L2 Prof. E. Edlund on the Electrical 



korff induction-apparatus, that the quantity of heat evolved in 

 a Geissler tube is proportional to the quantity of electricity 

 that has passed through the tube, and is almost independent 

 of the diameter of the tube. The quantity of heat evolved at 

 the negative electrode was in the same way proportional to the 

 quantity of electricity which had passed through the tube, and 

 several times as great as at the positive electrode ; at this 

 latter the quantity of heat evolved was so very small that it 

 was impossible to determine distinctly its proportionality to 

 the amount of electricity. 



From the observations above cited we can already deduce 

 some results which will find their confirmation in the obser- 

 vations expounded below. When the electric current passes 

 through rarefied gas, the gas is heated. This could not take 

 place unless the gas opposed an obstacle to the electric motion. 

 The current is forced to consume a certain amount of work to 

 overcome this obstacle ; and it is the work consumed that is 

 transformed into heat. We must admit, on the ground of this 

 observation merely, that the gas exerts a resistance to the pro- 

 pagation of the electricity ; it is, besides, of no importance 

 whatever for our consideration to know the nature of that 

 resistance, whether it is or is not homogeneous with that which 

 occurs among solid and liquid bodies. The observations above- 

 mentioned sIioav that the quantity of heat evolved may be 

 regarded as proportional to the intensity of the current, and 

 not to the square of that intensity as Joule's law would require 

 it to be. It follows directly from this, that, all other circum- 

 stances remaining equal, the amount of heat evolved must be 

 independent of the section of the tube containing the gas. 



Suppose, for example, two tubes of equal length, filled with 

 the same gas, the section of one tube being n times as large 

 as that of the other, and the same current s passing through 



both. Then - of the current passes through each - of section 



n l " n 



of the wide tube, whence it follows that the evolution of heat 



in each equal part is proportional to -, and consequently that 



the heat evolved in all the parts united is proportional to s 

 ( that is to say, of an amount equal to that in the narrow tube). 

 Of course, moreover, cceteris paribus the evolution must in- 

 crease with the length of the tube, and be in general proportional 

 to that length. If r x denote the specific resistance in a column 

 of gas of unit length, the heat evolved in I units of length will 



