No one can examine Mr. Dugdale's book without being impressed with the quiet unassuming modesty and worth of the author, and yet in the hands of those who have so often carelessly and unscientifically generalized from his studies, it has possibly brought more harm than good.

The error, when there is error, does not arise from generalizing too extensively; that is, from including too wide a range of particular cases in a single proposition.

Of course, in all this I generalize very rashly from a very narrow experience.

My excuse is that these things must be discussed if we are ever to generalize more safely, or to learn that we must not generalize at all.

We all feel the essential unreality of such a conception of 'history' as this; but if such a synechistic pluralism as Peirce, Bergson, and I believe in, be what really exists, every phenomenon of development, even the simplest, would prove equally rebellious to our science should the latter pretend to give us literally accurate instead of approximate, or statistically generalized, pictures of the development of reality.