282 



THE BAROMETER. 



formably to the principle explained by Torricelli, the column was observed 

 gradually to diminish in height, as the elevation of the apparatus was increased. 

 The same experiment was repeated by Pascal himself, with similar success, 

 upon a high tower in the city of Paris. 



Meanwhile other effects were manifested which not less unequivocally 

 proved the truth of Torricelli's solution. The apparatus being kept for a length 

 of time in a fixed position, the height of the column was observed to fluctuate 

 from day to day between certain small limits. This effect was, of course, to 

 be attributed to the variation of the weight of the incumbent atmosphere, ari- 

 sing from various meteorological causes. 



The apparatus which we have just described is, in fact, the common barom- 

 eter. By the principles of hydrostatics it appears that the height of the col- 

 umn E F, sustained by the atmospheric pressure, will be the same, whatever 

 be the magnitude of the bore of the tube. If we suppose the section of the 

 bore to be equal to a square inch, then the column E F will be pressed up- 

 ward, and held in equilibrium by the weight of a column of atmosphere pres- 

 sing upon a square inch of the external surface F ; consequently the weight of 

 the column E F, must be equal to the weight of a column of the atmosphere 

 whose base is a square inch, and which extends from the surface of the mer- 

 cury in the cistern to the top of the atmosphere. If there be another tube 

 whose bore is only half a square inch, then the pressure which will support 

 the column in it will be that of a similar column of atmosphere, whose base is 

 half a square inch ; such pressure, then, will only be half the amount of the 

 former, and therefore will only sustain half the weight of mercury. But a 

 column of mercury of half the weight, having a base of half the magnitude, 

 must necessarily have the same height. Hence it appears that so long as the 

 atmosphere presses upon a given magnitude of the surface F, with the same 

 intensity, the column of mercury sustained in the tube will have the same 

 height, whatever be the magnitude of its bore. 



In adapting such an apparatus as this to indicate minute changes in the pres- 

 sure of the atmosphere, there are many circumstances to be attended to, which 

 I propose to explain, so far as they are necessary to render intelligible the 

 general principles and use of the barometer. 



It is, in the first place, necessary to have the means of measuring exactly 

 the height of the column E F, fig. 2. If the surface F were fixed, and the 

 tube B A maintained in its position, it would be sufficient to mark a graduated 

 scale upon the tube, indicating the number of inches and fractions of an inch 

 of any part upon it, from the surface F. But it is obvious that this will not be the 

 case when the pressure of the atmosphere is increased, as an additional quan- 

 tity of mercury is forced into the tube, and consequently an equal quantity is 

 forced out of the cistern. While the surface E rises toward B, the surface 

 F therefore descends, and the distance of E from that surface is increased by 

 both causes. 



A graduated scale marked upon the tube would then only indicate the change 

 in the position of the surface E, but would not show the change in the length 

 of the column E F, so far as that change is affected by the fall of the surface 

 F. There are several ways in which this defect may be remedied. 



If the instrument be not required to give extremely accurate indications, it 

 will be sufficient to use a tube the bore of which is small compared with the 

 magnitude of the cistern. In this case, a small change in the height of the 

 column will make but a very inconsiderable change in the whole quantity of 

 mercury in the cistern, and therefore will produce a very minute effect upon 

 the position of the surface F. If such a change in the level F, be so small as 

 to affect the indications of the instruments in a degree which is unimportant 



