the Structure of Language. 175 



the case with i-egard to the primary biliterals 5)D (CP), and 

 -)D (CR). 



Thus, from t]^, we not only get the biliterals t)p, t\^, PjH, yp, 

 a:i, nn but also the biliterals 5)D, ID^ ^V, P)^j> ^^j» (^I^^ ^^») 

 ^2i, (nii), each of these being primarily dependent for its im- 

 port on the primary biliteral P]D, to which the significant D im- 

 parts its ideal characters I. and II. 



So "ID (CR), not only communicates its import to its deriva- 

 tive in (and, in several instances, to "1p and to ")j), but it 

 imparts it also to the biliterals ID, *1D^ 11, l'^, and, in many 

 instances, to "1t_, It:^, and "1J1. 



Thus a second set of secondary biliterals is derived from the 

 significant D, and from each of these (as from the other bilite- 

 rals, both primary or secondary) are formed various triliterals, 

 which are constructed in the manner already pointed out, 

 namely, by adding a formative letter, either as an affix, or as an 

 epenthetic, or as a prefix. 



Jn these several ways, then, are triliterals, (both primary and 

 secondary) constructed from biliterals, which biliterals are either 

 primary or secondary (there being, as we have seen, both a first 

 and a second class of secondary biliterals in some instances) ; 

 and thus, from about eight or nine primary biliterals, which 

 owe all their import to one single significant (the bow-letter D), 

 may be derived and constructed above five hundred of those 

 words which are set down in the dictionaries and vocabularies 

 of the Hebrew language as arbitrary roots. So that from 

 this apparently simple symbol, from the extension of the 

 three general ideas — incurvation, restriction, smiting or fw- 

 fixing, is constructed more than a fourth part of the whole 

 language. 



In a similar manner, as we have before obsened, do other 

 significants communicate their imports to the biliterals and tri- 

 literals which are derived from them. But the great length to 

 which this paper has already been extended, obliges us to reserve, 

 for some other communication, all remarks respecting them. I 

 may briefly observe, that the same simplicity of structure which 

 has been traced out in the formation of biliteral and triliteral de- 

 rivatives of the bow-letter, is also seen in the construction of 



