526 SPECIAL METHODS OF DISCOVERY. 



the variation of atmospheric pressure with altitude. In 

 1647 Pascal showed practically that if we alter the super- 

 incumbent column of air by going to a high place, we alter 

 the weight which it will support. This celebrated experi- 

 ment was made by Pascal himself on a church steeple in 

 Paris, the column of mercury in the Torricellian tube 

 being used to compare the weights of the air; but he 

 wrote to his brother-in-law, who lived near the high 

 mountain of Puy de Dome in Auvergne, to request him 

 to make the experiment there, where the result would be 

 more decisive. ' You see,' he says, ' that if it happens that 

 the height of the mercury at the top of the hill be less 

 than at the bottom (which I have many reasons to believe, 

 though all those who have thought about it are of a 

 different opinion), it will follow that the weight and pres- 

 sure of the air are the sole cause of this suspension, and 

 not the horror of a vacuum ; since it is very certain that 

 there is more air to weigh on it at the bottom than at the 

 top ; while we cannot say that nature abhors a vacuum 

 at the foot of a mountain more than on its summit. M. 

 Porrier, Pascal's correspondent, made the observation as 

 he had desired, and found a difference of three inches of 

 mercury, " which," he says, " ravished us with admiration 

 and astonishment." ' l Boyle proved by experiment that 

 air was elastic. In 1661, wishing to know how much air 

 was compressed by increase of weight put upon it, he 

 devised his well-known bent tube experiment, and found 

 that double the pressure reduced the volume one-half. 

 Mariotte also, by means of similar experiments, a few 

 years later made the discovery that the volume of air 

 varied inversely with the degree of pressure. Bessel, by 

 employing many different substances to form the bob of a 



1 Whewell, History of Inductive Philosophy, vol. ii. 3rd edit. p. 63. 



