182 Proceedings of hidtcnia Academy of Science 



rods in contact it is seen (figure ?>) that the space of closest contact is 

 one of high potential gradient. This device of using glass rods in an 

 electric field has been tried out at Purdue by K. B. McEachron', who 

 has obtained some high yields of ozone. 



Similar reasoning would show that a sheet of dielectric material 

 between two parallel electrodes would prodi^ce no distortion of the field. 

 This theory has been tested experimentally by C. W. Rice^ using glass 

 cylinders with experimental conditions such that there were no "end 

 effects". It was found that the presence of the cylinder always lowered 

 the arc-over voltage. The condition of the surface and especially the 

 nature of the gar, or vapor adsorbcl en it were found to have impor- 

 tant effects. Moisture is always very potent in aiding the arc-over'. 

 These observations led to the idea that the mechanism of this promo- 

 tion of arc-over was catalytic in nature. Just as .solid surfaces pro- 

 mote chemical reactivity by adsorbing the reacting substances and 

 activating them by loosenirg their bonds S3 here the adsorbed molecules 



Via. 1. The linos of fnicc 

 in an electric field are flistoitcd 

 by a rod or .sphei-c of dielectric 

 material. 



Fi.u'. ''). Two such bodies in contact 

 distoi-t the lines of force as shown. 

 The maximum discharjje in the iras 

 occurs in t'lc shaded areas. 



in an activated condition would be much more easily ionized and woukl 

 aid materially in establishing a path of discharge". 



For this catalytic action of promoting creepage along the .surface 

 the name, "creepage corona" is proposed. In any discharge involving- 

 dielectrics there will be two effects, that of the specific inductive capacity 

 in increasing the discharge density and the creepage action in increasing 

 the discharge current. In addition, the purely catalytic effect on any 

 chemical reactions of the solid-gas interfaces will produce effects which 



<>. 



-)8-108 



'McEachron and GeoiKe. Purdue Univ. Enj;. Exp. Sta. Hul. N< 

 (1022). 



*C. W. Rice. Trans. Am. Inst. Elec. En^'. 36, 1947 (1917). 



•'■ This action of moisture is due to the polarity of the water molecule. Ueplacinsi 

 the polar water with nonpolar oil lowers the cieepa^c effect and raises the arc-over 

 potential. 



«G. V. Hevesy. Z. Physik. to, 80-:! (1922). who found that small crystalline 

 conglamerates had fifty times the conductivity of laiue crystals, explains the rcsu'ts as 

 due to "loosened spots" in the crystal lattice which would I'esult in an irrejrularity in the 

 order of the ions especially at the surfaces. H. S. Taylor. Chem. Ajrc. .id, :i09 ; .J. 

 Franklin Inst. J.9.;, 1 (1922), explains the catalytic action of reduced copper on the 

 reduction of cojiper oxide at the point of contact of the two phases, as due to a distortion 

 of the adsorbed reducins: gas when in contact with the different electronic configurations 

 in the two phases. 



