1880.] Prof. J. Dewar. Studies on the Electric Arc. 87 



As the yield of cyanogen compounds did not appear to be diminished, 

 and it appeared almost impossible to get the large volume of air in the 

 tin vessel perfectly dry, another plan was adopted. The poles were en- 

 closed in an egg-shaped glass globe about eight inches long and six inches 

 in diameter, thus diminishing the volume of air to be dried and dis- 

 pensing with the water covering. The globe, balanced by a system of 

 pulleys, was firmly attached to the lower or negative pole, with which 

 it moved without impeding the action of the lamp. 



Dry air was sometimes forced through the negative carbon itself, at 

 other times through a glass tube passing up the side of it into the 

 globe, the products being drawn through the positive pole as before. 



As the globe soon became intensely heated, and as a far larger supply 

 of dry air was forced in than was drawn out, it is inconceivable that 

 any moisture could remain near the arc after it had been burning for 

 a few minutes. 



Seven consecutive experiments, each of ten minutes' duration, made 

 with the same purified carbon poles, did not show any diminution in 

 the quantity of hydrocyanic acid, unless in one of the experiments, 

 when the arc would not be drawn into the interior of the carbon tube, 

 but persisted in rotating round it.* 



The inference drawn from these experiments was, that the drilled 

 carbons, even after prolonged treatment with chlorine, still con- 

 tained a quantity of combined hydrogen, and a few organic analyses 

 showed that the amount of ash and combined hydrogen in the various 

 samples was never less than about 0'75 of the former and as much 

 as 0*1 of the latter. Poles made with especially purified carbon by 

 Messrs, Siemens for these experiments proved to be no better in respect 

 to the quantity of hydrogen and ash they contained. 



The well-nigh impossible problem of eliminating hydrogen from 

 masses of carbon such as can be employed in experiments of this kind, 

 proves conclusively that the inference drawn by Mr. Lockyer,f as to 

 the elementary character of the so-called carbon spectrum from an 

 examination of the arc in dry chlorine, cannot be regarded as satis- 

 factory, seeing that undoubtedly hydrogen was present in the carbon, 

 and in all probability nitrogen in the chlorine. 



Experiments with Carbon Tubes. 

 In order to ascertain whether the formation of hydrocyanic acid and 

 acetylene in the arc was really due to transformations induced by some 



* Cyanogen is difficult to recognise in presence of prussic acid when in small 

 quantity, especially when impurities from the carbons complicate the tests. In 

 speaking generally of the formation of this acid in the arc, I do not mean to exclude 

 the possibility of cyanogen being formed as well. 



f " Note on the Existence of Carbon in the Coronal Atmosphere of the Sun." 

 " Proc. Roy. Soc," vol. xxvii, p. 308. 



