THE SCIENCE OF LANGUAGE. 43 



from a comparatively early time, may 

 be otherwise supposed in respect of other 

 families of languages, which do not possess 

 those exponents of their earlier forms. We 

 therefore know positively from the obser- 

 vation of collected facts that languages 

 change as long as they live, and for 

 this knowledge we are indebted to the art 

 of writing. 



But for the invention of the art of writing 

 the student of language would never have 

 imagined, up to the present day, that such 

 languages as Eussian, German, and French, 

 for example, are descended, after all, from 

 one and the same stock. Nay it is quite 

 possible that nobody would ever have 

 hit upon the idea of a common origin 

 for any languages whatsoever, however 

 closely related, or ever would have sup- 

 posed that a language is subject to any 

 change at all. Without wa-itten records 



