VELOCITIES. 97 



and the strips of tin (Fig. 12). We may also mention the 

 absence of detonation in certain grains of fulminate which had 

 been slightly greased by accident, a circumstance which retarded 

 the heating without otherwise modifying the explosive property. 

 Again, when the flame is extinguished in its course, as we 

 observed with the capillary glass tube, the furthest registering 

 apparatus remains intact. We could not however feel assured 

 on this point until we succeeded in reproducing our experiments 

 and obtaining the same velocities in a tube of caoutchouc, a 

 substance which could not be suspected of propagating the 

 vibratory movement like metals. This was rendered possible 

 by the fact that the internal combustion of the gaseous mixture 

 is so rapid that it does not affect the material of which the tube 

 is made. This caoutchouc tube was 40109 metres long, and 

 several mms. thick, its external diameter being 5 mms., and it 

 was capable of supporting either a vacuum or an interior pressure 

 of several atmospheres without any appreciable deformation. 

 It was fixed upon the frame already described, in parallel lines 

 (Kg- 9). 



These were the results : 



Velocity per second. 



2685 metres. 

 2911 

 2994 , 

 2672 



2788 



Mean ... 2810 ,, 



This mean agrees with the value, 2841, obtained with the 

 leaden tube, within the limits of error. 



The propagation of the explosive phenomenon is thus inde- 

 pendent of the composition of the tube, provided that the 

 internal diameter remains the same. 



We now come to some experiments made with a system of 

 glass tubes, the total length of which was 43*24 metres, but the 

 mean internal diameter only '0015 metre. These were capillary 

 tubes, each 2 metres in length, connected end to end by means of 

 caoutchouc tubes, the whole being fixed upon the frame before 

 mentioned (p. 91). The bends were made with the same glass 

 tubing. We found velocity per second, 2403 and 2279 metres ; 

 mean, 2341 metres. 



These figures are rather lower than the foregoing ones, no 

 doubt owing to the difference in the diameter ; the propagation 

 of the explosion being impeded in a capillary tube, as is the 

 case also with the propagation of sound. The experiments 

 made in glass enable us to see the propagation of the flame. 

 Working in darkness, we see the entire length of the tube 

 lighted up at the same moment, the eye not being able to 

 perceive the progress of the flame. 



H 



