CHAP. XII.] CORRESPONDENCE. 377 



personally ; and I could not help saying, with regard to the 

 sentence which begins it, p. 264, that I thought you would 

 object to inferences founded on comparison by contrast, and 

 that the proper way was to compare by matching recognised 

 browns with a compound. 



Listing's paper, mentioned in p. 102, was to me rather 

 a paradox, I had got to regard the subdivisions of the 

 colour-scale which are assumed in language, as something so 

 arbitrary. If you cared to see it, I have that number of 

 Poggendorff ; I think he would hardly agree with your J. J. 

 Miiller. I wish you or Benson could eradicate the insane 

 trick of reasoning about colours as identified ly their names. 

 People seem to think that blue is blue, and one blue as 

 good as another. Benson's book I have seen (since I heard 

 of it from you), but not read. His way of mixing by 

 means of a prism is very happy. ... I wish, with your new 

 set up box, you would just put the prism observations into 

 relation with the disk ones. It would be very easy. White 

 we have got; and it would only be strictly necessary to 

 determine two other standard colours, such as vermilion or 

 emerald, by reference to the spectrum primaries. You don't 

 say whether your dwellers in Mesopotamia and elsewhere 

 agree on the whole better or worse than " J " and " K," 

 who, I suppose, agree for better and for worse. To judge 

 by their case, the discrepancy would be a little diminished 

 by taking as units of colour co-ordinates for any given pair 

 of eyes, not the intensities of the primaries as they appear 

 in the spectrum, but their intensities as they appear in the 

 combination which to that pair of eyes makes (say) white. 

 This amounts to transforming from trilinear to Hamilton's 

 anharmonic co-ordinates with white for the fourth point, 

 in the language of " scientific metaphor." On the other 

 hand, ought not all your co-ordinates to be cooked by 



,,. , . , d. scale, page 68 _ 

 multiplying by - ** ? 

 d . wave-length 



You know where I learnt scientific metaphor. I have 

 read the address in Section A more than once with much 

 pleasure, and, I hope, profit in proportion. The pleasure, I 



