1915] on Gaseous Explosions 279 



within the walls, and the flame reaches it much later than it touches 

 B. D also, although near the w\alls, has only been in contact with 

 the flame for a very short period before maximum, so that some 

 other explanation than heat loss must be found to account for the 

 great temperature differences. The temperature in the side thermo- 

 meter, 850° C, is easily accounted for by heat loss, as the flame 

 touches the side long before it reaches the ends, the diameter being 

 23- 4 inches, w^hile the length is 28-75 inches. Hopkinson accounts 

 for the differences in the following manner : — When the flame 

 strikes the thermometer B it is burning nearly at atmospheric 

 pressure. The volume ignited at first does not disturb the pressure 

 sufficiently to require consideration. As the flame progresses the 

 pressure increases to 6| atmospheres absolute. The flame temperature 

 therefore of about 1200^ C, which is attained at atmospheric pressure, 

 is increased by the compression towards the centre as pressure rises, 

 so that so far as the centre thermometer is concerned, the gas is 

 treated as if a cylinder had been filled with flame at 1200° C. and 

 then rapidly compressed to 6 J atmospheres pressure. This compression 

 accounts for the rise of temperature above the temperature due to 

 combustion ; that is, flame has been produced first, then it has been 

 adiabatically compressed from about 1200° C. to 1900° C. At the 

 point C the same reasoning applies, but the range of compression is 

 somew^hat less, so the temperature only rises to 1700° C. At the 

 point D the gases are compressed, as in a gas engine cylinder, to 6 

 atmospheres before ignition occurs ; and, consequently, the subsequent 

 rise of half an atmosphere is not sufficient to appreciably disturb the 

 temperature. The temperature there is therefore between 1100 and 

 1300° C, simply the temperature attained by explosion without 

 subsequent compression. At the side thermometer the flame reaches 

 the walls when the pressure in the vessel is about 2 atmospheres. 

 The temperature then instantly rises to 1200° C. or 1300° C, and at 

 once begins to fall, so that by the time maximum temperature has 

 been attained it has fallen to 850° C. 



Half a second after maximum pressure, the distribution of 

 temperature is very different ; convection has now had time to take 

 effect. The distribution of temperature is broadly as follows — 



Degrees C. 

 Mean temperature inferred from pressure . . . 1100 



Mean temperature, exclusive of layer 1 cm. thick at 

 walls, determined by long platinum wire from 



B to D 1160 



Temperature at centre of vessel, thermometer B . 1100 to 1200 



In this experiment the thermometer D has been dispensed with, 

 and a thermometer consisting of a long platinum wire stretching from 

 B to D has been used to get the mean temperature between these 

 points. Another thermometer B has also been used at the centre. 



