McChKi.LAND AND GriLL — Sclf-Ljnition of Ether- Air Mixtures. Ill 



B. Experiments with air or other gas without etlier vapour witlr the 



object of determining the distribution of temperatures by 

 means of a thermo-couple at different parts of the apparatus. 



C. A determination of the actual temperature reached by air under 



the conditions of the ether-air experiments. 



A. — General Effexts Observed. 



Provided that there was a sufficient quantity of ether present, the exact 

 proportions of air and ether in the reservoir did not appear to be of impor- 

 tance. 'J'he ordinary ether supplied to the laboratory was used without any 

 special purification. The humidity of the mixture did not appear to interfere 

 with the result of the experiments. Explosions were obtained both when the 

 reservoir was perfectly dry and wheir it had just been washed with water 

 inside. The presence of dust had apparently no special significance. 



The pressure of the mixture in the reservoir could be varied between 

 fairly wide limits without interfering with the result of the experiments. In 

 our experiments the pressure was usually atmospheric. 



The influence of the pressure of the residual air in the exhausted 

 explosion-tube was studied in some detail. Ignition was obtained at the 

 lowest pressures possible in our apparatus. When the pressure exceeded 

 5 cm. of mercury, iro ignition was observed. The most suitable pressure 

 seemed to be about 3 or 4 cm. in the explosion-tube when the pressure of 

 the mixture in the reservoir was atmospheric. Considerable difficulty was 

 experienced in obtaining very low pressures in the explosion-tube, as it was 

 difficult to obtain a large tap which was at the same time perfectly air-tight. 

 By employing the following device, a low pressure was obtained ; — Two large 

 taps, as air-^ight as possible, were joined in series, with a short space between 

 them. The tap near the explosion-tube was surrounded by a bath of thick 

 oil, so that no leak could take place into the explosion-tube except through 

 the space between the taps. This space was kept exhausted by means of an 

 air-pump until just before the experiment, when the tap near the reservoir 

 was opened. By this means, using a mercury air-pump, the pressure in the 

 explosion-tube was reduced as low as O'OOS mm. 



The length of the connecting tube did not seem to have any marked 

 influence on the ignition. Tubes of varying lengths were employed with 

 success. It is prudent to have the tube of sufficient length to ensure that the 

 flame will not reach the reservoir. 



The length of the explosion-tube is a matter of importance. Ignitions 

 or explosions were obtained in tubes varying from somewhat less than 

 three feet to eight feet. Longer tubes would probably have served as well. 



o2 



