2G ON THE MATHEMATICAL PROBABILITY 



worthless. On the contrary, most of them have gained a new value, since they not only 

 serve to strengthen the most important conclusions of comparative grammarians, but they 

 also furnish a rich mine for the exploration of the searchers after verbal roots. Although it 

 may be demonstrated that a given Latin word was not derived from the Greek, as was 

 once mistakenly believed, the resemblance which led to the mistake 1 still remains as an 

 evidence^ of kindred origin, and if it be rightly studied it may help us to useful results. 

 Although the French word suivre is undoubtedly more closely associated with the Latin 

 root sec, than with the Chinese suy, the latter may possibly have arisen from a similar 

 law of verbal detrition. 



2. The unfair advantage that lias been taken of the confessed errors of etymologists, 

 will undoubtedly, in time, be followed by a favorable; reaction, and meanwhile it will have 

 little influence in deterring those who are on the lookout for new discoveries. The 

 unfairness is in itself an evidence that the science which is impugned is still in its infancy, 

 and that its held is consequently mostly unexplored. Therefore, as soon as tin; fact becomes 

 established that there is a sun; basis for accurate; research, the great probability of attaining 

 satisfactory results will draw crowds of investigators. 



3. Students avIio have devoted themselves the: most earnestly to philology, have been 

 the most thoroughly convinced that all languages exhibit bonds of connection which point 

 to a common origin of some kind. The more profound their research, the deeper does 

 this conviction usually become, and it is no mark of a candid spirit, for erne' whose acquire- 

 ments are not such as to qualify him for pcrtinemt criticism, to charge it to the fascination 

 of a hobby, or of a pet theory* Such a charge might merit consideration, if the con- 

 viction we;rc solitary or exceptional, or if it were 1 made the 1 main support of a questionable 

 hypothesis. But its approach to universality should be regarded as sufficient <-vi<lene'e; that 

 it is well grounded, while; the 1 different reasons that have been imagined by different 

 investigators to explain the connection, show that they have not been led astray by devo- 

 tion to a system. Whether any of their theories be rejected or believed, the facts on 

 which the-y we:re built are incontrovertible 1 , and deserving of careful scientific investi- 

 gation. 



* I am aware that some distinguished scholars, like M. Renan, deny the existence of any traces of linguistic 



unity, but such a denial can have little weight with any one who is at all familiar with the profound researches of 

 the German philologists. The a. priori probability that there was a, primitive significance to every syllable, and 

 even to every sound, and that traces of such significance are still to he round in the languages which Max Miiller 

 has classed as "Turanian," is greatly strengthened by a comparison of such dialects as the Chinese, Egyptian, and 

 Yoruban, and it seems to me much more! natural and reasonable to regard identical roots as evidences of family 

 identity, than to attempt to explain their existence in any other way. 



