82 GRAVITATION. 



with prodigious violence, might exist in bodies so small, 

 as the balls upon the table, and yet not be sufficient to 

 overcome the difficulties which impede their motion. The 

 little inequalities of the table, too small to be perceived 

 by the senses ; the resistance of the air which must be 

 removed from between them if they come together, and 

 the want of perfect regularity in their form, would, per- 

 haps, be sufficient to prevent the motion taking place, 

 when there was a real tendency to it. If the two balls 

 are suspended by strings, and are brought nearly into 

 contact, the obstruction to motion would be less than 

 before ; still the balls in moving towards each other must 

 evidently rise slightly, on account of the nature of their 

 suspension. Consequently their tendency to come to- 

 gether must be sufficient partially to lift them, or its ef- 

 fects would not be visible. 



Various ingenious plans have been devised for sus- 

 pending bodies in such a manner as to render sensible 

 their gravitation towards each other. An instrument 

 called, from the name of its maker, Cavendish's machine, 

 accomplished the object. Two leaden balls suspended 

 by an apparatus so contrived as to diminish as much as 

 possible resistance and friction, gravitated sensibly to- 

 wards each other. The method adopted was very similar 

 in principle to that employed by Mr Coulomb, for render- 

 ing sensible other weak attractions. This method, very 

 simple and easily imitated, he applied not only to demon- 

 strating gravitation between small bodies, but also to ren- 

 dering sensible, and to measuring very many other weak 

 forces. His instrument was substantially a bar connecting 

 two heavy leaden balls, a and 6, 

 and suspended by the string d, 

 c, attached to the centre of the 

 bar. If now a heavy mass be 

 brought near the ball b, the lat- 

 ter may move towards it with- 

 out being lifted at all, for it 

 may move round horizontally 

 only twisting the string cd. If 

 it move but little, the twist or (~~\_ 



slight force, and as"this is all 



torsion of the string is a very 



O 



