under the Influence of the Electric Discharge. 373 



They were joined to each other by fusion, placed horizontally, 

 and calorimeters pushed over them as shown in fig. 4. The 

 narrower tube had a diameter of 0*6 millim., and the portion 

 surrounded by the calorimeter a length of 42 millims. ; for 

 the wider tube the corresponding numbers were 2 millims. 

 and 45 millims. The ratio of the sectional areas of the tubes 

 was therefore 1:10; the water-equivalent was nearly the same 

 for the two calorimeters. 



The following table gives, in the usual way, the amounts 

 of heat observed in the two calorimeters, which were always 

 simultaneously observed : — 



Hydrogen. 



P 



Air. 







P 



F= 

 wide 



=0 

 narrow 



F= 

 wide 



= 10 

 narrow 



F=0 



F= 



=10 



32-8 



0-85 



0-96 







21-5 



0-74 



065 



0-82 



0-60 



21 



0-66 



0-88 



0-66 



077 



8-1 



034 



044 



_ 



047 



84 



0-30 



0-48 



0-38 



0-44 



3-4 



018 



0-24 



0-36 



033 



32 



015 



0-20 



0-28 



0-32 













06 



006 



017 



0-12 



024 













X 



007 



0-24 



0-19 



0-32 













The tables show (1) that, in agreement with the earlier 

 results, the difference of heating with and without spark is not 

 very great, both in wide and in narrow tubes ; that, further, 

 the amounts of heat produced in the tubes of different width, 

 with the same length of spark, do not differ very greatly. In 

 both cases, however, at low pressures there appear considerable 

 deviations from these laws. In these, as in the former obser- 

 vations, the amount of heat increases perceptibly when the 

 spark is included ; and, moreover, the heating in the narrow 

 tube is decidedly greater than in the wide one. Further 

 insight into these relationships can only be obtained by ad- 

 ditional experiments. 



A further series of experiments was occupied w r ith the 

 behaviour of still wider tubes. 



Two tubes, a capillary tube and a wider tube, of diameters 

 0*5 millim. and 4 millims., and thickness of glass 2 millims. 

 and 1 millim., were united and included in the circuit, the 

 calorimeters filled with turpentine, and the amounts of heat 

 determined. The values obtained are given in the following 

 tables. Those under A relate to the case in which the 

 electrode next the capillary tube was put to earth, those under 

 B to that in which the electrode next the wide tube was in 

 connexion with the earth. The signs + and — mark the 



