In part, or even entirely, biology is deprived of certain methodical processes which are utilised by the sciences which precede it. It cannot avail itself of calculation. Undoubtedly each of the elements which go to make up a physiological phenomenon varies according to a definite law. But the sum of these elements forms such a complex whole, that we shall never be able to express their relations in the terms of an equation. Further, the numbers which are relative to the phenomena of living bodies present continual and irregular variations, which do not allow us to establish the data of a mathematical calculation.[161] Each living being has its individuality, its personal formula, its characteristic reactions, which prevent us from treating it as identical with the other beings of the same species. Each physiological or pathological “case” is distinct from any other case. That is why Comte distrusts statistics. In his judgment they are misleading in physiology, and fatal in medicine. In the same way, Claude Bernard will protest vigorously against averages. Is the inductive method at least a convenient one to make use of in biology? Simple observation cannot lead us far in the study of such complex phenomena, of which many are not directly accessible to our senses or to our instruments. Experimenting is very difficult in biology, for nothing is easier than to disturb, to suspend, or even to bring about the entire cessation of the phenomena of life. But it is almost impossible to introduce an exactly determined perturbation, whether of kind, or, _a fortiori_, of degree. Indeed a modification of a single condition of the phenomenon almost at once affects the greater number of the other phenomena, by reason of their _consensus_. In principle, experimenting is not forbidden in biology. On the contrary it is of remarkable efficacy, but it is often impracticable. Nevertheless, as we know, it is not man’s intervention in phenomena which constitutes experimenting properly so called. It consists, before all things, in the rational selection of cases, (natural or artificial, it matters little), which are most appropriate for bringing out the law of variation of the phenomenon under observation. Nature gives us such, for illnesses resemble experiments which we can follow through their entire course and to their termination. They are often difficult to interpret, on account of their extreme complexity, but less so, however, than the majority of the experiments which we bring about ourselves. For are they not more or less violent diseases, suddenly produced by our intervention, without our being able to foresee all their indirect and future consequences? It is pathological anatomy which led Bichat to his fine discoveries in histology and in physiology. And to pathology we must join teratology which is, as it were, its prolongation. Here again, nature supplies experiments which we should not know how to institute.[162]