a 

 meaning to convey, he must, if he can succeed in con- 

 veying it, convey it by some visible, audible, or tactile 

 sign. The employment of any one must be due to an 

 internal impulse, and the employment also of any one 

 kind of sign is fundamentally as wonderful as are either 

 of the others. If existent dumb sign-making is due to 

 ancestral speech, and ancient speech due to still more 

 ancient gesture — as Mr. Romanes represents — to what 

 was the original gesture due ? 



As we have already pointed out,* the nervous ana- 

 tomical conditions which favoured and were further 

 developed by one kind of expression, could never have 

 favoured the other. 



We are quite sure that Mr. Romanes is entirely sincere 

 and honest, and does not see the equivocal nature of his 

 argument. Nevertheless, to represent that the origin of 

 each kind of language was developed from the other, 

 and to withdraw whichever conception of origin an 

 inquirer may seem disposed to select, is practically to 

 shuffle with ideas in a way which reminds us not a little 

 of the well-known " three-card trick." To this question 

 we shall, however, be compelled to revert f when we come 

 to examine Mr. Romanes's eighth chapter — that on " the 

 relation of tone and gesture to words." 



Our author candidly makes the noteworthy admis- 

 sion X that it would " be wrong to say that a higher 

 faculty is required to learn the arbitrary association 

 between a particular verbal sound and a particular act 



* See above, p. 141. t See below, pp. 163. 



X p. 123. 



