of the H and K Lines of the Spectrum of Calcium. 435 



Kayser and Runge found 106 lines of the calcium spectrum to 

 belong to the series of triplets ; among the remaining lines they 

 pointed out pairs with constant differences of wave-frequencies. 

 Notably H and K, with a difference of wave-frequency of 222*9, 

 and the more refrangible pair at 3737*08 and 3706T8, with a differ- 

 ence 223 1. 



Messrs. Humphreys and Mohler in their experiments on the effect 

 of pressure on the wave-lengths of metallic lines, found that in the 

 case of calcium, the H and K lines were shifted only one-half as 

 much as the blue line at 4226*9. We know far too little to justify 

 us in forming any theoretical conclusions from this peculiarity of 

 behaviour. Indeed there are no certain reasons why the lines of any 

 substance should be equally shifted. 



It is well known that calcium, in common with nearly all sub- 

 stances, gives a more complex spectrum under the conditions of the 

 arc and spark than under those of a flame. Now in the Fraunhofer 

 lines we have, as first shown by Kirchhoff and Bunsen, absorption 

 spectra of the elements which correspond, speaking broadly, with those 

 of the bright-lined spectra of the same substances as they are pro- 

 duced by the spark. In order, therefore, to study the modifications 

 which the calcium undergoes in the higher regions of the chromo- 

 sphere, in the prominences, and possibly in lower parts of the corona, 

 as well as in the atmospheres of stars of different orders, it was 

 clearly desirable that we should start with an ordinary spark spec- 

 trum. It w T as suggested to us strongly by the known rarer state of 

 the gases in the regions above the photosphere, as well as by my 

 long experience with the behaviour of calcium in comparison spectra 

 that the modifications of the calcium spectrum which we were seeking^ 

 would be likely to show themselves under conditions of greatly 

 reduced density of the calcium vapour. 



Experiments. 



For reasons which will be obvious later on, we elected to use 

 throughout the experiments a spark of very small intensity. 



1. The break of a 6-inch Apps coil was fixed at the position of 

 smallest acting force of the spring. So much battery power only 

 was employed as would be just sufficient to move the break. Under 

 these conditions, when a jar was not in connection, the feeble spark 

 would not pass when the distance between the points exceeded 

 1£ inches. 



2. In all the experiments a jar was intercalated. 



3. The same length of exposure, a very short one of a second and 

 a half, sufficient to bring out only the strongest lines of the spec- 

 trum, was used in each experiment. 



