Hydrogen ivith Oxygen at Low Pressures. 



471 



which corresponds to the lowest pressure is, however, distinctly 

 higher. All these temperatures are, I believe, well below 

 the temperature o£ redness ; and in view of the large variations 

 of the pressure, it may be said that 9 is, roughly, inde- 

 pendent o£ the pressure. 



Before making the series of observations from which 

 Table I. was constructed, I had made similar experiments of 

 precisely the same nature with a wire of similar dimensions, 

 but of commercial platinum. It measured about 23 cms. in 

 length and '105 mm. in diameter. Its temperature-rate of 

 increase of resistance was only '00185, just half that of the 

 pure platinum wire. It was thoroughly cleaned as the 

 former one had been, and was fastened into glass tubing 

 in exactly the same way which is described above (see 

 fig. 1), except that the connexions (P,Q in the figure) of the 

 wire with the pieces fused into the glass were not completed 

 by solder. 



Instead of the solder, a spring contact was made at the 



point Q, a little 

 This contact, althom 



coil in the wire serving as the spring, 

 jh precarious, must have answered very 

 well, for its equivalent resistance evidently, from Table II., 

 remained fairly constant. 



In Table II., p has the same meaning as before, but R is 

 the resistance of the wire together with the resistance of its' 

 supports, when the wire was just sufficiently heated to start 

 the combination of the gases. The same joint resistance 

 when the temperature of the wire was 18° C. was found to be 

 5*37. The resistance of the wire alone was nearly 5'1 ohms. 

 Accordingly, if 6 is the temperature of the wire corresponding 

 to R> 



0= 18 + 110 (R-5-37). 



Table II. 



, 



K, 



9. 



31-2 



7*37 



238 



29 



754 



257 



11-8 



7-51 



254 



7 



723 



223 



4-4 



7-37 



238 



4 



7-37 



238 



3-9 



7-4 



242 



2-02 



7-47 



250 



Owing to the defective contact in this apparatus, to which 

 I have alluded, the calculated values of 6 in Table II. cannot 

 be considered nearly so accurate as the corresponding 



