166 



difficulties: much error and confusion have arisen from these 

 difficulties, yet the source of perpetual blunders and of the greatest 

 confusion has never in this instance been even suspected. It is my 

 present business to indicate some points which belong to the topic. 



16. Perhaps it may be said that all the phenomena of life 

 which can become the subjects of investigation are matters of in- 

 ference; hence the great uncertainty which must attach to them, 

 and the more especially as the few facts upon which the inferences 

 are grounded are not those of every-day experience, so that their 

 force and connexions are known but partially. In agencies of 

 which the senses can take account, it is easy to discriminate be- 

 tween a phenomenon which results from a cause conferred and 

 one taken away. I need only advert to the operations of chymis- 

 try: though here a great deal of the subtler business is matter of 

 inference; and therefore this science is not certain to the full ex- 

 tent to which it is investigated, or to the extent for which it has 

 obtained credit. To return to the inquiry respecting our 

 criterion, &c. 



17. If the connection between two continuous parts may be 

 separated, and each part preserve unchanged the properties which 

 belonged to it, it may be inferred that the properties existing in 

 each of these parts have no essential dependence upon those 

 existing in the other. 



18. A primary, producing a secondary affection in a different 

 seat, proves of itself only that the properties of the two seats are 

 liable to communicate. 



19. In order to determine the mode of the secondary affec- 

 tion, it must first be ascertained what are the effects of a separa- 

 tion of the two continuous seats? If such a separation might be 

 made, and each seat preserve unchanged the character and pro- 

 perties which before belonged to them, then it might be inferred 

 that the secondary affection does not arise from the privation of 

 properties naturally and regularly imparted to its seat, but that 

 the secondary affection results from a foreign influence, conferred 

 in consequence of the preternatural condition which the properties 

 pf the primary seat bad assumed. 



20. If in consequence of the separation of connected parts 

 the characteristic properties pf one should be rendered extinct, 

 the alternatives to be inferred are, eitjier that the injury involved 

 in the separation has produced the extinction of these properties 

 by an influence conferred, or that their extinction happens in con- 

 sequence of an habitual source of the properties in question being 

 simply intercepted. 



21. In order to decide this matter in a way which is the least 

 likely to mislead us, we must recur to our nearest analogies. In 

 a general way, then, in consideration of these analogies, we are 

 warranted to infer that a simple division of continuous parts does 

 not operate to the destruction of properties independently main- 

 tained; one iottaope pf this has been cited i" the division pf 



