598 SCIENTIFIC RECORD FOR 1885. 



the inanuer lu which the iuflainmation is produced regulates geuerallj 

 the mode of coinbustioii, flame producing deflagration and shock de- 

 termining explosion; while for gaseous bodies deflagration changes 

 spontaneously into explosion, at least for certain mixtures where the 

 propagation is very rapid. There is, moreover, for gases a mode of 

 vibratory propagation very variable and very irregular, which is, in a 

 certain sense, intermediate between deflagration and explosion. These 

 plicnomeua the authors explain by supposing that the slow deflagration 

 corresponds to the propagation of the temperature of inflammation by 

 conduction ; while on the other hand the explosive wave is produced at 

 the instant when the pressure produced by the ignited layer upon the 

 following one is sufQcieut to raise this latter to the point of inflamma- 

 tion. From this moment the inflammation is itself propagated with the 

 speed of transmission of this pressure, *. e., with the speed of sound, 

 accelerated, of course, by the high temperature produced. The prog- 

 ress of the slow wave was studied by means of photography. A glass 

 tube containing the gaseous mixture — generally a highly photogenic 

 mixture of carbon disulphide and oxygen, or nitrogen dioxide — was 

 placed near a revolving cylinder and parallel to its axis. As the flame 

 progressed along the tube it developed a curve on the sensitive paper 

 with which the cylinder was covered, the abscissas of which gave the 

 place of the flame and the ordinates the time required for it to reach 

 this point. The displacement of the flame along the tube is perfectly 

 uniform and constant for the same mixture, being for one volume car- 

 bon disulphide vapor and three of nitrogen dioxide 1-25 meters per 

 second. The temperatures of combustion are given in tabular form. 

 {J. Phys.. February, 1885, IT, iv, 59.) 



In their researches on the speed of the explosive wave in gases, Ber- 

 thelot and Vieille have shown that this wave is propagated with a speed 

 approximately equal to the mean speed of translation of the molecules 

 of the gaseous products of combustion, upon the hypothesis that all 

 the heat involved in the reaction is found for the first instant in the 

 compounds formed. Thus, for electrolytic gas, the mean of several de- 

 terminations gave a speed of 2,810 meters per second, the mean speed 

 of molecular translation for the vapor formed being 2,831 meters. But 

 this law did not appear to hold for mixtures of carbon monoxide with 

 oxygen or with nitrogen monoxide. Dixon has now pointed out the- 

 fact that water- vapor is necessary to the combustion of carbon mon- 

 oxide, and has shown that as the proportion of vapor augments the 

 speed of inflammation increases also. From experiments made in a 

 lead tube 55 meters long and 13 millimeters in diameter, he found the 

 speed of the explosive wave to be 2,817 meters at 10° C., confirming 

 the above results. -With carbon monoxide and oxygen nearly dry, the 

 explosive wave was not established until the flame had traversed a dis- 

 tance of 700 millimeters from the firing point. The speed observed was 

 rather over 1,500 meters per second. After tbe explosion a fine layer 



