and their Mayneto-optic Properties. 269 



the red or the violet end of the spectrum according as 



This result seemed at first discouraoino- since the discoverer 

 of the resonance spectra, Prof. R. W. Wood, speaks in his 

 Clark University Lecture (of 1909, published 1912) and in 

 his paper in Physik. Zeitschr. xii. p. 1204 (1911) of constant 

 wave-lengths, and not frequency-intervals. But a glance 

 at Wood's numerical table, given in the last-mentioned 

 paper (p. 1207), for iodine vapour has sufficed to notice the 

 tendency of 8\ to increase towards the red, so that a detailed 

 comparison of the above law (9) with Wood's measurements 

 has seemed worth undertaking. The result is given in the 

 following Table, in which A, B, C are Wood's three resonance 

 series of iodine vapour excited by light of wave-length 

 X = 5461'0, 5769-5, 5790'5 respectively. The n-columns 

 contain the observed frequencies, i. e. the reciprocals of 

 AVood's wave-lengths reduced to vacuum, and 8n are their 

 differences. Since Wood's wave-lengths are, according to 

 himself, correct only within about + 1 ^.U, the last figures 

 in our n-columns are unreliable. It may be well to remark 

 also that the lines made use of for this Table seem to be the 

 chief lines of small groups or " orders" given by Wood in a 

 later publication (Phil. Mag. xxvi. p. 828, 1913). To the 

 fainter lines accompanying the chief ones in each " order/' 

 I hope to return at a later opportunity. 



The fundamental or exciting frequencies N, at the head 

 of each of the three series, are printed in clarendon. At the 

 bottom of the Table are given the mean values of 8n. 



The constancy of &n along each series is quite satisfactory. 

 Since the observed Vs are uncertain almost within +1 A.U., 

 the deviations from 8n are well within the limits of experi- 

 mental error. 



Using the above mean values for Sn in (9 a), we have for 

 1 — p, which in some way is a measure of deviation from a 

 perfectly Hookean resonator, for the three series, 



l-p A =-0111, 1-/> B = -0118, l-p o =-0119, . (10) 

 and for p itself 



p A =0-9889, ;^ B = -9882, p c = «9881. . . (10 a) 



* Wood's series of iodine, A, B, C, consist of 17, 13, and 13 lines, 

 respectively, on the red side of the corresponding fundamentals, and only 

 of two ultra-N lines, in each case, thus snowing a marked tendency to a 

 unilateral development. The ultra-N lines can easily be accounted for 

 by an extra term x~ p , but since they are but two, we can avoid 

 complication by disregarding them for the present. 



■ 





