442 M. F. L. Ekman on the Electrical Insulating -power and 



glasses employed. After the glasses had been well cleaned, two 

 narrow strips of tinfoil coating were fixed parallel to each other 

 on one side of each glass. The one strip which was intended 

 to be charged with free electricity was fastened as far as possible 

 from the edges of the glass, the other was placed in communica- 

 tion with the earth. These tinfoil slips were all of the same size, 

 namely 43 millims. long and 8 millims. broad, with the comers 

 rounded off; the distance between them was 48 millims. This 

 coating having been fixed, the glasses were warmed and covered 

 with shellac varnish, leaving, however, the tinfoil slips and the 

 glass surface between them bare. 



The glasses being thus prepared and the lower coating of each 

 connected with the earth, the upper coating was charged from a 

 Ley den jar, to which the same quantity of positive electricity 

 was always communicated by means of a Lane's electrometer. 

 The insulating quality of the glass was now measured by the 

 length of time during which the electricity remained in the 

 upper coating. As the last traces of electricity disappear very 

 slowly from well-insulating kinds of glass, I considered it suffi- 

 cient to observe how long the charge in the upper coating con- 

 tinued sufficiently strong fully to deflect the gold-leaf in a 

 Bohnenberger's electroscope on the transfer of the first charge 

 from the coating to the electroscope. This test, though it can- 

 not make pretensions to extreme accuracy, is yet abundantly suf- 

 ficient to establish the order of precedence in which the different 

 kinds of glass must be classed as insulators. The relative mois- 

 ture of the air varied during the experiments from 47 to 52 per 

 cent., at a temperature of from 15° to 18° C, and may therefore 

 for these experiments be regarded as sufficiently constant. 



The times of insulation thus determined for the different 

 glasses were found on an average to be — 



Glass No. 1. ... At least 6 hours. 



„ No. 2. . . . About 51 minutes. 



„ No. 3. ... „ 8i „ 



„ No. 4. . . . „ 50 seconds. 



No. 5. . . . „ 12 



» 



The differences were, as we see, strikingly great, and tend on 

 the whole in the direction which one, on chemical grounds, would 

 be led to expect. But here there is a remarkable exception. 

 Specimen No. 2 had proved less hygroscopic and less susceptible 

 to the action of carbonic acid than No. 1, and yet is far surpassed 

 by the latter in insulating-power. 



It now appeared to me worth while to try whether any of the 

 badly insulating kinds of glass could be improved by the re- 

 moval of the old surface of the glass. I tried this experiment 



