511), of the instances which show that he was ill-informed about the advances of knowledge in his own time.

And his mind was often not clear when he came to deal with complex phenomena.
Thus, though he constructed a table of specific gravities--"the only collection," says Mr.Ellis, "of quantitative experiments that we find in his works," and "wonderfully accurate considering the manner in which they were obtained;" yet he failed to understand the real nature of the famous experiment of Archimedes.

And so with the larger features of his teaching it is impossible not to feel how imperfectly he had emancipated himself from the power of words and of common prepossessions; how for one reason or another he had failed to call himself to account in the terms he employed, and the assumptions on which he argued.

The caution does not seem to have occurred to him that the statement of a fact may, in nine cases out of ten, involve a theory.

His whole doctrine of "Forms" and "Simple natures," which is so prominent in his method of investigation, is an example of loose and slovenly use of unexamined and untested ideas.