1882.] producing the reversal of spectral lines of metals. 257 



be observed, and the metallic vapour not only in a layer of some 

 thickness, but heated up till it is capable of freely emitting the 

 light of those lines which we want to reverse. In the case of the 

 less volatile metals such as iron and aluminium it would not be 

 expected that these conditions should very easily be fulfilled. 



After trying the oxy-hydrogen jet to ignite the bottom of a 

 tube bored in a block of lime, which gave us reversals of the well- 

 known green and orange bands of lime, and of the blue line of cal- 

 cium which was barely reversible in tubes heated in a crucible 

 furnace, we turned our attention to the electric arc as a source of 

 heat both to give the requisite bright background and to vapourize 

 the metals examined, but enclosed it in a block of lime or mag- 

 nesia by which we were able to maintain a considerable amount of 



a. block of lime or magnesia. 



b. hole bored to the centre of block through which observations were made of 

 the arc. 



c. hole by which metals were dropped into the arc, usually covered with a piece 

 of lime. 



d.d. carbon rods forming electrodes of dynamo-electric machine. 



refractory metal in a state of vapour, and from the tube-like form 

 of the opening through which the arc was viewed could observe the 

 absorption of a tolerable thickness of such vapour gradually 

 diminishing in temperature as it receded from the arc. The 

 greater part of our observations were made with some one or other 

 of several modifications of this apparatus. 



For the ultra violet part of the spectrum photography was 

 used. Generally 15 photographs were taken on the same plate 

 in order to vary the time of exposure, as good impressions of faint 

 lines cannot be obtained without over exposing the strong lines : 

 and also in order to catch different phases of effect as the metal 

 introduced evaporates. 



