198 THE PERSISTENCE OF FORCE. 



by which body manifests itself as existing and the persist- 

 ence of the extrinsic force by which body acts. For it is 

 from these equal units of linear extension, through the 

 medium of the equal-armed lever or scales, that we derive 

 our equal units of weight, or gravitative force ; and only by 

 means of these can we make those quantitative comparisons 

 by which the truths of exact science are reached. Through- 

 out the investigations leading the chemist to the conclusion 

 that of the carbon which has disappeared during combustion, 

 no portion has been lost, what is his repeatedly-assigned 

 proof? That afforded by the scales. In what terms is the 

 verdict of the scales given? In grains in units of weight 

 in units of gravitative force. And what is the total con- 

 tent of the verdict ? That as many units of gravitative force 

 as the carbon exhibited at first, it exhibits still. The valid- 

 ity of the inference, then, depends entirely upon the con- 

 stancy of ike units of force. If the force with which the 

 portion of metal called a grain-weight, tends towards the 

 Earth, has varied, the inference that matter is indestructi- 

 ble is vicious. Everything turns on the truth of the as- 

 sumption that the gravitation of the weights is persist- 

 ent; and of this no proof is assigned, or can be as- 

 signed. In the reasonings of the astronomer 

 there is a like implication; from which we may draw the 

 like conclusion. No problem in celestial physics can be 

 solved without the assumption of some unit of force. This 

 unit need not be, like a pound or a ton, one of which we can 

 take direct cognizance. It is requisite only that the mutual 

 attraction which some two of the bodies concerned exercise 

 at a given distance, should be taken as one ; so that the other 

 attractions with which the problem deals, may be expressed 

 in terms of this one. Such unit being assumed, the motions 

 which the respective masses will generate in each other in 

 a given time, are calculated; and compounding these with 

 the motions they already have, their places at the end of that 

 time are predicted. The prediction is verified by observa- 



