30 FRAGMENTS OF SCIENCE 



be the weight of the air overhead. He caused a friend 

 to ascend the Puy de Dome, carrying with him a baro- 

 metric column; and it was found that during the ascent 

 the column sank, and that during the subsequent descent 

 the column rose. 



Between the time here referred to and the present, mil- 

 lions of experiments have been made upon this subject. 

 Every village pump is an apparatus for such experiments. 

 In thousands of instances, moreover, pumps have refused 

 to work; but on examination it has infallibly been found 

 that the well was dry, that the pump required priming, 

 or that some other defect in the apparatus accounted for 

 the anomalous action. In every case of the kind the skill 

 of the pump -maker has been found to be the true remedy. 

 In no case has the pressure of the atmosphere ceased; 

 constancy, as regards the lifting of pump-water, has been 

 hitherto the demonstrated rule of nature. So also as re- 

 gards Pascal's experiment. His experience has been the 

 universal experience ever since. Men have climbed moun- 

 tains, and gone up in balloons; but no deviation from 

 Pascal's result has ever been observed. Barometers, like 

 pumps, have refused to act; but instead of indicating any 

 suspension of the operations of Nature, or any interference 

 on the part of its Author with atmospheric pressure, ex- 

 amination has in every instance fixed the anomaly upon 

 the instruments themselves. It is this welding, then, of 

 rigid logic to verifying fact that Mr. Mozley refers to an 

 1 ' unreasoning impulse. ' ' 



Let us now briefly consider the case of Newton. Be- 

 fore his time men had occupied themselves with the 

 problem of the solar system. Kepler had deduced, from 

 a vast mass of observations, those general expressions of 



