1832.] producing the reversal of spectral lines of metals. 263 



another carbon rod b, imperforated, is passed into the lime block 

 through a second hole at right angles to the first, so that the end 

 of the rod b meets the rod a in the middle of the block of lime. 

 The rod b is connected with the negative electrode of the dynamo 

 machine, and after contact is made between the two carbons is 

 raised a little so that the arc discharge continues between the two 

 carbon rods within the block of lime. 



By passing a small rod of carbon into the perforation from the 

 farther end, a luminous background can be obtained all across the 

 field, and then, as the walls of the tube are hotter than the metallic 

 vapours between them and the eye, the metallic lines are only seen 

 reversed. A very slight alteration in the position of the small 

 carbon rod forming a core to the tube makes the lines disappear, 

 or reappear, or shew reversal, and as the core is adjusted by eye 

 observation before photographs are taken, all the conditions of the 

 experiments are known and are under easy control. 



By this arrangement we are able to make observations as the 

 temperature rises and as it falls, "and so to trace the influence of 

 temperature in many cases in which the extent of that influence 

 was before doubtful. The temperature attainable is doubtless far 

 below that of the arc, but still it is quite sufficient to maintain 

 iron and aluminium in the state of vapour, and shew the reversal 

 of the lines of these elements with singular sharpness. The 

 temperature of the interior is sufficiently high to transform the 

 diamond into coke, even in a current of hydrogen, and the result 

 may be taken as proving that the temperature is above that of the 

 oxyhydrogen flame. We have taken photographs of the violet 

 and lower part of the ultra-violet spectrum given by the tube at 

 successive intervals while the temperature was rising, and noted 

 the following results. When commercial carbons were used, which 

 always contain iron, aluminium, manganese, calcium and mag- 

 nesium as well as traces of other metals, the first lines to be seen 

 as the temperature rose were the potassium lines, wave-length 

 4044—6, next the two aluminium lines between H and K became 

 conspicuous, then the manganese triplet about wave-length 

 4034, and the calcium line, wave-length 4226, then the calcium 

 lines near M and an iron line, probably M, between them, and 

 then gradually a multitude of lines which seem to be all the 

 conspicuous iron lines between O and h. At this stage, when the 

 core is inserted to give a background, the bright continuous spec- 

 trum is crossed by a multitude of sharp dark lines, vividly recalling 

 the general appearance of the solar spectrum. In the higher region 

 the continuous spectrum extends beyond the solar spectrum, and 

 the magnesium line, wave-length 2852, is a diffuse dark band, 

 while all the strong iron lines about T, and the aluminium pair 

 near S, are seen as reversed lines. The behaviour of the calcium 



