A good task to set before yourself is to take every idea you agree with in a book and try to treat it as a “germ.” Tell yourself that you will develop it beyond the point where the author left off. Of course this will not always be possible. You will seldom succeed. But there is nothing like hitching your wagon to a star, and it will do no harm to set this up as an ideal. * * * * * A few miscellaneous problems remain to be considered. How should we deal with authors with whom we disagree fundamentally? Herbert Spencer relates that he twice started Kant’s _Critique of Pure Reason_, but disagreeing fundamentally with the first and main proposition he ceased reading. Now to do this is to give an author too much credit for consistency. For even if every other proposition he sets forth is ostensibly a corollary from his leading one, some of them will contain much truth. It is impossible to be consistently wrong. Add to this the possibility that the author may be right on his first proposition after all. However, no book with a viewpoint radically different from our own should be used as a main text, for we would get little benefit from it. If the book is by an obscure author we may safely lay it aside altogether. But if it is by so famous and so bepraised a philosopher as Kant we should at least glance through the entire volume for suggestions. How many times ought we to read a book? I have already partly answered this in formulating the law of diminishing returns. Few books are worth re-reading. Rather than read one book twice on any given subject it will most often be more profitable to read another book on it. For the second will not only serve as a review of previous knowledge, but will furnish you with new ideas, different aspects and new problems. Certain books, however, can never be replaced by others. They occupy this position either because they deal with a subject not elsewhere dealt with or because they take an entirely novel aspect, or solely because they are the works of supreme genius, for while the main conclusions reached in works of this last type may be found elsewhere, the _manner of thinking_ can never be. These books should be read twice. The main text-book selected on any subject will usually be chosen because it is the best and most comprehensive work on that subject. For this reason it should be read a second time even if such reading is only of the hop, skip and jump variety.