On the common Barometer. 



331 



some mechanical contrivance ; but none of these can be 

 added I') nn instrument, without adding to its price. 



But this inipeiFection may be removed, uMthout any ad- 

 ditional apparatus or expense; for we need only find how- 

 high the surface of the mercury in the cistern would be 

 raised by anv given fall in the tube, to find the correction 

 for e\erv other point in llie scale. 



Si'.ppose it were required to find the correction for one 

 inch rise of the quicksilver above the zero, or that point 

 in the scale v\ here no correction is necessary : Let a = 

 the diameter of the cistern, and /' = the inside diameter 

 of the tube ; then, as equal cylmders have their altitudes 



reciprocally as their bases, we have a a : hi) : : ] : — , 



which is the correction to he added to the altitude of the 

 column of quicksilver, when it rses one inch above the 

 zero, and subtracted from the alti*ude of the column, when 

 its surface falls one inch below thai point. 



Example. — Suppose the inside diameter of the tube be 



— of an inch, and the diameter of the cistern be two 



inches = -^ : Then 



= "01 = the correction for one- 



inch rise or fall of the mercury in the tube. 



Suppose the barometer needs no coriection. when the 

 surface of the quicksilver stands at ^9'5 on the scale, then 

 the corrections for the wliole scale, according to the pre- 

 cedmii example, would be as in Table 1. 



But Mr. De Luc observes, that^ the cistern ban^meter 

 /does not (five the true ])ressure of the atmosphere; the 

 quicksilver in it being a little depressed on the same pnnciplc 

 as in capillary tubes. It is therefore thought necessary, 

 where so much nicety is required, to determine by experi- 

 ment how much the quicksilver is depressed m tubes oi a 

 given bore, and to allo^v accordingly, by some experi- 

 ments which have been made on this subject by Lord Charles 

 Cavendish, the tlepres'iou appears to be -067 of an inch 

 ibr a tube whose inside diameter is 0"2.* 



Now, if OG? be added to ea:;h of the corrections in 

 Table 1. we shall have the corrections in Table 2. 



• Philos. Trans, vol. Ixvi. for the year J776. 



Table 



