292 THE COMPUTING BOY. 



Observations may be removed with respect to the even digit, though I think 

 support ofthe * ma ^ sa ^ e b' challenge him to produce a single instance of a 

 talents, and child from six to eight years of age, who would be able to 



itemeth%sof COm P re ^ lend the met * 1<xl > mucn less to apply it with facility and 

 computing by rapidity. Be this as it may, it is confessed by Mr. A. that the 

 Zerah Col- case f numbers ending with 5 is one which l( can deceive," 

 and I accordingly expected to find that Mr. A. had given the 

 boy various examples of this am biguous case, and that he had 

 uniformly found the boy incapable of answering such questions 

 correctly, or that he had obtained from him an a(knowledge?nent 

 that such questions were beyond the reach of his powers to 

 answer. Yet nothing of this kind is mentioned by Mr. A. 

 who leaves us totally in the dark upon the very point which 

 would have cleared up the difficulty. Are we to imagine, then, 

 that Mr. A. though aware of the importance of putting such 

 questions, for the purpose of ascertaining whether M. Ralliers 

 method was employed or not, yet omitted to ask them ? Or, if 

 he did ask questions of this kind, and received wrong answers, 

 (which must have been the case if the boy employed the me- 

 thod alluded to,) how is it that he has neglected to avail himself 

 of the statement of this circumstance, so materially affecting 

 his claims to a discovery which he evidently considers to be an 

 important one. 



But allow me, Sir, to examine the merits of this rule in its 

 application to the square root. Let us suppose the boy was. 

 requested to extract the square root of the number 42436 , 

 here it is obvious the first figure of the root would be 2, and 

 the last either 4 or 6 ; — if 4 be taken, then 4 or Q would be 

 found to be the middle figure j but if 6 be used, then O or 5 

 would be the middle figure j hence there would be no fewer 

 than four different roots obtained by M. Rallier's method, of 

 which four the boy could not possibly know the correct one, 

 and he might assign either 20(3, 256, 244, or 2^4 for the root 

 of the required number. This is no particular example, se- 

 lected for the purpose of exhibiting M. Rallier's rule in the 

 most unfavourable point of view ; for it will be found upon 

 trial, that had any other number been proposed, four different 

 results would have been obtained by this rule; and that if a 

 number ending with 5 had been proposed, no less than ten 

 different results would have been produced, since all square num- 

 bers 



