measuring the Velocity of Sound in the Atmosphere. 37 



internal arrangement fig. 5, Plate I., gives a ground-plan. On 

 one of the ends are the two embouchures A and B ; on the bottom, 

 partitions 7 centims. in height and 7 centims. distant from each 

 other are glued, as represented in fig. 1. The box was closed by a 

 cover which fitted tightly the sides and ends, as also the partitions. 

 It is readily seen that the box thus represents a coiled square tube, 

 the two ends of which, A and B, are close to each other. On the 

 cover there is, moreover, a rectangularly bent tube of square sec- 

 tion, C D, fig. 6, open at D, and also a pair of shallow fillets. The 

 former serve for receiving the sound which starts from the mouth 

 of a small pistol E which acts as source of sound, the latter to 

 give this pistol a firmer bearing. The tube C D takes up the 

 sound at D, conducts it to the opening in the cover C, through 

 this down to the lower space at F, fig. 1, where its course divides 

 into two, opposite in direction, — one in the direction of the arrow 

 m, the short wayto A ; and the other in the direction of the arrows 

 n,o,p, g, r, s the path to B, which is exactly 6 metres longer. Hence 

 the sound will reach B at an interval of time after A, the magni- 

 tude of which interval depends on the length, 6 metres, of the 

 path of the sound between the points F and Gr. To determine this 

 interval, a disk 33 centims. in diameter, moveable by a wheel 

 O P, fig. 7, was used, along with an additional piece of appa- 

 ratus, K I, consisting of a board M N and the short tubes H and 

 I, K and L, fastened on it. The latter part, with the additions 

 H and I, was inserted in the openings A and B, and fastened by 

 means of screws. The metal tubes K and L, 65 millims. in 

 width, are exactly in the directions of A and B in the perforated 

 board ; so that the sound can pass unaltered from A and B to the 

 caoutchouc membranes stretched on K and L, and can excite 

 them. Exactly in the centres of the external surfaces of the latter, 

 wooden blocks are fastened, supporting the iron rods / and g, 

 which are 5 millims. long and 1 millim. in thickness. 



To obtain an approximate idea of the mode in which a dis- 

 charge from the gun E acts, it was charged with 3 decigrammes of 

 powder and fired, after a board 25 millims. in thickness and about 

 two pounds weight had been so placed that it touched the rods 

 /and g without pressing them inwards. The board was projected 

 at once with a violent shock. In a second experiment the same 

 board was coated with white paper, and the ends of the rods, 

 which were quite smooth, coated with printers' ink. The board, 

 which was at a distance of 15 millims. from the rods, was not 

 hit after a second equally strong shot; but the membrane d 3 in 

 consequence of too violent agitation, was burst, so that a new 

 and less tight one than e had to be stretched. As the charge of 

 3 decigr. was obviously too strong, it was reduced to 2 decigr.; and 

 by continued trials, the board being gradually brought nearer to 



