112 Scientific Procccdinga^ Royal t)ublin Society. 



No explosion or ignition was observed in tubes less than about two feet long. 

 The diameters of tubes in which the effect was observed varied from two to 

 somewhat more than three inches. Tubes of about one inch in diameter 

 were not suitable. 



Tlie nature of the end of the explosion where the ignition began liad a 

 marked effect on the result of experiments. A fiad of cotton wool, about 

 an inch and a half thick, prevented ignition. A bulb, three inches in 

 diameter, was fixed into a hole of the end plate of the explosion-tube. The 

 flame began in the bulb, but did not travel up the tube. The explosion- 

 tube was replaced by a bell-jar of about ten inches diameter, and having 

 appi'oximately the same volume as the tube, but no ignition was obtained in 

 the jar. 



Experiments were made to find out what effect the presence of obstacles 

 in the explosion-tube might exert. Ignition took place when a pad of wire 

 gauze, up to six layers thick, was placed across the middle of the tube. A 

 pad ten layers thick prevented ignition. 



B. — Temperature Experiments. 



As the cause of the ignition seemed to be due to the heating of the ether- 

 air mixture, it seemed useful to examine the rise in temperature of pure air 

 when used in the place of ether-air mixtures. The conditions under which 

 these experiments were made were identical with those in the case of tlie 

 experiments described. 



Small pieces of phosphorus fixed axially in the explosion-tube were 

 ignited at distances from the closed end varying from a few millimeters to a 

 few centimeters. At distances greater than about 5 cm. the phosphorus was 

 not ignited. 



Experiments were then undertaken witli the object of studying the dis- 

 tribution of temperatures in the different parts of the explosion-tube by 

 means of a thermo-couple. The thermo-junction consisted of thin copper 

 and constautan wires soldered together and to thicker leads of the same 

 materials. The leads passed througli a glass tube, which, in turn, passed 

 through a small metal tube fixed on to the metal disc closing the end of the 

 tube. By means of a clamp the junction could be placed at any required 

 distance from the end of the tube. The aperture through which the tube 

 passed was kept air-tight by means of plasticine. In all cases the junction 

 was placed axially in the tube. The numbers on which the curves are drawn 

 refer to the deflections of a ballistic galvanometer in series with the couple. 

 The swing was reduced by putting resistances of from 500 to 700 ohms in 

 series with the galvanometer. The ditlereut curves serve to indicate the 



