72 WISCONSIis ACADEMY SCIENCES, ARTS, AND LETTERS. 



reliable. The distrust generally arises from an insufficient ac- 

 quaintance with the instrument and its defects. For instance, a 

 mercurial barometer may be taken from one temperature to a 

 higher, and instead of rising as it should do the mercury may fall; 

 everything else being correct, the vacuum cannot be perfect, a small 

 quantity of air is above the mercury which expands according to Gray 

 Lussac's law with the increase of temperature and hence instead of 

 rising, as the mercury should from expansion, it falls. 



The barometer consists of two parts; a tube and a basin. The 

 tube is first filled with mercury and then the open end is inserted 

 into the basin of mercury, the other, or upper end, being closed in 

 the manufacture. The mercury in the tube will now descend until 

 the height of the column measured from one surface to the other, 

 is just in equilibrium with the weight of the atmosphere. It makes 

 no difference how long the tube may be, whether three feet or three 

 hundred, the difference of level will measure the weight of the at- 

 mosphere. If the tube is inserted deeper into the basin a corres- 

 ponding rise will take place in the tube. Such being the case, how 

 does the expansion of the mercury have an influence on the height 

 of the column as we have seen that the quantity has nothing what- 

 ever to do v/ith the same; this will be accounted for from the fact 

 that the mercury has lost in specific weight, in other words, has 

 brcome lighter and the atmosphere is consequently enabled to sup- 

 port a correspondingly greater column of equal weight. I would 

 here remark that the greatest distrust to the barometer may prob- 

 ably arise from too many observations being made indoors, for al- 

 though the detached thermometer would indicate a dilation of the 

 atmosphere, yet this local dilation does not aflect the tension but 

 acts merely as a cushion. Cases may occur, however, where the 

 atmosphere of a room may be in an abnormal condition, that is 

 when the heat currents are such as to carry small objects as though 

 supported by a denser medium. 



I have under such circumstances found a difterence of tempera- 

 ture between the \ipper and lower ends of the instrument amount- 

 ing sometimes to as much as ten degrees. At such times the bar- 

 ometer is very seriously affected and is entirely unreliable. 



In making observations for heights, the same should be made 

 simultaneously at both stations, as the density sometimes changes 

 as much as 200 feet in a few hours ; the instruments should be care- 



