If this is possible, the advantages are immense. The teacher appears as the friend of the pupil, not as his enemy. The child learns faster, because he is co-operating. He learns with less fatigue, because there is not the constant strain of bringing back a reluctant and bored attention. And his sense of personal initiative is cultivated instead of being diminished. On account of these advantages, it seems worth while to assume that the pupil can be led to learn by the force of his own desires, without the exercise of compulsion by the teacher. If, in a small percentage of cases, the method were found to be a failure, these cases could be isolated and instructed by different methods. But I believe that, given methods adapted to the child’s intelligence, there would be very few failures. For reasons already given in connection with accuracy, I do not believe that a really thorough education can be made interesting through and through. However much one may wish to know a subject, some parts of it are sure to be found dull. But I believe that, given suitable guidance, a boy or girl can be made to feel the importance of learning the dull parts, and can be got through them also without compulsion. I should use the stimulus of praise and blame, applied as the result of good or bad performance of set tasks. Whether a pupil possesses the necessary skill should be made as obvious as in games or gymnastics. And the importance of the dull parts of a subject should be made clear by the teacher. If all these methods failed, the child would have to be classified as stupid, and taught separately from children of normal intelligence, though care must be taken not to let this appear as a punishment. Except in very rare cases, the teacher, even at an early age (_i.e._, after four, say) should not be either parent. Teaching is work requiring a special type of skill, which can be learnt, but which most parents have not had the opportunity of learning. The earlier the age of the pupil, the greater is the pedagogical skill required. And apart from this, the parent has been in constant contact with the child before formal education began, so that the child has a set of habits and expectations towards the parent which are not quite appropriate towards a teacher. The parent, moreover, is likely to be too eager and too much interested in his child’s progress. He will be inordinately pleased by the child’s cleverness and exasperated by his stupidity. There are the same reasons for not teaching one’s own children as have led medical men not to treat their own families. But of course I do not mean that parents should not give such instruction as comes naturally; I mean only that they are, as a rule, not the best people for formal school lessons, even when they are well qualified to teach other people’s children.