Taylor.] 



328 



[Oct. 21, 



well as for the places occupied by them, would become even more 

 necessary than a distinctive form. The terms "ten," "hundred," 

 "thousand," especially, are too essentially decimal in their origin, and 

 too ineffaccably stamped by usage in their significance, to permit their 

 use in any novel application. The names, like the symbols, should be 

 both as simple and as distinct as possible. The simplest name is a mono- 

 syllable, containing but one consonant and one vowel sound. Let this 

 then be the rule of our numerical vocabulary. It will be convenient and 

 even advisable to preserve a resemblance to the popular numerical lan- 

 guage, that the analogy of structure may be the more apparent. The word 

 one will naturally give us the French " un ;" two will give us "du;" 

 three will give us " the ;" the consonant sound being really a simple one, 

 although requiring two letters in our language. The word "tre " would 

 have been better, as being very near the Latin tres, the Greek treis and the 

 original Sanscrit tri ;* but the double consonant excludes it under our rule. 



* It is a matter of curious philological interest to trace the Sanscrit or ancient Indian 

 parentage of all our modern European languages, especially in the names of the numer- 

 als. In this particular the different vocabularies of the numerous and wide-spread 

 race's,— of the Celtic, the Romaic, the Sclavonic and the Gothic, with its two great fami- 

 lies of the Scandinavian and the Teutonic, appear only as dialects of each other. The 

 names of the first ten numbers, in a few languages, are here selected, mainly from the 

 Introduction to Bosworth's Anglo-Saxon Dictionary : 



That these Sanscrit terms should have been so widely diffused, while yet no traces of 

 the Hindoo arithmetical notation should ever have been found outside of India, would 

 seem to show that this derivation was antecedent to the formation of a written language, 

 or, at least, prior to the invention of the cipher. A nomenclature may be easily trans- 

 mitted orally by tradition ; a notation could be communicated and preserved only by 

 records. 



To the Sanscrit we are indelsted for the denominations of our lowest two coins. From 

 the Sanscrit Safa or Shatum (a hundred), through the Latin centum, we obtain our 

 "cent;" and from the Sanscrit Dasa or Dai^han (ten), through the Latin decern and the 

 French disme or dime, M'e obtain the name of our "ten-cent piece." 



