XLII

Among Prerogative Instances I will put in the nineteenth place Supplementary or Substitutive Instances, which I also call Instances of Refuge. They are those which supply information when the senses entirely fail us, and therefore we fly to them when appropriate instances are not to be had. Now substitution is made in two ways: either by gradual approximation or by analogy. To take an example: There is no medium known by the interposition of which the operation of the magnet in drawing iron is entirely prevented. Gold placed between does not stop it, nor silver, nor stone, nor glass, wood, water, oil, cloth or fibrous substances, air, flame, etc. But yet by nice tests some medium may possibly be found to deaden its virtue more than any other; comparatively, that is, and in some degree. Thus it may be that the magnet would not attract iron as well through a mass of gold as through an equal space of air, or through ignited silver as well as through cold; and so in other cases. For I have not made the trial myself in these cases. It is enough to propose such experiments by way of example. Again, there is no body we are acquainted with which does not contract heat on being brought near the fire. And yet air contracts heat much more quickly than stone. Such is the substitution which is made by gradual approximation.

Substitution by analogy is doubtless useful, but is less certain, and should therefore be applied with some judgment. It is employed when things not directly perceptible are brought within reach of the sense, not by perceptible operations of the imperceptible body itself, but by observation of some cognate body which is perceptible. For example, suppose we are inquiring into the mixture of spirits, which are invisible bodies. There seems to be a certain affinity between bodies and the matter that feeds or nourishes them. Now the food of flame seems to be oil and fat substances; of air, water and watery substances; for flame multiplies itself over exhalations of oil, air over the vapor of water. We should therefore look to the mixture of water and oil, which manifests itself to the sense, since the mixture of air and flame escapes the sense. Now oil and water, which are mingled together very imperfectly by composition or agitation, are in herbs and blood and the parts of animals very subtly and finely mingled. It is possible, therefore, that something similar may be the case with the mixture of flame and air in pneumatic bodies, which, though not readily mingling by simple commixture, yet seem to be mingled together in the spirits of plants and animals, especially as all animate spirit feeds on moist substances of both kinds, watery and fat, as its proper food.

Again, if the inquiry be not into the more perfect mixtures of pneumatic bodies but simply into their composition, that is, whether they be readily incorporated together; or whether there be not rather, for example, certain winds and exhalations or other pneumatic bodies which do not mix with common air, but remain suspended and floating therein in globules and drops and are rather broken and crushed by the air than admitted into or incorporated with it — this is a thing which cannot be made manifest to the senses in common air and other pneumatic bodies, by reason of their subtlety. Yet how far the thing may take place we may conceive, by way of image or representation, from what takes place in such liquids as quicksilver, oil, or water, and likewise from the breaking up of air when it is dispersed in water and rises in little bubbles; and again in the thicker kinds of smoke; and lastly, in dust raised and floating in the air; in all of which cases no incorporation takes place. Now the representation I have described is not a bad one for the matter in question, provided that diligent inquiry has been first made whether there can be such a heterogeneity in pneumatic bodies as we find there is in liquids. For if there can, then these images by analogy may not inconveniently be substituted.

But with regard to these supplementary instances, although I stated that information was to be derived from them in the absence of instances proper, as a last resource, yet I wish it to be understood that they are also of great use even when proper instances are at hand — for the purpose, I mean, of corroborating the information which the others supply. But I shall treat of them more fully when I come in due course to speak of the Supports of Induction.