MORSE. — SPECTRA OF WEHNELT. 



523 



ments with the eye, on plates where the exposures are by no means com- 

 parable, are only rough approximations. The attempt has therefore 

 been made to confine any conclusions drawn from such comparisons to 

 cases so obvious and striking that there could be no possible chance of 

 error introduced by difference in exposure or width of slit. 



Careful comparison of the spectra produced by a platinum or carbon 

 point in solutions of various salts of the same metal shows that there are 

 no differences whatever corresponding to different salts. This has been 

 proven for zinc by photographing the spectrum of a platinum point 

 in zinc sulphate, nitrate, chloride, bromide, and iodide, and also that of 

 a zinc point in sulphuric, nitric, and hydrochloric acids. These spectra 

 are all identical within the limits of the method. The same has been 

 shown for aluminium for an equally extended series of salts and acids, 

 and in the cases of other metals for a less number of combinations. The 

 tables give the data for a stated combination in each case, but apply 

 equally well to any other for the same metal or for any salt of the same 

 metal. 



The present paper is concerned only with the photographs of the 

 Wehnelt in the first spectrum, where comparison can be made with the 

 full spectrum of spark and arc without disturbance from overlapping. 



o 



The dispersion is only sufficient to give an accuracy of about 1 Ang- 

 strom unit in the comparison of wave-lengths. 



LITHIUM. [Plate 2, Fig. 7.] 



Spark. Lithium chloride on carhon electrodes. 

 Arc. Lithium chloride on carbon electrodes. 



Wehnelt. Carbon point in solution of lithium chloride. (The impurities in the 

 carbon are iron, calcium, and aluminium.) 



