356 Prof. A. Griinwald on the Spectrum of Water-vapour 



results a more or less continuous spectrum, in which only some 

 light-maxima show the position of the former chief lines of 

 the normal spectrum. 



II. 



The most important of the highly remarkable relationships 

 between the line spectra of H, 0, and H 2 vapour are : — 



(1) " All the waves of the second or so-called compound- 

 line spectrum of hydrogen, which have been so admirably 

 determined by Dr. B. Hasselberg, of Pulkowa, and which are 

 due to a more complicated structure of H molecules, may 

 (after exclusion of the lines belonging to the elementary line- 

 spectrum) be converted into corresponding wave-lengths of 

 the water-spectrum by multiplying by J-." 



This proposition, which was discovered at first empirically 

 (for a large number of ra}^s) by comparison of the second 

 hydrogen-spectrum with the water-spectrum as far as known 

 at the time, and then theoretically by means of theorem I., 

 recognized as holding good generally, is, according to I., a 

 simple consequence of the fact that the modified hydrogen- 

 molecule H7 in H 2 0-vapour occupies exactly half the volume 

 it occupies in the free condition. This observation put me in 

 possession of a large number of previously unknown wave- 

 lengths of the water-spectrum, of those namely which hy- 

 drogen produces under the influence of the oxygen in water- 

 vapour, and enabled me to send (on May 9, 1887) to Prof. Gr. D. 

 Liveing, at Cambridge, a long list of hitherto unknown wave- 

 lengths of the water-spectrum for the purpose of experimen- 

 tally testing the 'correctness of my mathematico-chemical 

 disturbance-theory ; a prediction which, as I am informed by 

 letters of the 19th and 21st of June from Prof. Liveing, has 

 been verified in the most satisfactory manner, as far as the 

 observations have vet been made. The correspondence is 

 shown for the rays from A = 2800 to 2607*8 and from X=2603 

 to 2449, in the following Table. 



It has not yet been possible to verify the rest of the lines 

 predicted by me on the 9th of May, from.X = 2449 to \= 2207, 

 on account of their extreme feebleness. Below 2200 Messrs. 

 Liveing and Dewar have not yet been able to detect any lines 

 of the water- spectrum. Nevertheless, according to my theory, 

 there must exist a still more refrangible section of the water- 

 spectrum of altogether extraordinary feebleness, the wave- 

 lengths of which correspond to the wave-lengths of the 

 hydrogen-spectrum given by Dr. B. Hasselberg in his Zusatz 

 zu seinen Untersuchungen uber das II. Wasserstoffspektrum, 

 1884. 



