THE CONSTITUTION OF THE ELECTRIC SPARK. 213 



ceased. Further experiments are in progress, and the matter is only mentioned 

 here because any one repeating our experiments on the influence of self-induction 

 \\ill be struck by the intensity of the H and K lines when the spark is taken 

 from silver poles. The extraordinary persistency with which the slightest trace of 

 calcium is known to givr tin- lines in ((Ufstion renders any investigation on the actual 

 source of the calcium present very difficult. The effect of polishing, which was 

 necessary to obtain good sparks, undoubtedly results in calcium contamination, as we 

 were able to ascertain by experiment, but silver poles carefully prej>ared and not 

 brought into contact with any foreign material, but polished by friction Hgainst each 

 other, still gave the H and K lines very prominently. 



10. Conclusion. 



We point out in conclusion the principal results arrived at in the preceding 

 investigation. 



We do not consider that the numbers arrived at for the molecular velocities can 

 lay claim to great accuracy, owing to the irregularities in the results, some of which 

 have not quite been traced yet. But we think that we are able to say generally 

 that metals of light atomic weight, like aluminium and magnesium, are projected from 

 the poles with greater velocities than the heavier ones we have tried, such as zinc, 

 cadmium and mercury. Different curvatures of different lines are marked in the case 

 of bismuth, and are not easily explained, except by assuming the presence of different 

 kinds of molecules having different masses, the lighter ones diffusing more quickly. 

 We have thus established a method which is likely to prove of extreme value in 

 separating the effects of different molecules. 



Our experiments also allow us to draw another important conclusion. When two 

 lines of a metal are of unequal intensity, it is not always due to the fact that at 

 any period of the discharge the vibrations which appear the strongest are really the 

 most intense. Our eye or the photographic film only perceives the total energy sent 

 out, and the time of luminosity is in many cases very different for different lines. 

 A vibration which is weak but persists may appear stronger than one of greater 

 intensity which only appears for a very short 'time. 



We have been led in addition to a new method of taking spark spectra with an 

 induction coil, by the discovery that self-induction in the spark circuit leads to the 

 disappearance of the air lines, which are often very troublesome in the investigation 

 of spark spectra. 



Finally, the appearance of the calcium lines in the photograph of the silver spectrum 

 lias shown the existence of a new type of spectroscopic lines, namely, one that starts 

 in the centre of the spark and is propagated towards the poles. 



