216 Mr. Meikle on the Construction 



ever, is, in my opinion, an instrument greatly preferable to 

 Mr, Murray's, particularly in being more sensible, and only 

 requiring one correction on account of temperature. Its con- 

 struction is totally different from his ; but the principles em- 

 ployed have little novelty to boast of, though, perhaps, they 

 may not have been applied to the same purpose, at least in 

 the same form ; and therefore I would hope that a brief ac- 

 count of it may not be altogether unacceptable to your readers. 



This instrument, which has some resemblance to an aii'- 

 thermometer, consists of a hollow ball of glass containing air, 

 from which a vertical tube, open at bottom, descends, and 

 terminates in a cistern of mercury*. The mercury is likewise 

 designed to occupy a part of the tube, more or less, according 

 to the state of the atmosphere. Another tube, equal to the 

 former, and placed close by its side, is also immersed in the 

 quicksilver, though open at top. But in order that the air in 

 the ball and first tube may always be readily brought to the 

 same tension as the air without, the cistern consists of a 

 leathern bag, inclosed in a box, the bottom of which is move- 

 able by a screw precisely as in a mountain-barometer. The 

 mercury in the cistern is, however, open to the external air 

 no where but through the tube, v.hich is open at top. 



Now it is manifest, that if the screw at the bottom is turned 

 till the mercury in both tubes stand at the same height, the 

 elasticity of the air within Avill just balance the weight of the 

 atmosphere : and shice in this case the spring of the included 

 air, allowing for change of temperature, cannot sensibly differ 

 from being invei'sely as its bulk, the space which it occupies 

 will always be inversely as the atmospheric pressure. If, 

 therefore, the tube connected with the ball, or a scale by its 

 side, is graduated, and numbers attached proportional to the 

 contents of the ball, and of that part of the tube which lies 

 above them, these numbers* being inversely as the densities, 

 or inversely as the mercurial altitudes in a common barome- 

 ter, are also ordinates to a logarithmic curve, equal that em- 

 ployed in the usual mode of investigation ; and hence the dif- 

 ference of their logarithms has still the same proportion to 

 the difference of elevation f; wherefore these numbers will be 

 equally convenient for the purpose of calculation, as the num- 

 bers on a counnou barometer 1. The mode of ajoplying a 



vernier, 



* Dr. Hook long dgo employed air in the construction of his marine baro- 

 meter ; but that instrument is very different from this in various respects. 



-|- Or, more simply, the difl"erencc of the logarithms of two numbers is 

 equal the difference of the logarithms of their reciprocals: the logarithms of 

 any number being the arithmstical complement of that of its reciprocal. 



% These numbers are equally well suited to the very ingenious method 



of 



