1818 Biographical Memoir of Henry Cavendish. 



this object, and had constructed, for the purpose, an apparatus 

 which was nearly the same as that already employed by our 

 deceased fellow member M. Coulomb, for measuring the power 

 of electricity, and that of the magnet. A lever, six feet long, 

 bearing at each extremity a small lead ball, was suspended hori- 

 zontally, by the middle, to a vertical thread. This lever once 

 at rest, a large mass of lead of a given diameter and weight, was 

 brought near each of its extremities in a lateral direction. The 

 attraction, exerted by the masses upon the balls, put the lever 

 in motion. The thread became twisted in accommodating itself 

 to this action, and tending to return to its first state, made the 

 lever describe small horizontal arcs, that is to say, the attrac- 

 tion of the earth made it describe arcs perpendicular to the pen- 

 dulum ; and, by comparing the extent and duration of these 

 oscillations and those of the pendulum, the relation of their 

 causes was obtained, or, in other words, the relation of the attrac- 

 tive power of the masses of lead, and of that of the terrestrial 

 globe. But this presented only a rude idea of the apparatus, and 

 of the precautions and calculations which the experiment required. 

 The mobility of the lever was such, that the slightest difference 

 of heat between the two balls, or only between the different 

 parts of the air, occasioned a current strong enough to make it 

 vibrate. It was even necessary to estimate the attraction of the 

 walls of the wooden case in which it was contained ; and the at- 

 tention required in measuring the extent of its vibrations, and 

 even in observing it without altering them by approaching too 

 near, was almost infinite. All these difficulties became appa- 

 rent only at the moment of performing the experiment ; and 

 the dehcate means which procured their removal, and of which 

 the necessity had not even been foreseen by Mitchell, belong 

 entirely to Mr Cavendish. The result was singular ; the mean 

 density of the globe was found to be 5.^^% times, or something 

 less than 5 J times that of water. According to this result, it 

 would be necessary not only that the globe should have no 

 vacuities in it, but also that the materials of its interior should 

 be heavier than those of the surface ; for the substances, of which 

 the common rocks are composed, are only about three, or rarely 

 four, times the weight of water, and no known stone has a specific 

 gravity so high as five. It might therefore be imagined that the 



