Royal Society. 221 



where b is the variable boiling-point on Fahrenheit's scale and (3 the 

 corresponding barometric pressure, with the formula of Laplace for 

 the determination of the difference in level of two stations from ba- 

 rometric observations, he obtains the formula 



H = 54-7-99 (b - V) 1 1 + (t - 32°) -00222 j , 



where b and b' are the boiling-points on Fahrenheit's scale at the 

 two stations, t the mean temperature of the air at the stations, and 

 H their difference of level in English feet. 



The author describes the particular instrument he employed in 

 his observations, and his mode of determining the correction which 

 it required ; and then gives, in a table, the observations he made on 

 the boiling-point of water at thirty-eight different stations in the 

 Alps ; the heights of the corresponding stations above the sea level, 

 deduced from these observations ; and, for the purpose of compari- 

 son, the heights of the same stations deduced by other observers. 

 The difference between these and some of the author's results are 

 considerable ; but as they are not greater than would probably arise 

 from ordinary barometric measurements, and as there is a close ac- 

 cordance between his results and the determinations on which the 

 greatest reliance can be placed, he concludes that the results are on 

 the whole satisfactory. Considering it, however, desirable to obtain 

 some test of the accuracy of each observation independently of the 

 rest of the series, the author avails himself of the barometric ob- 

 servations made at the Observatory at Geneva and at the Convent 

 of the Great St. Bernard ; and determining from these the corre- 

 sponding temperature of boiling water, deduces the difference of 

 level between each of his stations and these two places considered 

 as fixed points : the sum of the height above Geneva and the de-? 

 pression below the Great St. Bernard should in all cases be the dif- 

 ference of level between the two fixed stations. Although there are 

 here again considerable discrepancies, yet in most cases, where the 

 height of the station may be considered as well-established, the height 

 deduced from the observations agrees with it in a very remarkable 

 manner. 



In another table, the author gives the difference of level between 

 the Observatory at Geneva and the Great St. Bernard, deduced from 

 the recorded observations at those places simultaneous with his own 

 at his various stations; and then remarks that the differences of height 

 determined by the two methods do not differ from one another, in 

 any single case, by so large a quantity as do the greatest and least 

 differences of height deduced from the barometric observations ; 

 while in many cases the accordance is almost perfect. 



The conclusion drawn from the comparisons in these tables is, 

 that the barometric thermometer is capable of affording highly ac- 

 curate and satisfactory results, perhaps even more so than the com- 

 mon form of barometer, but that there is considerable uncertainty 

 attached to its indications. This uncertainty, far from being wholly 

 attributable to the imperfections of the instrument as a measure of 

 the atmospheric pressure, might, the author thinks, arise from an 



