of Sulphuric- Acid Solutions. 143 



sources. I thought the latter was the preferable method, and 

 adopted it. 



On p. 311 Prof. Riicker says : — " Why, for instance, are 

 we to admit an error of 22, or three times the lower limit to 

 the maximum error, at 56*89 per cent., and insist that an 

 error of 16, which is only twice the same limit, is impossible 

 at 63*08 per cent. ?" Such a question is open to serious mis- 

 construction, for I neither rejected nor specially insisted on 

 either of these experimental points, but quoted them without 

 comment in my tables, and inserted them in my diagrams, 

 treating and accepting both of them to exactly the same 

 extent. 



In forming an estimate of how far a criticism like that now 

 under discussion, based on the examination of a limited portion 

 of one set of my results, can upset my conclusions, it is 

 necessary to bear in mind the exact nature of these con- 

 clusions. From the study of any one, or any few, particular 

 breaks, I concluded — nothing : from a study of a whole series 

 of density-results I only concluded that it was advisable to 

 make other series at other temperatures : from the study of the 

 series at four different temperatures my conclusions were 

 merely that I had " strong presumptive evidence " of the 

 existence of changes (p. 79), but that confirmatory evidence 

 from the study of independent properties was necessary before 

 such changes could be regarded as established ; and it was 

 only after obtaining such evidence from the study of three or 

 four other properties that I ventured to call this evidence 

 proof, and then only with the oft-repeated caution " that many 

 of these changes were admittedly of a very doubtful nature." 

 When, further, it is remembered that those who have called 

 my conclusions in question have confined their attention to a 

 limited portion of one particular series of results (and that 

 portion one which I myself pointed out to be most open to 

 attack), or have even selected only one particular and ex- 

 cessively doubtful break ; that in spite of this, they have, 

 while adding confirmation to my conclusions in every other 

 respect, only succeeded so far as to bring some additional 

 evidence against (without altogether disproving) the one 

 change which I myself never regarded as proved ; remember- 

 ing all this, 1 cannot fail to believe that my conclusions were 

 better founded, and my method of working more trustworthy, 

 than I had ever imagined. 

 October 1891. 



Postscript. 



I have made a detailed examination of several breaks by 

 the bent-lath method, and also by the application of parabolic 



