264 Temperature of Mercury in a Siphon Barometer, 



By (14) and (15) 



1.73 -H^ 1.48 -/i' 

 a' is corrected to a'+ „ ' ^' is corrected to h' o 



a" " 394.60; 6" " 356.60. 



a''' " 389.58; b"' " 352.51. 



The left readings are those found in the formula, which are 

 supposed to have been corrected for the heights of the menis- 

 cuses. The readings found in their equivalents, are those taken 

 from the brass scale. Substituting the observed temperature and 



corrected readings in (35,) we have ^ = 1.0013 ; so that the di- 

 ameter of the longer branch being denoted by unity, that of the 

 shorter branch would be 1.0006. 



For the temperature, we have, after the necessary substitutions 

 and reductions 



f = 10.74(a'-1.0013 6'- —^-33.02). (38.) 



In which («',) (6',) are the observed readings, and H', Ii', the 

 height of the meniscuses. 



This formula is easy of application, although it provides for 

 imperfections in the instrument, which are not necessary to it, 

 and which may in a great measure, or perhaps entirely, be avoid- 

 ed by ingenuity and care in the construction. 



Remarks. — If the tubes of the two branches are cylinders of 

 equal diameters, and their interior surfaces are free from tarnish, 

 or any foreign substance, such as dust, humidity, &c., and are 

 filled with pure mercury, according to the well known rules for 

 this process, the heights of the meniscuses, I am confident, would 

 not vary in such a manner as to require a correction in the for- 

 mula for temperature. If barometers were thus constructed, with 

 that care which all exact instruments demand, the temperature 

 could be derived from a formula like (10') in but little more time 

 than would be requisite to read it accurately from a thermometer, 

 and record it ; and what is still more important, the height of the 

 column would not be subject to those errors which a construc- 

 tion faulty in these respects must occasion. 



The means above set forth for determining the mean tempera- 

 ture of the mercury, throughout the whole extent of the column, 

 and for detecting and correcting the defects of the instrument, 

 are peculiar to siphon barometers, and give this form a decided 

 advantage over all others. 



