Mr Glover on Forms of induction. 133 



been described, and it is found, that, in some plants of that class, 

 the particular configuration of the family exists along with a pecu- 

 liar medicinal or culinary virtue ; it yet cannot be stated, on the 

 grounds of such a knowledge, what may be the virtues of other 

 plants of the order. A presumption may be formed from the 

 investigated cases, as to the nature of the relations of certain 

 properties in the individuals not examined ; but it must be a 

 mere hypothesis. If, however, the botanist had the pri.vcr of 

 experimenting upon a cruciferous plant, so that he co; 'd find 

 the four crosswise-placed petals and the peculiar shaped pod to 

 be essentially connected with the virtues of a plant, — in such a 

 manner, as that while the structure was modified, the other pro- 

 perty should also undergo modification, — and in fact so that a 

 certain not-to-be-doubted relation could be detected between the 

 two sets of properties, — he might tiien conclude, on the convic- 

 tion of the uniform order of nature derived from past experience, 

 what were the virtues of all cruciferous plants. But he cannot 

 perform such a precise experiment, and therefore must be con- 

 tent to collect every instance, before drawing a general conclu- 

 sion. 2. There seems to be a class of sciences, of which we shall 

 take Medicine as an example, in which most of the inductions 

 are defective ; or, more properly, where there cannot be formed 

 in most cases any complete induction, as in all such cases logi- 

 cally defined genera cannot be procured, since new instances are 

 being continually created, — beings with diverse constitutions ; 

 so that from the examination of already existing cases, no per- 

 fectly certain inference can be drawn as to the whole class of 

 analogous cases; and where, besides, the differences between in- 

 dividuals are such as to prevent one individual being taken as 

 a sufficient example of another ; so that the induction can only 

 conclude with certainty as to an individual. And the history 

 of medicine affords proof of the enormous difficulty thus oppos- 

 ing generalization. Suppose that the diseased structure of the 

 intestinal glands could be found in one case to preserve a con- 

 stant relation with the symptoms of fever ; yet this invariablc- 

 ness of accompaniment should only afford a presumption as to 

 what may exist in other cases ; for, there can be no conviction 

 in the uniformity of nature, when the actions of nature in each 



