APPENDIX. 549 



This difference is apparently due to the greater viscosity of the 

 liquid at lower temperatures. 



Average velocities of 1310 m., 1015 m., and 1286 m. were 

 observed in lead, Britannia metal, and tin tubes 3 mm. in diameter. 

 A Britannia metal tube 9 mm. in diameter gave 1386 m. Abel 

 found 1672 m. under slightly different conditions. 



VII. DYNAMITE. 



Velocities of 2333 m. and 2753 m. were observed in Britannia 

 metal tubes 3 mm. internal diameter, while in tubes of the same 

 metal or lead 6 mm. diameter the average velocity was 2668 m. 



Abel found 5928 m. to 6566 m. for a train of dynamite cartridges, 

 30 mm. in diameter, placed end to end and fired in the open air. 

 These much higher velocities are no doubt due to the much 

 greater diameter of the explosive cylinders. 



VIII. PANCLASTITE. 



Owing to the extreme volatile nature of this mixture, bubbles of 

 gas formed in the interior of the tube, and caused irregularity in 

 the results. A mixture of equal parts of liquid nitric peroxide and 

 carbon disulphide, contained in a leaden tube 3 mm. internal 

 diameter, gave 4685 m. velocity ; another similar experiment gave 

 5470 m. in the first half of the tube, and 6658 m. for the total 

 length. On the whole, these figures are similar to those found for 

 gun-cotton. 



To sum up, principally from the experiments made with gun- 

 cotton 



The velocity increases with the density of charge. It increases 

 with the diameter, at least within the limits of the very narrow 

 tubes experimented with. 



It appears to increase with the resistance of the envelope (the 

 latter being pulverised by the explosion). 



Finally, comparative measurements made with a tube of 200 mm. 

 very much curved, and a similar but straight tube, gave practically 

 the same velocity. 



These experiments should be regarded as applicable to practical 

 conditions comparable to those under which they were made, 

 although the indications of the correlation between the velocity 

 and the density of charge or the resistance of the envelopes 

 appear conformable to theory. 



To further develop this study, experiments were made with a 

 homogeneous and very mobile liquid explosive, methyl nitrate 

 (p. 420), contained in tubes of caoutchouc, glass of different thick- 

 nesses, Britannia metal, and steel. 



The details of these experiments are to be found in a com- 

 munication on the " Explosive Wave," by M. Berthelot, " Annales 

 de Chimie etde Physique," 6* serie, torn, xxiii. pp. 485-503 (1891). 



1. Canvas-covered caoutchouc tubes. The tube had an internal 



