100 



PROCEEDINGS OF THE AMERICAN ACADEMY 



Fig 



might be entirely remedied by supplying an additional cup, n, Fig. 2, 



within the other; the inner cup 

 being of such capacity that the 

 mercury expelled from the meas- 

 uring-tube during an experi- 

 ment would be certain to cause 

 an overflow of mercury (no 

 matter how little) from the 

 inner into the outer cup. With 

 such an arrangement it is ob- 

 vious that at the time of meas- 

 urement the inner cup would be 

 always full of mercury, and con- 

 sequently that the latter would 

 always stand at the same level 

 on the measuring-tube ; hence 

 that this level may be previously 

 ascertained, once for all, and 

 thus obviate the necessity of 

 making this reading during an 

 experiment. It will then only 

 be required to make the single 

 reading at the upper level of 

 the mercurial column. The dif- 

 ference between this level and 

 the constant level, previously 

 ascertained, at the top of the mercury in the inner cup, will give di- 

 ^•e- 2. rectly the height of the column of mercury, which, corrected 

 for temperature, is to be deducted from that of the barometer 

 to find the pressure to which the vapor is subjected. 



With this slight alteration of Bunsen's apparatus, it ap- 

 pears to me far preferable to the more complicated and 

 expensive apparatus of Gay-Liussac. It has not only the 

 advantage of greater simplicity, and of economy in the 

 quantity of mercury required, but recommends itself also 

 for its convenience, and the facility with which the vapor 

 may be brought to and maintained at a constant tempera- 

 ture. 



