RECENT PROGRESS IN PHYSICS. 



369 



psitive. This is exhibited in the following table, drawn from the 

 ixmer experiments. The range was — 



In air 



oxygen 



nitrogen 



hydrogen 



carbonic acid 



olefiant gas 



coal gas 



muriatic acid gas. 



s and S. 



[Althoiigh, as Faraday himself remarks, these numbers require con- 

 sierable correction,, the general result is striking and the differences 

 i several cases very great. 

 It appears clearly from these experiments that different gases have 

 Bt equal capacities for insulation. Considering the mean values of 

 t (for positive charges of * and S,) we perceive that a stratum 0.62 of 

 a inch of — 



Oxygen 

 Nitrogen 

 Hydrogen 

 Carbonic acid 

 Olefiant gas 

 Coal gas 

 Muriatic acid gras 



insulates as well as a stratum 

 of air, whose thicknes is 



0.505 

 0.615 

 0.370 

 0.640 

 0.750 

 0.490 

 1.105 



tat is, an electrical discharge passes as easily through a stratum of 

 ar 0.370 of an inch thick, as through one of hydrogen of 0.62 of an 

 ich ; an electrical spark penetrates a stratum of air 1. 105 inch thick as 

 elsily as one of 0.62 of an inch of muriatic acid gas; an electrical 

 s|ark passes with decidedly more ease through oxygen, hydrogen, and 

 cal gas than through an equal stratum of air ; but muriatic acid 

 §s and olefiant gas present a decidedly greater resistance to the trans- 

 Eission of the spark than an equal thickness of air does. 



iSimilar results were obtained from later but less reliable experi- 

 rents.— (Pog. Ann., XLYIII, 281.) 



iThe mean values of w are not equal with positive and negative 



trges of S and s ; for many gases u has a greater mean value with 

 ositive charge of S and s than with a negative; for other gases it 

 i<the reverse; but to draw the conclusion from these experiments that 

 r|iny gases more readily allow the negative and others the positive 

 (^charge through them, seems to me unwarranted, because the differ- 

 , elces of the above table in this respect are within the limit of errors 



q observation. 

 . fThus, according to the table given above, the mean value of w with 

 , positive charge of s and ^ is, for — 



' 24 s 



