OF ARTS AND SCIENCES. 63 



in dictionaries ; naniely, that there is a small number of characters, 

 reckoned at 214, of very frequent occurrence, and that either alone, 

 or as a component part of a larger character, some one of these fre- 

 quent signs occurs in every sign-word of the language. Hence they 

 have arranged their sign-words under these frequent signs as heads of 

 groups, and denominate these last the keys of the language. 



" Mr. Marshman observed that so much as remains of a compound 

 sign-word, after the key is removed, is likewise a substantive char- 

 acter or sign-word of the language, occurring both by itself and in 

 combination with different keys, so as to furnish another and distinct 

 mode of grouping or classifying the characters. This remaining part 

 of the character after removing the key was called by Mr. Marshman 

 the primitive. The key is then the modifier. (It is also badly de- 

 nominated the radical. ) Mr. Marshman supposed that the primitive 

 represents the meaning of the whole character in a general way, and 

 that the modifier then renders it definite, much in the same way as 

 the primitive or root word of a Latin or Greek compound verb is 

 modified by the several prepositions prefixed to it, and he adduced 

 a moderate amount of examples to sustain this theory. His observ- 

 ance of related meanings extended only to those few obvious ones 

 which appear at a casual glance, and offered no clew to an integral 

 development of the scheme. No successor of Mr. Marshman has 

 therefore had more success than himself in demonstrating his theory, 

 and M. Gallery comes forward to throw discredit upon it altogether, 

 by asserting one quite different from it, and, as he evidently thinks, 

 incompatible with it. 



" It has just been shown that the compound sign-words consist 

 each of two parts, one of which is called the -primitive., and the other 

 the modifier. The modifiers are not so numerous as the syUahic 

 words of the spoken language, while the primitives are much more 

 so, being by M. Gallery's computation 1040. It has been observed 

 by the Ghinese themselves, that, as the general rule, all the sign-words 

 which have the same primitive are homophonous, or, in other words, 

 signify the same syllaMc word, while those having the same modifier 

 have no such established relationship of sound, but generally differ 

 from each other throughout. This fact M. Gallery has brought out 

 into a much clearer light, and has made it the basis of his arrange- 

 ment of the sign-words of the language. He advances and contends 

 for the theory, that the primitive as previously called, which is usually 



