1^ 



PORTABLE BAKOMETEll. 



CORKECTIOX FOR TEMI^ERATURE. 



For 04" = 6 hundredths, 25 . 3 



2 = 2 four hundredth, 2 . 

 V = 1 eight hundredth, . 5 



Correction, + 27 . 8 



Do. by Sir G. Shuckburg^ 27 • 2 



Error, -f 0.6 



By me. By Sir G. Shuckburg, 

 Approximate height, 422 . 6" 422 . 9 



Correction for Temper. 27 . 8 — • 27-2 



459 .4 450 . 1 



These two examples shew how near the truth the me- 

 thod here recommended will come, even in coisidcrable 

 heights. 



Estiinat.c of the jf [-^^^ fjgp,^ already observed, that in observatrons made 



proi)able error • , -i *' 



which may ^^^"^ tiic barometer I have described, a small correction is 



differencTi'n ^^^^^^^^Y' ^n account of the rise in the mercury in the cis- 

 the proportion ^rns as the barometer falls. Altitudes being in all cases mea- 

 tL'hlberthat''''^'^-^^'*^''' differences of the heights of the mercury at th^ 

 ofthe cistern; it '^"'o stations, and these differences being evidently always too 

 ihaTone'thou! ""'""^^ '" *^'' barometer, the correction is obviously always 

 sandth part, additive. As, in constructing different barometers, the inte- 

 rior and exterior diameters of the tube will not always be 

 exactly similar, though the cisterns may be turned always 

 alike, this error, and of course the correction for it, should 

 be in each instrument, deduced, from a comparison .witli a 

 barometer of known accuracy, at different heights. It will 

 probably vary in dilicient instruments from a ninetieth to a 

 seventieth. Indeed, if it were always taken at an eightieth, 

 in instrumonis constructed as above directed, the possiblas 

 error could only amount to about one foot on a thousand; -c 

 quantity of very little importance. 



witr^^e^'s^ngle ^^ "^''' ""'"'"^"^ ^^ '''>' ^ ^^^ -words on the necessity of twe 

 barometer pos- baro- 



