170 Mr. Nixon on the Theory/ of the [SE^i-i 



wards so affixed as to exhibit the heitrht of the column at the 

 reduced, in lieu of the measured quantity, the calculations may 

 be made as before without introducing any correction whatever 

 for capacity or capillarity. 



Calculation » 



Measured height above the level of the mercury .... 30*000 in. 



Correction for capacity l^j *600 



Capillarity -100 



-500 



Height to be indicated by the scale. 29*600 



The correction for capacity is in the ratio of the square of the 

 interior diameter of the tube to the square of the diameter of 

 the cistern rninus that of the exterior diameter of the tube.^ As 

 an incorrect estimate of the capacity will influence the adjust- 

 ment of the scale in the ratio of the height of the column, the 

 artist should make the measurement at as low a pressure as 

 practicable. 



The Englefield barometers as at present constructed have 

 their neutral points marked at or about 30 inches, so that the 

 computer has frequently to make the correction for the lower 

 barometer additive, and the one for the instrument at the upper 

 station sub tractive. These corrections may indeed be shunned 

 by increasing the calculated altitude in the ratio of the capacity,t 

 but the method is only an indifferent approximation, capable of 

 introducing an error (in defect, when the mean of the two 

 pressures falls below the neutral point) equal to the 100th part 

 of the altitude. 



When the absolute pressure is required, it may be readily 

 found with the zero adjusted as proposed by increasing the 

 observed pressures in the ratio of capacity, and then correcting 

 them for temperature. To adjust the zero of a scale already 

 attached to the barometer, raise the scale, or lower the cistern 

 with the connected tube by a quantity equal to the height of 

 the neutral point divided by the fraction indicating the capacity 

 minus the correction for capillarity .J 



• When the tube and cLstem are not formed of materials expanding alike from 

 change of temperature, the ratio of capillarity will be (slightly) variable. 



-}• To reduce the errors of this approximative method, the neutral point should be 

 equal to the mean of the pressures likely to occur in general barometrical observations, 

 for instance, 26 or 27 inches. 



J Were we to immerse in the cistern of a stalionari) barometer a vcrti<!al rod of glass 

 of the same diameter as the mercurial colunm within the tube, and so connected with 

 the index as to move with it in a vertical line; but in an opposite direction, an exact 

 compensation would be effected for the drainage and filling of the cistern at pressures 

 difiering from the neutral poinu Should the length, equal to tlie range of the pressure, 



