Progress of Inquiry.

Being disposed to think that the investiga­tion could be carried further, I proceeded to look about for any examples in horses which might show the transmission of these artificial results to their descendants, and had to wait awhile before I could see which of the regions affected by the pressure of harness were likely to afford the required phenomena. These were in due time forthcoming, and will form the chief subject of the present chapter. I look upon them as cases of an undesigned experiment and will describe them later.

In the present stage of science all hypotheses must be submitted to the test of experiment before they can enter the charmed circle of natural laws. For this reason one must endeavour to apply the test of experiment to the hypothesis before us.

The Nature of Experiment.

Hitherto I have gone no further than the region of experience and observa­tion, from which, Jevons says, “all knowledge proceeds.” There has been abundance of observa­tion of phenomena in this quest and I have ventured even on hypothesis. Experiment is shortly defined by Jevons as observa­tion plus altera­tion of conditions. He points out that when we make an experiment we more or less influence the events which we observe, as when we bring together certain substances under various conditions of temperature, pressure, electric disturbance or chemical action and so on, and then record the changes observed; and, that experiment may be of two kinds, experiments of simple fact and experiments of quantity. It is unnecessary here to describe all the rigorous rules that the man of science so rightly imposes upon himself before he claims to have proved his hypothesis, merely adding that among others he requires, Exclusion of Indifferent Circumstances, Simplifica­tion of Experiments, Removal of Usual Conditions, Removal of Interference of Unsuspected Conditions, Blind or Test Experiments, Negative Results of Experiment, and he lays down the limits of experiment. Those who have not for themselves investigated some scientific problem may learn from this statement some of the difficulties of the work of scientific men and will not fail to respect and admire the caution, patience and honesty of the scientific worker, and will perhaps feel the more gratitude to a class of men by whose self-denying labours they live and move and have their being in a modern state, and by whose discoveries, thus established, they are frequently preserved from premature death.

Experiments for the Present Purpose.

Now in the matter of experiment for the proof of the thesis that changes in the habits of an animal cause the changes observed in their hair, it is at once seen that, ex hypothesi, no one can impose and work with such calculated conditions as are ordained by experiment, strictly so-called. The action of a habit is a slow process and the movement of a hair is slow; moreover the lifetime of a man is too short and that of a horse, for example, too long to allow of any individual experimenter applying artificial pressure through many generations of horses, so as to be able to verify his assertion that the effects of artificial pressure do what is claimed, and that these effects are transmitted from one genera­tion of horses to another. One can conceive a calculated experiment of the kind made with numerous individual rats, and successive generations, but it is hardly likely that effectual pressure could be applied to the hairy coats of such small and elusive mammals as would serve to test the hypothesis.

Undesigned Experiments.

We are thrown back, then, on such experiments as may be provided for us by the uncalculated operations of man through many ages. This class I call undesigned experiments and have had more to say about numerous examples of these in another place.[57] Using the term experiment broadly we see many occurrences which consist in an accidental observa­tion of a fact, and Jevons mentions five of these which have led to organised results in science—the double refrac­tion in Iceland spar by Erasmus Bartholinus, the twitching of a frog’s leg under stimuli by Galvani, the light reflected from distant windows with a double-refracting substance by Malus, the form of a vertebra by Oken, and the peculiar appearance of a solution of quinine by Sir John Herschel. But he notes something further than this, that is, the way in which astronomers make the earth’s orbit the basis of a well-arranged natural experiment. He says further that “Nature has made no experiment at all for us within historical times” among animals living in a state of nature, allowing at the same time that man has made an approach to experiment in his domestica­tion of many animals. Huxley himself kept an open mind until the last as to the validity of Natural Selection in the Origin of Species, because of the fact that races which are sterile together have not yet been produced by human cultiva­tion, for example, the sterility of mules, the human product of the jackass and the mare. I allude to this to show that such a result, if effected, would have constituted a valuable experiment in biology in favour of Natural Selection.