TRANSACTIONS OF SECTION B. 685 



They attached to a leaden coil of two litres capacity a strong glass tube of 

 10 to 25 cubic centimetres capacity, closed at one end, and separated at its other 

 end from the coil by a gas-tight steel tap, the bore of which had the same diameter 

 as the glass tube. The glass gauge had previously been calibrated and graduated. 

 The coU was then filled with the explosive mixture, while the glass gauge i to- 

 gether with the bore of the tap was filled with air or hydrogen. The instant before 

 the mixture was fired the tap separating the coil from the gauge was opened, and 

 the flash due to the exploding gases darted into the gauge. The experiment con- 

 sisted in taking a photograph upon a sensitive plate of the flash in the gauge. It 

 was found possible to obtain photographs which showed the flash to come to a 

 well-defined and abrupt termination, marking the volume into which the air in 

 the gauo-e had been compressed. By this means the authors hold that the air in 

 the gauge is compressed adiabatically, and with the data aflbrded by the photograph 

 they'had originally intended to calculate the pressure produced in the explosion by 

 Ilankine's well-known formula 



P' = P 



(;)■ 



The assumption was therefore made that there is no loss of heat up to the time 

 that the flash attains its furthest limit. To obtain some data to justify this 

 assumption the authors took a photograph of the flash upon a swinging plate 

 attached to a pendulum chronograph. The plate moved past the flash at the rate 

 of 2i- metres per second. The photographs so obtained could scarcely be dis- 

 tinguished from those taken upon a stationary plate ; they were perfectly symme- 

 trical ; only the slightest possible increase in breadth could be detected. These 

 photographs showed that the compression is approximately instantaneous, and that 

 the flash ceases to have any sensible photo-chemical action in an interval of time 

 less than one five-thousandth of a second. In so short an interval of time the 

 compressed gas cannot lose a sensible quantity of heat. It would appear, therefore, 

 rio-ht to assume that the compression is adiabatic. The authors believe also that 

 no appreciable diffusion of the compressed gas into the compressing gas can occur 

 during the very brief time of compression. 



Further, since thermo-dynamic theory proves the ratio of the two specific heats 

 of any perfect gas to be approximately constant for all temperatures and pressures, 

 Eankine's formula may be employed' to calculate the pressure produced by the 

 explosion when the gauge is filled with air or hydrogen. 



Essentially, therefore, the method gives a record of the adiabatic compression 

 of a gas by the explosion, a piston of indefinitely small thickness and mass being 

 interposed between the compressed and compressing gases. 



Experiments have been made with gauges of different volume, varying from 



11-55 cubic centimetres to 23-20 cubic centimetres, and the ratio y,^ is found to be 

 independent within these limits of the volume of the gauge. Further, the ratio 



X is found to be the same, within the error of experiment, whether air or hydro- 

 V 



o-en is the gas compressed. The photographs obtained, however, in the case of 

 hydrogen show the flash to be in general iU-defined and pointed, owing probably 

 to the°high thermal conductivity of hydrogen, combined with a much greater inter- 

 diffusion with the compressing gases in consequence of the greater molecular 

 velocity of hydrogen in comparison with that of air. 



The mean value from six diflerent sets of experiments with air and hydrogen, 



electrolytic gas being the explosive mixture, gave — = 8-6. Calculating P' by 



means of Rankine's formula, and taking k = 1-408, it is found that P' = 20-7 

 atmospheres. This pressure is about twice that measured by former experimenters. 

 Some experiments were also made with carbonic acid gas ; in these the explosion 

 compressed the gas into a much smaller volume, owing to the great variation of 

 the two specific heats of carbonic acid gas -s^itb increase of temperature and 

 pressure. 



