xxvi.] A MODIFICATION. 371 



series of facts from which to draw analogies instead of one, it 

 would have been necessary to point out that such lines if they 

 existed might either be exactly coincident on the calcium 

 analogy or nearly coincident on the shift analogy. But the 

 modification after all is a small one. Instead of insisting upon 

 the exact coincidence of lines in different substances, at one 

 particular wave-length ; a massing of the lines round that 

 wave-length would have been suggested in addition. I am 

 most anxious that this should not be regarded as special 

 pleading or as an ingenious evasion of a difficulty. I am only 

 stating now what I certainly should have stated in 1878, had 

 the facts at present known been then available. 



From the very first, as I have already stated, I have guarded 

 myself carefully against the statement that the basic lines re- 

 corded by Thaleii and myself in my earlier work were anything 

 more than coincidences with the dispersion employed. Indeed 

 in 1881, in referring to the basic-line part of the hypothesis, 

 I wrote as follows : l 



" Here we must confess both our imperfect instrumental and 

 mental means. We cannot talk of absolute coincidence because the 

 next application of greater instrumental appliances may show a 

 want of coincidence. On- the other hand there may be reasons 

 about which we know at present absolutely nothing which should 

 make absolute coincidence impossible under the circumstances 

 stated. The lines of the finer constituents of matter may be liable 

 to the same process of shifting as that at work in compound bodies 

 when the associated molecules are changed ; but however this may be 

 the fact remains, whatever the explanation may be, that the lines 

 of the elementary bodies mass themselves in those parts of the 

 spectrum occupied by the prominent lines in solar spots and 

 storms." 



We now undoubtedly know that with the higher dispersion 

 which we can now employ the lines indicated by Thalen and 



1 Nature, 1881, vol. xxiv. p. 370. 



B B 2 



