Belatioiiship between Spectra and Atomic Weights. G99 



•very close to poine })oiiit representing a speclnil line. There- 

 fore it* no account is taken whether or no this spectral line is 

 honioloi>ous to the spectral lines correspondino- to the points 

 where the same straight line intersects the other spectra, the 

 whole arounient Falls to the ground. 



Dr. Watts has secondly availed himself of a remark made 

 l)v Rvdberg, and by Kayser and the writer, that the differences 

 of oscillation-frequencies of homologous doublets are ap})roxi- 

 matelv })roportional to the square of the atomic weight. But 

 here, as in the first case, he does not keep to homologous lines. 



By neglecting this restriction, duly insisted on by Ramage 

 and by the other authors, it is possible to make the atomic 

 weight of a substance come out anything we please. There 

 is no ditficulty for instance in making the atomic weight of 

 radium come out eijual to 100. We need only draw the 

 spectrum of radium in damage's dingram in the parallel 

 corresponding to the atomic weight 100, then choose a point 

 in the radium s])ectrmn and a point in the strontium spectrum 

 such that the straight line connecting them wdien prolonged 

 intersects the si)ectrum of barium in some part where the 

 spectral lines lie close. Then We take the nearest point 

 representing a spectral line of barium. In this manner we 

 get three lines and have only to assume that chey are 

 homologous to make out 100 as the atomic weight of radium. 

 In a similar way doublets may be found in the spectrum of 

 radium and the spectrum of, say, mercury, from w^hich w^e 

 make out 100 as the atomic weioht of radium. 



Take for inst^mcc the following lines 



Sr. 

 X. 



(1) 4892-20 



(2) 4784-43 



(3) 4729 93 



(4) 4678-39 



(5) 4338 00 



(6) 472^J-93 

 From the 



Ea. 



20440-7 

 20901-1 

 21142-0 

 21374-9 

 23052-1 

 211420 

 atomic 



X. 



5081-20 

 4971-88 

 4918 61 

 4826-12 

 4533-33 

 4837-59 



1;X. 



19680-4 

 20113-1 

 20330-9 

 20720-6 

 22058-9 

 20671-4 



Ba. 



X. 



5965-06 

 5853-91 

 5805-86 

 5473-94 

 5473-94 

 5294-40 



1/X. 



16764 

 17083 

 17224 



18268 

 18-268 

 18888 



weight 137-4 of barium and 87*0 of 



strontium we find for the atomic weight of radiu 

 to Dr. Watts' first method of calculation : — 



(1) 99-95 



(2) 99-9:5 



(3) 99-97 



(4) 100-17 



(5) 100-00 



(6) 100-07 



Mean 100-015 



m accordino' 



