Deductions from Mr.Glaisher's "Meteorological Corrections." 227 



whether the distm-bance is a condensation or a dilatation, or 

 compounded of both. 



To express these principles symbolicallyj let p be the original 

 density, t the original temperature, measured from a point 274°"6 

 Centigrade below the temperature of melting ice for a perfect 

 gas, and 273° for air in its ordinary state ; let p be the original 

 pressm-e, and let 8p, St, Bp be momentaiy small variations of 

 those quantities. Then 



-=(7-1)^ 



T p 



8p Bp Bt Bp 

 P P T p 



and the velocity of propagation 



This is the theoiy of Laplace and Poisson as to the propaga- 

 tion of sound in gases. I cannot help believing that the objec- 

 tions m-ged against it have arisen chiefly from its having been 

 misunderstood. 



London, February 1851. 



XXXI. Deductions from Mr. Glaishei-^s " Meteorological Cor- 

 rections." By S. M. Drach, Esq., F.R.A.S., B.M.S. ^c* 

 THESE empu'ical tables strikingly confirm my investigations 

 in the Philosophical Magazine for June and July 1842, 

 wherein I showed, that if the temperatui-e, &c. for any time t 

 from the epoch be put equal to 



HH-Asin^-|-«cos/ + Bsin2^ + 6cos2^ + Csin3^ + cco3 3^ 

 + E sin 4/ + e cos 4it, 

 that the average of four readings respectively a quai-ter day apart 

 would give the mean =aH+E 8in4/ + ecos4/, a nearly constant 

 quantity if E and e be small. The second paper illustrated this 

 conclusion by Sir D. Brewster's Leith observations (Edin. Phil. 

 Trans., vol. x.), and the annexed tabular Greenwich results cor- 

 roborate it. 



Thus the diurnal average temperature, &c. may be confidently 

 deduced from the mean of four equidistant observations, whatever 

 be the commencing hour ; and thus one observer recording the 

 instmment at 5 and 11 a.m., 5 and 11 p.m., will deduce the 

 mean moi-c accurately than at any fancy hours ; all factors of 

 latitude and solar dechnation in the quantities A, a, &c. being 

 climi)iatcd. For mountain-height determinations these remarks 

 are of some value. 



* Communicated by the Author. 

 Q2 



