352 THE SCIENTIFIC PAPERS OF 



experiments to determine the velocity of carriers in pneumatic 

 tubes by electrical markers, with records of the observations on a 

 chronograph. The results thus obtained must, he thought, be 

 accepted as indisputable ; but he was inclined to doubt some of 

 the generalisations attempted in the paper. It was perfectly true 

 that when two carriers followed one another in a tube worked by a 

 continuous current, the time occupied by each carrier in traversing 

 the same section of the tube from one marker to another must be 

 the same, because the current flowing through the tube was always 

 the same ; but it did not follow that the absolute speed, the 

 number of feet traversed per second, should be the same in each 

 portion of the tube. M. Bontemps appeared to have found that 

 that was substantially the case that after a short period of 

 acceleration, the speed of the carrier fell into a uniform rate 

 until almost the very end of the journey, when it again increased, 

 and he stated that these results seemed to verify Fournier's 

 theorem, according to which "equal impulses given throughout 

 the journey of an accelerated body must produce the same velo- 

 city." These results did not coincide with the common-sense 

 view of an elastic fluid expanded behind a light working piston, 

 but he thought that an explanation of the experimental results 

 \vas nevertheless possible. The air, say of 2 atmospheres pressure 

 at one end expanded down gradually to atmospheric pressure, 

 and the same index of air between the two carriers must elongate 

 as the carriers went along ; and expansion must take place 

 throughout the course because working power was required at 

 every point. But in taking the case of a carrier not fitting the 

 tube entirely, and yet causing some friction against the sides, he 

 should expect the results which were stated in the paper. In 

 that case the impulse given to the carrier in the tube would be 

 carried by the rush of air past it, and this would be the same 

 throughout, and there would practically be the same power active 

 to overcome friction at every step of the course. The result would 

 l>e a uniform speed for the chief part of the course, till the very 

 end, when the rush of air past the piston would greatly increase. 

 It was to be hoped that the author would continue his observa- 

 tions with the appliances he had made in order to obtain further 

 information on the interesting subject of gaseous friction in long 

 tubes. An explanation had been attempted by Mr. Preece, of 



