322 



by tfie cessation of the symptoms of the disease. It is a point 01 

 nice discrimination, to decide how far remedies are to be urged 

 When the condition of disease appears to be subverted, so ais on 

 the one hand to subvert also the tendency among constituents, 

 which first produced it (or in other words to prevent a relapse) ; 

 and, on the other, id avoid the danger of deteriorated disease, 

 which may end in death, by the processes which succeed to the 

 continued employment of the remedy. This matter is exemplified 

 familiarly in syphilis; and the point of discrimination is to ad- 

 minister enough mercury to cure venereal without producing mer-' 

 Curial disease: the same or a greater nicety of discrimination is 

 required in the treatment of many chronic diseases. For the pur- 

 purposes of this distinction we must trust to our experience Irs 

 particular cures; as if squill and calomel should cure a dropsy, 

 which will return when the medicine is left off, we must consult 

 odr experience in order to judge what continuance of the preter- 

 natural, or remedial affection is sufficient, not only to remove the 

 present symptoms, but to subvert that tendency to dropsy which 

 first produced it, and may produce it again ; or how long the 

 remedy must be employed to produce a totally new predisposition : 

 our experience also in these cases must be consulted, in order to 

 decide how far, under certain states of constitution, remedies may 

 foe pushed, with this view of subverting diseased tendencies, with- 

 out risking the substitution of a new, and perhaps a worse disease. 

 37. To repeat: the basis of every treatment is the analogy of 

 disease, and the experience of the results of remedies. We say 

 experience of the analogy of disease, and experience of the effects 

 of remedies; but our experience in these respects is imperfect 

 or equivocal. If we were capable of experiencing the similitude 

 of diseases we should never be mistaken in theii* character ; but 

 we often are mistaken. We may experience a resemblance in certain 

 Symptoms, but different latent causes may be associated with a simi- 

 lar exhibition of certain symptoms; these causes are developed in 

 the progress of disease, and then we confess the inadequacy of our 

 experience. In strict truth, this which we call an experience of 

 the similitude of diseases is only an inference, or in part an in- 

 ference: we infer a similitude of the entire state of disease from 

 a sensible analogy in some circumstances. So also with respect 

 to our experience of the effects of remedies: we experience only 

 the succession of events; and disease frequently ceasing from 

 latent, spontaneous processes of causation, under the exhibition of 

 remedies, we infer that the remedies produce the cure because 

 the recovery of health succeeded to their employment. Hence it 

 is that so many remedies have obtained credit for curing diseases, 

 which, upon further trial, have been found to possess no such 

 power. There is only one way (rarely practicable) of guarding 

 against error in making these conclusions, and that is, not to admit 

 succession as the proof of causation, until the succession by its 

 frequency approaches td the invariable, upon which analogy an 



