406 Prof. Whewell on Isomorphism, 



If now in the analyses thus expressed, we add together the 

 numbers in the columns A and Fes, and also those in the co- 

 lumns Fe, Mn, C, M, we have certain numbers which we 

 must compare with each other, as follows : 



Now in these columns the approximate equality of the 

 numbers in each column is remarkable. None of those in the 

 second column differ much from 2; none of those in the third 

 column much from 3 ; the first column being invariably 4. 

 We have a degree of resemblance altogether different from 

 what appeared in the former table. And this becomes stronger 

 when we notice, that in all the cases the number in the third 

 column is very nearly one and a half times that in the second, 

 and that this proportion obtains even where the numbers them- 

 selves differ from the average, as in (2). We may observe that 

 in this analysis (2), which is much the furthest from the mean, 

 we have only to suppose an excess of 7 per cent, of silica, 

 in order to make it agree with the rest; and that in this case 

 there must be some inaccuracy in the analysis, since the pro- 

 portions of the separate ingredients amount to 102 instead 

 of 100. In (4), where the second column is too large, and the 

 third too small, we are led to ask, whether it is possible that 

 in the analysis any portion of Fe has been converted into Fes : 

 the probability of such an inaccuracy must be left to chemists 

 to determine. 



But even without making any allowances, I think the agree- 

 ment of the above numbers is as close as we could expect, 



I will leave the matter to bo decided by those who may attend to it. 

 I shall use the notation which I have explained, because it is the only one 

 which will answer my purpose, as well as the only one which is algebrai- 

 cally consistent. Those who do not find any inconvenience, or see any 

 incongruity, in that of Berzelius, have no reason for relinquishing it. The 

 relative advantages of the two will be discovered by any one who has to 

 work with them : and there would be little use in a further discussion of 

 these on general principles. 



and 



