EXPERIMENTAL PHONETICS. 250 



ancient Sanskrit, or how Demosthenes spoke in Greek or Cicero in 

 Latin? Would it not now be interesting^- to hear the exact iiccent of 

 old English, or the Scotch of the tifteenth century? All dialects 

 should be carefully registered and put aside for future consultation, 

 and thus we would do for the ear what we do for the eye. No dou])t 

 such a collection of phonographic records would help onward the 

 science of language. 



