140 An Essay on the Relationship of [Jan. 



Remarks referring to the numeral 3.— Also this numeral has in a few 

 of these languages some similarity with the 3d person of the pronoun. — 

 Ur (we) in Hungarian, may possibly be the root of the numeral harom, as 

 well as har (= her) in the Thu'ringian provincial dialect ; and er (he) in 

 German, may perhaps, as well as ur, be the root of erek. — Sie (= she) 

 German, may be identic with the Zend and Persian numeral se, si. 



Trin* in the Gipsy language seems to form the transition from the 

 Sanscrit tri, to the Hindostani tin. 



Remarks on the numeral, 4. — The forms reaaapa, tr iropev, peduor 

 and quatuor, shew us evidently, how the same root may be diversified 

 and corrupted in different languages and dialects. Even the forms, 

 fiorer, four and vier (pronounced fir), appear to be corruptions of 

 7rnope<i by dropping the letter t, just as the Persians have changed the 

 / into h, (we might as well say, dropped it) in the Sanscrit, catur, or 

 the Zend, catvere ; or in a similar form of the same word in some other 

 cognate ancient dialect. 



Busbeck, a learned Dutch ambassador at the court of Constantinople, 

 has in his operibus (Epistolis Turcicis) Amst. Elzev. 1660, an account of 

 some men, whom he met with at Constantinople, and whom he believed to 

 belong to avery ancient German colony in the Crimea. He gives a list of 

 words of their dialect, many of which are evidently Teutonic, others be- 

 longing to quite a strange language. The numerals are : 1, ita, (in Icelan- 

 dic : einr (masc.) ein (fern.) eitt (neuter) ; 2, tua ; 3, tria; 4, fyder ; 5, fynf ; 

 6, seis ; 7, sevene ; 8, athe (Icelandic ath). —Fyder is an important 

 middle-link between irnope?, peduor, fiorer, four and vie, v r, and shews 

 that the Teutonic four and vier are corruptions of ^tropes as well as the 

 Persian chahar (cahar) is a corruption of chatur (catur). 



Remarks on the numeral, 5.— Just as the English have dropped the 

 letter n, in the German funf and say five, so the Turks have dropped n 

 in panj, and say, besh ; and the Bohemians, according to the same 

 analogy, have dropped the n in irevre and formed pet (or rather p*et 

 the pi being p mouille), neither can I doubt that panj, ^rev-re, pimp, funf, 

 and five on the one hand, and hing and quinque on the other, are all of 

 them, modifications and corruptions of one and the same root. The 

 change, in sound, from quinque into cinq in French, makes this very- 

 probable, t 



* I am not sure, however, whether my manuscript is here correct; perhaps the Gipsy 

 word might be triu ; for the materials contained in this treatise were collected more than 

 20 years ago, in Germany. Since that period, I had more important duties than "to ask 

 every word for its passport, to know, whence it comes, and whither it goes." 



t I do not mean to say that the Bohemians have borrowed their form of pent or pet (p*et) 

 from the Greeks ; but the Bohemians, as well as the Greeks, have either received this 

 word from a tribe of middle Asia, from which both these nations branched out, or they 

 (the Greeks and Bohemians) changed the (probably) original vrordpanj (after their separa- 

 tion) analogously into Trevie and pent or p*et. This remark is applicable to many similar 

 cases throughout this essay. 



