15^ 



NATURE 



{June 17, 1 8 So 



meter-discharger, shown in Fig. 2, under the bell-jar of 

 an air pump "to which was attached a gauge about 36 

 inches long in order to indicate the pressure of the con- 

 tained gas. In the first instance the disks were adjusted 

 to the striking-distance at atmospheric pressure for the 

 battery of 1 1,000 cells. Afterwards a less number of cells 

 was connected with the disks and the bell-jar gradually 

 exhausted until the discharge occurred ; the height of the 

 mercuiy in the gauge was then read off. Then a less and 

 less number of cells was connected with the disks and the 

 operation was repeated. 



In air the discharge took place at ordinary atmospheric 

 pressure with 11,000 cells when the disks were o'i3 inch, 

 3'3 mm. distant, and with 600 cells at an average pressure 

 of 10 mm. 



In hydrogen it took place at atmospheric pressure with 

 11,000 cells when the disks were 0-22 inch, 5'59 mm. 

 distant ; and with 600 cells at an average of 14 mm. 

 pressure. 



In carbonic acid, at atmospheric pressure with 11,000 

 cells, when the disks were o'i22 inch, 3'096 mm. 

 distant ; and with 600 cells at an average pressure of 

 5 '2 mm. 



The numbers obtained for air, hydrogen, and carbonic 

 acid respectively were plotted down on millimetre scale 

 paper, the abscissa; being i mm. = 2,500 M, the ordi- 

 nates i mm. = 25 cells, and curves drawn to give a mean 

 of the several observations. These appeared to resemble 

 hyperbolic curves so closely that true hyperbolic curves 

 were found partly by a geometric construction, partly by 

 computation, which would intersect the mean experimental 

 curves in two points. The results of experiment were 

 again laid dow-n on these new curves, and it was found 

 that they did not differ more from them than they did 

 from each other. 



The ratio of the transverse axis (pressure) to the 

 conjugate axis (potential) of the hyperbolas set out on the 

 above-mentioned scale was — 



For air 

 ,, hydrogen 

 ,, carbonic acid 



0-9665 

 I '0170 

 I '0690 



The striking distances at atmospheric pressure for 

 spherical surfaces 3 inches radius and i'5 inch diameter, 

 with various potentials, as given in Part I. page 68, curve 

 VIII. and at page iiS, also those for nearly flat surfaces 

 in pages 73 and iiS, were reduced to millimetres distance 

 and plotted down in the same way, but not on precisely 

 the same scale as the preceding curves for constant 

 distance and various pressures. Hyperbolic curves were 

 also found which intersected the experimental curves in 

 two points. 



It was seen in the case of spherical surfaces, the result 

 h.iving been obtained as the average of a great number 

 of experiments, that the hyperbola coincided closely with 

 the observations, while for plane surfaces, for which only 

 a few experiments were made, the coincidences were not 

 quite so perfect. Xcvcrthelcss, it would appear that the 



law of the hyperbola holds ecjually well for a constant 

 pressure and varying distance as it does for a constant 

 distance and varying pressure ; the obstacle in the way 

 of a discharge being up to a certain point as the number 

 of molecules intervening between the terminals.' 



In the two cases of spherical and plane surfaces the 

 ratio between the transverse (distance) and conjugate 

 (potential) axes of the respective hyperbolas was — 



For spherical surfaces 

 ,, disks 



I '240 

 1-285 



With the data already published in Part I., the authors 

 have laid down a fresh curve for the striking distance 



■ Dr. Alexander Macf.irhne h.is published in the Transacfwns of ihe 

 Royal Society of Edinburgh, 1E78, vol. xxvii., an elaborate and cnreful 

 research of the "Disruptive Discharge of Electricity " in air and different 

 ga'^es, and between terminals of various forms. An .abstract of this p.aper 

 will be found in Natuke, vol. ,\i.v. pp. 184, 185. Dr. Macfarlane used n 

 Holtz machine and employed higher potentials than those we used : hefj' nd 

 that the results for the discharge between two disks 4 inches in diameter at 

 various distances up to 1*2 centims. and with various pressures were s.Ttis- 

 factordy represented by ihe hyperbola. 



