Leaving aside the fact that theory is the best guide for experiment—that were it not for theory and the problems and hypotheses that come out of it, we would not know the points we wanted to verify, and hence would experiment aimlessly—a more serious objection is that experiment is seldom if ever perfect, for it nearly always involves some unverified assumption. I have referred to empirical observation and experiment as two different methods. But the difference is mainly, if not solely, one of degree. If we experimented to find out whether acquired characteristics were inherited, it is obvious that our experiments would have to be confined to animals. If we found, let us say, that no acquired characteristic was ever transmitted to offspring, we could not say that this would be equally true of man, but would be justified in concluding only that the acquired characteristics of _animals_ are not transmitted to descendants. Nay, we could not go even this far. We would have to confine ourselves to the statement that certain acquired characteristics of the few score animals we had experimented upon were not transmissible. But even this statement would involve assumption. We could say only that certain acquired characteristics of the few score animals we had experimented upon had not been transmitted in these particular instances. We would have to limit ourselves to a bare statement of fact; we could draw no conclusion whatever. But if we had attacked this problem from the deductive standpoint, and had concluded that owing to certain conditions holding alike in all animals and in man, acquired characteristics _could not possibly_ be transmitted, we would have sufficient ground for deriving from our experiments a broad generalization. Experiment and deduction are not the only methods which can be checked up against each other. We can do likewise with the comparative and the experimental, the historical and the theoretical—in fact, all viewpoints applicable to any one problem. * * * * *