960 Mr. Beer and Prof. Tyndall on Manometric 



Experiments with Gas-denuded Carbons. 



It appeared, however, .that after the carbon had been 

 denuded of its gases' there was still a residual pressure 

 which might possibly be attributed to an electric wind. 

 This was borne out in the following long series of experi- 

 ments with carbons which had been as far as possible 

 denuded of gas by running the arc at a high current for 

 5 minutes before readings were taken, the current used 

 for denudation being in all cases greater than that of which 

 pressure measurements were subsequently taken. 



The carbons used were filed down to a diameter of about 

 8 mm., partly to ensure a more complete denudation of gas 

 and still more to minimize as far as possible the tendency 

 of the arc to leave its central position. This tendency was 

 one of the main experimental difficulties of the work 

 Considerable care and patience had to be exercised before 

 a satisfactory set of readings could be obtained, because 

 lack of centrality was often a matter of degree only. If the 

 discharge shifted — as it frequently did — to one side of the 

 carbon, this was immediately apparent without looking at 

 the arc, as the gauge readings dropped instantly. But when 

 it was nearly central, it was quite possible to mistake the 

 appreciable reading which was obtained for the maximum 

 reading which would have resulted from perfect centrality. 

 In all cases of doubt the arc was adjusted and kept burning 

 for some time, and the maximum gauge reading during this 

 period was the one recorded. 



For small currents and large holes, however, the incan- 

 descent patch of light on the anode was smaller than the hole 

 itself, and it is doubtful how far the readings then are 

 quantitatively valuable except for purposes of comparison. 

 For the cathode with any hole but a small one this was always 

 the case, and the readings were most variable and uncertain. 



Another experimental difficulty was " hissing," which 

 considerably restricted the scope of the observational work. 

 It was particularly noticeable with post-war carbons. When 

 the arc was hissing, the luminous patch on the anode was 

 quivering and travelling about, with the result that the 

 effective pressure was greatly reduced on that electrode 

 owing to the departure from centrality, if from no other 

 cause as well. On the other hand, hissing caused the 

 cathode pressure, which was generally negative for a large 

 hole, to become positive. 



It might also be mentioned that to obtain consistent 

 readings for such minute pressures, precautions had to be 



