METALS. 



307 



From 2 to 3 /A there is a weak continuous radiation, which appears to 

 be due to the oxides of the metals in the arc. For Li this is not so 

 intense, and, since the soHd material (dust) coming from the arc is 

 also less, the evidence is strengthened in favor of the emission at 2 to 

 2, IX. being due to oxides. The emission band at 4.52 /^ is also to be 

 noticed. Since it occurs only when the salts of the metals are in the 

 arc, and is not to be found in the carbon arc its source remains unde- 

 termined. If it be due to CO,, from the air, then from the shifting of 

 the maximum to the longer wave-lengths (found by Paschen, loc. cit., 

 being at 4.40 /* for the Bunsen flame) it would appear that the tempera- 

 ture of the arc is about 4,000° absolute. 



Lithium (Li), 



The chloride of lithium vaporizes so easily that no true measure of 

 the radiation from the dense vapor could be obtained. In like manner 

 the radiation from the oxides at 2 to 3^11 is also weak. The bands at 



MM 

 12 



10 



3 

 Fig. 139. 



6/^ 



o.6y fx and 0.81 1 /x (fig. 139) have the same ratio of intensity as found by 

 Snow. Beyond this point no lines could be detected, except a slight 

 band at 4.52 p.. 



Potassium (K). 



In figure 140 is shown the emission curve of potassium. The KCl does 

 not vaporize so easily and is readily adapted to the arc. Snow found 

 strong lines at 0.768, 1.155 ^^^ 1.22 /x, the first one being four times 

 the intensity of the band at 1.15 /i. In the present work the last two lines 

 were not quite resolved and no comparisons can be made. The band 

 at 1.47 ft was also found by Snow. The usual region of continuous 

 radiation is found from 2 to 3 ^u.. The 4.52 fi emission band is strong. 

 The emission band of CO2 at 4.40 /x, using a Bunsen burner, is also given, 



