176 Mr. R. Threlfall on the 



presence of these vibrations. The first set of experiments 

 bearing on this point have been already discussed, with the 

 result that the hypothesis is perhaps unnecessary. There are, 

 however, a great number of other experiments, some of 

 which cannot be so easily explained. In one case an explo- 

 sion was induced, in a charge of fulminate of silver placed at 

 the end of a tube, by the explosion of a, similar charge at the 

 other end. This effect was not interfered with by placing 

 diaphragms across the tube; but the state of the internal 

 surface of the tube seemed to exercise considerable influence. 

 Experiments were also made on the action of fulminate of 

 mercury on gun-cotton through tubes (Proc. Roy. Soc. 1874, 

 vol. xxii. p. 160). The great influence exerted on the deto- 

 nating power by the smoothness or roughness of the walls of 

 the tube seems a strong argument against the supposed 

 synchronism having much to do with the effect in these 

 cases. On the other hand, it is just what we should expect if 

 there was a bodily motion of air down the tube, or even if, as 

 in the case where diaphragms were inserted, the motion was 

 transmitted from layer to layer without any great amount of 

 displacement in each individual particle. It seems possible 

 that some of the vortex motion caused by the " break-down " 

 might be transmitted through the tube, and that the dia- 

 phragms merely served to change the portions of air of which 

 the rings were actually composed. I admit that this is not 

 very satisfactory ; but if the roughening of the internal sur- 

 face of the tubes actually exerted the effect attributed to it, 

 we are, I think, justified in supposing that the explosions were 

 not caused by the transmission of vibrations through the 

 material of the pipe itself. Again, vibrations, to be of any 

 effect in producing chemical change, must be comparable, as 

 to period, with the molecular vibrations. If such vibrations 

 are transmitted through ether, it is difficult to see where the 

 influence of chalking the inside of the tubes can come in, 

 and if through air, their wave-length would be too small (as 

 will be shown) to be likely to be much influenced by particles 

 of the size of chalk-dust. 



The similar experiments of Campion and Pellet (Comptes 

 Rendus, lxxv.) are sufficiently explained by their statement that 

 they used iodide of nitrogen. Unless any one likes to suppose 

 that the period of a fiddle-string may be comparable with the 

 period of an iodide of nitrogen molecule, the further experi- 

 ments of Campion and Pellet cannot be held to have much 

 bearing on the subject. One can only wonder that they 

 found a string that would vibrate slowly enough not to fire 

 their iodide. As to their experiments with mirrors, black- 



