474 



Profs. G. D. Liveing and J. Dewar. [Apr. 3, 



3827 -6 3745 3 



3825 -2 3736 -5 



L 3819 '8 3734 -5 



3815-3 M3727 



3799 -3 3719 '6 



3795 3709 



3787 3705 '5 



3766 -6 3647 



3763 -4 3631 



3757 7 3618 



3749 -5 3608 "2 



3747 '2 



N 3580 -5 

 3568 -9 

 3564 

 3525 7 

 3496 -8 

 3489 '8 

 3476 

 3465 -5 



3440 



T 3019 '8 



As a rule no iron lines above make their appearance ; in a few 

 plates T is visible, and it is possible that other lines may be obscured 

 by the water spectrum, which always comes out and extends from 

 near s to below R. Above T no line at all is visible in any of the 

 photographs, though the spark lines come out strongly enough, and 

 several of the strongest groups of iron lines, both of spark and arc 

 lines, are in the region beyond T. 



The spark by which the gas was fired passed in general between a 

 platinum wire and the side of the small brass tube, and was out of 

 view ; but in order to make quite sure that the lines were not due at 

 all to the spark, the brass tube was lined with a tube of platinum 

 foil which projected beyond the brass tube a short distance into the 

 larger tube, and the spark passed between the platinum wire and the 

 platinum tube. It was found that the same iron lines made their 

 appearance in the flash whichever way the spark was passed. 



Other experiments were made with explosions of carbonic oxide 

 and oxygen, and with coal-gas and oxygen. The explosions of these 

 gases were attended with much more continuous spectrum, and the 

 metallic lines were not always as well developed as they were with 

 hydrogen and oxygen, but on the whole there were as many metallic 

 lines photographed from tbe flashes of carbonic oxide as from those 

 of hydrogen. There is an uncertainty about the explosion of the 

 carbonic oxide mixture which we cannot account for, even when we 

 take into account the remarkable effects of relative dryness of the gas 

 on the explosions discovered by Mr. Dixon. Sometimes the explosions 

 were so violent as to break the plate closing the end of the tube, 

 though this had resisted the explosions of the hydrogen mixture, 

 while at other times the wave of explosion passed slowly along the 

 tube. The gas was in all cases confined over water and passed 

 directly from the gasholder into the tube. 



When the iron tube was lined with copper foil, only one copper 

 line in the visible spectrum, X =5104*9, was seen, and in the ultra- 



