OF ARTS AND SCIENCES. 191 



of nature. No observer would be willing to risk the value of this 

 long labor by trusting to the new barometer, until its peculiarities are 

 better understood than at present. It may possibly happen, that a long 

 series of observations which eliminates irregularities of weather will 

 eliminate instrumental irregularities at the same time. The same 

 objection applies with greater force to the application of the aneroid 

 barometer to the measurement of heights above the level of the sea. 

 An elevation of eighty-seven feet depresses the barometer by about 

 .1 of an inch only ; hence, a small error in the barometer will entail 

 a large error on the estimated elevation. Moreover, a long series of 

 observations in this case will generally be impracticable. I would 

 make one farther remark in this connection. The mercurial barometer 

 is liable to be broken when exposed to the perils of mountain travel. 

 In this case the damage, though great, is known and appreciated, and 

 no error is introduced into science. Unless the barometer is broken, 

 it is so simple in its construction that it is not likely to be injured at all. 

 It is otherwise with the aneroid barometer. To appearance it is 

 stronger, and can bear a greater strain without being broken. On 

 the other hand, we can easily foresee that it may be materially injured 

 without attracting the notice of the observer at the time, and in this 

 way may conceal its own infirmities under its apparent strength. It 

 should be added, in justice to the aneroid barometer, that it is far from 

 having been carried as yet to that degree of perfection in its mechani- 

 cal execution which the principle on which it is based will allow. When 

 it shall have received, at the hands of the artist, that amount of skill 

 and delicacy in its construction which is expended on the chronometer, 

 a more impartial comparison can be made between its claims and those 

 of the best mercurial barometers." 



