ibct. ii.] DISSERTATION SECOND. 97 



truth of his solution by the most clear and irresistible evi- 

 dence. 



The strict method of Bacon is therefore only necessary 

 where the thing to be explained is new, and where we have 

 no knowledge, or next to none, of the powers employed. 

 Tiiis is but rarely the case, at least in some of the branches 

 of Physicks ; and, therefore, it occurs most commonly in 

 actual investigation, that the inquirer finds himself limited, 

 almost from the first outset, to two or three hypotheses, all 

 other suppositions involving inconsistencies which cannot 

 for a moment be admitted. His business, therefore, is to 

 compare the results of these hypotheses, and to consider 

 what consequences may in any case arise from the one that 

 would not arise from the other. If any such difference can 

 be found, and if the matter is a subject of experiment, we 

 have then an instantia cruris which must decide the ques- 

 tion. 



Thus, the instantia cruris comes in real practice to be 

 the experiment most frequently appealed to, and that from 

 which the most valuable information is derived. 



In executing the method here referred to, the application 

 of much reasoning, and frequently of much mathematical 

 reasoning, is necessary, before any appeal to the experi- 

 ment can be made, in order to deduce from each of the 

 hypotheses an exact estimate of the consequences to which 

 it leads. Suppose, for instance, that the law by which the 

 magnetick virtue decreases in its intensity, as we recede 

 from its poles, were to be inquired into. It is obvious that 

 the number of hypotheses is here indefinite ; and that we 

 have hardly any choice but to begin with the simplest, or 

 with that which is most analogous to the law of other forces 

 propagated from a centre. Whatever law we assume, we 

 must en*er info a good deal of geomefrick reasoning, before 

 a conclusion can be obtained, capable of being brought to 



