MORSE. — SPECTRA OP WEHNELT. 541 



exhibition of an oxide spectrum when the dissociation temperature of 

 the oxide in question lies at a point higher than the dissociation tem- 

 perature of water. That the free hydrogen in the mixture of gases 

 could under these conditions form metal-hydrogen compounds * seems 

 more than doubtful, yet the variations between Wehnelt and spark and 

 arc in air bear a striking resemblance to the variations in the arc spectra 

 of the same metals in air and hydrogen, and in the latter case the 

 necessity for the assumption of such metal-hydrogen compounds seems 

 to be felt.f 



Although the spectra under discussion are in many ways similar to 

 those produced by allowing the spark to pass from a metallic point to a 

 solution of a salt, many of the characteristic bauds of compounds pro- 

 duced by the latter method are absent in the Wehnelt. Compound 

 spectra, varying from salt to salt, are the rule in the more volatile 

 metals $ when the spark passes to a solution of one of their salts. The 

 point used gives, in the Wehnelt, all the strong lines of the metal, even 

 of platinum, while in the other method the lines of the electrode are but 

 faint. 



The tables and plates show the following general points of especial 

 interest: — 



1. The lines of the Wehnelt spectra include some which have been 

 usually ascribed to the spark and some which have been ascribed to the 

 arc ; usually the spectrum is closely allied to that produced in the spark, 

 but often some of the strongest lines are missing. Other researches which 

 give evidence as to the effect of varying conditions on spectrum lines 

 serve to strengthen the conclusion which must be drawn : — that there is 

 no sharp boundary between arc and spark, and that the transition from 

 " arc " spectrum to " sjjark " spectrum is a gradual and continuous one. 



2. Under constant conditions a spectrum may contain the strongest 

 lines of the condensed spark and at the same time lines usually ascribed 

 to the flame. § 



3. Under the same conditions the " oxide bands " may be present at 

 the same time as the strong spark lines. § 



4. Under the same conditions the spectrum of zinc contains, beside 

 lines belonging to both spark and arc spectra, a band spectrum not 



* Vid. Liveing and Dewar, 1. c. 

 t Vid. Basquin, 1. c, and Porter, 1. c. 

 X Vid. Leroq. Spectres. Lum., atlas. 



§ See on these points Liveing and Dewar, 1 c. ; also Proc. Hoy. Soc, 32, 189 

 (1881). 



