33

Classification of the Sci
33

a sailor who never learned to swim fall overboard, and he can only try to strike out as best he may. That he must and will do, whether he has reason to trust to his natural swimming instinct or does trust to it, or not. Natural good sense is not a faculty that ought to be trusted in questions of a scientific character. But there are occasions when we are reduced to using it and making the best we can of it. Its suggestions will probably be wrong. But if we can find some way of putting them to the test, of thus refuting the first suggestion and can then ask good sense to propose something else to be treated in the same way, there is a hope that we may ultimately get set upon the road to truth. But in the present case, the only checks that the author has been able to devise contain themselves too much of the subjective to inspire much confidence; and the very fact that they turn out favorable to the first suggestion is a suspicious circumstance. The author would be very sorry to have his classification of