178 



is an addition to the treasury from which future investigators will 

 draw their wealth, and he will rather leave the rejection of the base 

 metal to those who can find a proper use for the good, than attempt 

 to depreciate the whole by directing our attention exclusively to that 

 which he imagines may be bad. 



All etymological researches may be arranged in three classes, viz. : 

 1. Immediate or Derivative ; 2. Comparative; 3. Radical or Ger- 

 minal. 



1. Derivative P]tymology is limited to the determination of the 

 language from which any given word is immediately derived, and 

 usually without much change of form. Its method is the one adopted 

 by ordinary lexicographers. 



2. Comparative Etymology traces similarities and subordinate dif- 

 ferences of sound and meaning, between different languages which 

 are generally, though not necessarily, of the same family or origin ; 

 and sometimes attempts to assign the cause of such similarity or 

 difference, whether accidental, organic, derivative, or otherwise. 



3. Radical or Germinal Etymology, endeavors to ascertain the 

 essential elements of words, without regard to their immediate deri- 

 vation or remote affinities. The most noteworthy example of this 

 class, is to be found in the labors of the Indian grammarians, who 

 compiled the lists of Sanscrit radicals, an example that might be 

 advantageously imitated by supplying similar lists for other languages. 



The philological labors of Oriental scholars, such as Grimm, Bopp, 

 and others, have been mostly comparative. In the sanguine enthusi- 

 asm of the early students of Sanscrit, the hope was often indulged, 

 that the key to all languages had been found, and that all dialectic 

 mysteries would be speedily removed. Since these extravagant ex- 

 pectations have been generally given up, there has been, perhaps, an 

 undue tendency in the opposite direction, to disparage the study of 

 Sanscrit for etymologic uses. In contending against such a ten- 

 dency it is not necessary to assume any theory with regard to the 

 relative antiquity and position of the Sanscrit and its kindred lan- 

 Kuao^es ; it is sufficient to show that the Sanscrit is one of the earli- 

 est known members of the Indo-European family, and that, therefore, 

 from its introduction into a comparative list with any other member 

 of the same family, it is reasonable to hope for aid in ascertaining 

 primitive roots and determining radical significance. 



Whatever value may be attached to the following vocabulary must 

 belong to it in virtue of its adaptedness to the wants of the com- 

 parative etymologist. It is a list, not of roots, but of analogues; 



