basis of the new science of statistics. But his consistency did not stop there. His works are full of mathematical computations of a more complex nature.
At this point we must make a distinction. In his purely statistical work it is unjust to criticise results when they are due solely to the want of accurate information. He was conscious of the imperfect sources which he was obliged to use, and he did all in his power to secure trustworthy data. It is quite another thing to object to the method itself, and this in many cases we can fairly do. For example, granting the correctness of his estimate of the population of England, his assumption that the population of a country always increases in the same ratio (107) is one whose falsity he should have seen. Again, in elaborate computations he uses the words "guess" and "suppose" frequently, showing that the whole argument is purely conjectural. Yet he does not hesitate to use such arguments to guide the policy of the State.
Petty's preference and constant use of the deductive method may also be regarded as a sign of Hobbes' influence. In many cases it is difficult to say whether he employs facts simply as illustrations or whether they have been used inductively. I should be inclined to say that the number of purely fictitious illustrations invented to prove or explain general principles, shows that he most frequently used observation to fortify a principle, which he had reached deductively. Taking his great principle that value is determined by labor, it can hardly be said that he derived it from observation. He seems to have worked it out independently, not taking the
