168 Prof. G. Wiedemann on the Laws of the 



metals attract positive more strongly than negative electricity, 

 and therefore that, if charged with an equal quantity of electri- 

 city, the negative electrode obtains upon its surface the tension 

 necessary for the commencement of a discharge sooner than 

 the positive. After the resistance has been overcome, the 

 column of gas between the electrodes should, on the hy- 

 pothesis now under discussion, conduct the electricity like a 

 metal. But we find that, with similar electrodes and similar 

 gas at various degrees of pressure, the quantity of electricity 

 requisite for a discharge is at first independent of the length 

 of the distance between the electrodes — as for instance in 

 capillary tubes when their length is so great that the influence 

 of the electrodes upon each other and that of the free electri- 

 city accumulated in the tube may be neglected. Again, in 

 capillary tubes of different lengths the distribution of 

 electricity upon the charged electrodes before the discharge 

 must be equal, and therefore also the resistance that has to be 

 overcome at the electrodes must be similarly equal. In fact, 

 if we put v for the potential energy which corresponds to the 

 electrical charge immediately before the discharge, a for the 

 action effected in overcoming the resistance, b for the action 

 effected in the " metallically conducting gas " upon a unit of 

 length in the shape of production of heat, and I for the dis- 

 tance between the electrodes (the wider spaces in the neigh- 

 bourhood of the electrodes being replaced by a suitable length 

 of capillary tube), the result will be 



v—a-\-bl, or b={y—a) : I. 



Thus the action effected, or the heat produced in the entire 

 length of the capillary tube would be inversely proportional 

 to the length of the tube. But experiments show that the 

 quantity of heat is, within certain limits at least, nearly inde- 

 pendent of the length of the tube. 



We find a similar contradiction if we measure the heat 

 produced by discharges under conditions altogether similar, 

 except the diameter of the tubes. 



If the gas be heated like a metallic conductor, we must 

 suppose that the whole heat produced in the gas by the dis- 

 charge passes over to the capillary tubes. Accordingly the in- 

 crease of temperature in a tube of diameter 4 would be only 

 one fourth of that in one of a diameter 1, if the gas in both 

 tubes were equally good as metallic conductors. But as a 

 matter of fact the increase of temperature is in both cases 

 almost the same. If we suppose the conductive power of the 

 gas to be itself dependent upon its temperature, we must first 

 explain, since the gas after each single discharge becomes 



