Mr. Galbraitli on the Velocity of Sound. 337 



periment, than that (1-362) which I had usually supposed the 

 more accurate. Now if this be introduced into formula (4), 

 page 217, and the proper value for g also, a slight modifica- 

 tion will be effected in the general formulae for the velocity in 

 metres or English feet. 



V = (105-9518 + 0-19845 (l + 51 f_2/ ) (3-14a43- 



0-0042 cos 2 X) + CO cos $ (A) 



the velocity in metres using the metrical barometer and 

 centigrade thermometer. 



V = [105-9518 + 0-1103 (^-32°)] (1 + ^TTTY^) (10-2739 



— 0-01378 cos 2 X) + o) cos ^ (B) 



the velocity in English feet, employing the English ba- 

 rometer and Fahrenheit's thermometer. 



As I have been able to find no experiments by which a di- 

 rect comparison with these formulae can be made, except those 

 of Professor Moll of Utrecht, with the omission of the last 

 term, namely, w cos ($>, I cannot say whether it is in this last 

 respect perfectly correct. As Dr. Moll took the precaution 

 of filing guns at each end of the measured base, the effect of 

 the wind was in this case obviated; and if my formula agree 

 nearly with his experiments, independent of this term, it may 

 be looked on in this state as verified by direct experiment. 



The truth of these remarks will be obvious, on consulting 

 Dr. Moll's paper in the Philosophical Ti'ansactions of the 

 Royal Society of London for the year 1824, pages 425, &c. 

 in which a full account of the whole steps of his experi- 

 ments is recorded. At page 445 there is presented a table of 

 the velocity of sound from 44 different experiments on a base 

 near Utrecht in Holland, of about nine miles in length. On 

 the 27th of June 1823, twenty-two shots were fired at each 

 station or end of the base of 17669-28 metres, or 9664-7044 

 fathoms, the metre being supposed 39-3824 English inches. 

 The sum of the times was 2286^-07, which divided by 44 

 gives 51'-96 for the mean of the whole, which Dr. Moll adopts. 



Whence = 340-06 metres, the experimental velocity 



in one second of time. Now while these 22 shots were fired 

 at each end of the measured base, the mean temperature at 

 both ends of it was 11°-16 = ^ of the Centigrade scale. Also 

 the mean lieight of the metrical barometer was -74475 = p ; 

 the mean tension of aqueous vapour by Daniell's hygrometer 

 was 0-00925307 metre —f. By substituting these quantities 

 in formula (A) neglecting the last term, « cos (p, since by 

 firing ginis at both ends of the measured base its cflect, as 

 New Series. Vol. 1. No. 5. Mut/ 1827. 2 X Dr. 



