and the chapter on the Classification of Fallacies, were drafted in the autumn of the same year; the remainder of the work, in the summer and autumn of 1840. From April following, to the end of 1841, my spare time was devoted to a complete re-writing of the book from its commencement. It is in this way that all my books have been composed. They were always written at least twice over; a first draft of the entire work was completed to the very end of the subject, then the whole begun again de novo; but incorporating, in the second writing, all sentences and parts of sentences of the old draft, which appeared as suitable to my purpose as anything which I could write in lieu of them. I have found great advantages in this system of double redaction. It combines, better than any other mode of composition, the freshness and vigour of the first conception, with the superior precision and completeness resulting from prolonged thought. In my own case, moreover, I have found that the patience necessary for a careful elaboration of the details of composition and expression, costs much less effort after the entire subject has been once gone through, and the substance of all that I find to say has in some manner, however imperfect, been got upon paper. The only thing which I am careful, in the first draft, to make as perfect as I am able, is the arrangement. If that is bad, the whole thread on which the ideas
