Intelligence and Miscellaneous Articles. 399 



I lowered the temperature by causing the gas to traverse a glass 

 tube immersed in a freezing-mixture whose temperature was indi- 

 cated by a thermometer. 



The rectilinear distance traversed was measured with particular 

 care, by means of iron measures 2 metres in length, which were 

 compared with the official standards. 



The bent part was directly estimated by taking the rectilinear 

 length equivalent to it : I determined it by comparing the lengths 

 of two organ-pipes — one straight, and the other comprising the 

 bend itself which had been used in my experiments — these two 

 tubes being brought to render exactly the same fundamental note. 



From the mean of several series made at the temperature of 

 melting ice, and comprising seventy-seven determinations, I deduce, 

 for the velocity of the propagation of a disturbance under the given 

 conditions, the number 330*66 metres in a second, — the possible error 

 not exceeding T £ m . 



In the introduction to my memoir I discuss the previous experi- 

 ments on the propagation of sound in the atmosphere ; I reduce 

 them to zero by means of Regnault's coefficient of expansion, and 

 make at the same time the probable correction in each case for the 

 hygrometric state of the air. Of eight numbers representing the 

 results, five are between 332 and 332*44 metres. On the other 

 hand, the number obtained in 1822 by Arago and the Bureau des 

 Longitudes, agrees almost exactly with my estimate. The English 

 astronomer Goldingham's number (33 1*1 metres) is very near Arago's 

 number. 



Experiments made in the open air on a base of several kilometres 

 can clearly only inspire a very limited degree of confidence, owing 

 to the considerable uncertainty which must prevail as to the true 

 value of the temperature of the air in the path of the sonorous im- 

 pulse. And the error in this respect is the more to be feared, 

 as most of these experiments have been made during the night. 

 Now the researches of modern physicists, such as MM. Babinet, 

 Becquerel, Martens, &c, have proved the existence of a maximum 

 temperature during the night extending to a greater or less height. 

 And Mr. Glaisher's balloon-ascents during the night have shown 

 that the temperature often continues to increase for considerable 

 heights. The influence of this cause would be to give velocities 

 which are too great ; and it is exactly the smallest of the numbers 

 found for the propagation in free air which is nearest to that I found 

 for the propagation in a cylinder. 



Hence it is probable that the two velocities of propagation, 

 spherical and cylindrical, are equal ; but to obtain a definite solution 

 of this question it would be necessary to work in the atmosphere 

 on a small basis, so as to be able completely to study the distribu- 

 tion of temperature in the space. My experimental method is 

 eminently fitted for such an investigation ; I had arranged ap- 

 paratus with this view ; but the perfect calm of the atmosphere 

 which is necessary for the working appears difficult to meet with 

 in our climates. — Comptes Rendus, March 4, 1S67. 



