492 Prof. R. Bunsen on the Temperature of the 



gas which takes no part in the combustion, the celerity of the 

 ignition becomes more and more diminished, until, when the 

 mixture is on the point of burning, the progress of the ignition 

 can be followed with the naked eye. If therefore the combustion 

 is excessively slow, owing to the presence of foreign gases, the 

 measurement of pressure is rendered completely erroneous, since 

 the first portion of gas which is ignited is cooled down again 

 before the remainder has become heated. In order to remove 

 this source of error, it is necessary to avoid too great a dilution 

 with foreign gases, to have the diameter of the column of gas to 

 be burned not too great, and to cause the inflammation of the 

 gases to be as instantaneous as possible. This last is best ac- 

 complished by sending a very powerful spark from a large Ruhm- 

 korff 's coil-machine through the entire length of the column of 

 gases. The exploding- vessel (fig. 1, Plate IV.) which I employed 

 was made of exceedingly thick glass, and contained a column of 

 gas 8*15 centims. in height and Wcentim. in diameter, so that the 

 entire length which the combustion had to travel was 8" 5 millims., 

 the distance from the path of the spark to the glass walls. The 

 shorter the time in which the maximum temperature is reached 

 in proportion to the duration of the cooling, the less is this source 

 of error to be feared. It is necessary here to say a word on the 

 celerity with which the propagation of the ignition takes place. 

 This is estimated as follows : the explosive mixture is allowed to 

 burn from a fine orifice of known diameter, and the rate of the 

 current of gas carefully regulated by diminishing the pressure^ 

 to the point at which the flame passes back through the orifice 

 and ignites the gas below it. This passing back of the flame 

 occurs when the velocity with which the gaseous mixture issues 

 from the orifice is inappreciably less than the velocity with which 

 the inflammation of the upper layers of burning gas is propa- 

 gated to the lower and unignited layers. It is then only neces- 

 sary to measure directly the rate of the current of incombustible 

 gas at the pressure under which the flame went back, in order to 

 obtain with approximate exactness the rate of propagation of the 

 ignition to which this rate of the current is equal. If we call 

 the rate of propagation c, the diameter of the orifice in the thin 

 plate from which the explosive gas issues d, and the volume of 

 gas which streams in t seconds under each pressure from the 

 orifice V, then 



___4V 

 C ~ irdH 



An experiment with pure hydrogen explosive mixture gave the 

 following numbers : — 



V—963000 cubic millions.; ^=1'2 millim. ; t=25 seconds. 



