The teaching of mathematics and science can only be begun towards the end of the years that we are considering in this chapter--say at the age of twelve. Of course I assume that arithmetic has already been taught, and that there have been popular talks about astronomy and geology, about prehistoric animals, famous explorers, and such naturally interesting matters. But I am thinking now of formal teaching--geometry and algebra, physics and chemistry. A few boys and girls like geometry and algebra, but the great majority do not. I doubt if this is wholly due to faulty methods of teaching. A sense for mathematics, like musical capacity, is mainly a gift of the gods, and I believe it to be quite rare, even in a moderate degree. Nevertheless, every boy and girl should have a taste of mathematics, in order to discover those who have a talent for it. Also, even those who learn little profit by the knowledge that there is such a subject. And by good methods almost everybody can be made to understand the elements of geometry. Of algebra I cannot say the same; it is more abstract than geometry, and essentially unintelligible to those whose minds are incapable of detachment from the concrete. A taste for physics and chemistry, properly taught, would probably be found to be less rare than a taste for mathematics, though still existing only in a minority of young people. Both mathematics and science, in the years from twelve to fourteen, ought only to be pursued to the point at which it becomes clear whether a boy or girl has any aptitude for them. This, of course, is not immediately evident. I loathed algebra at first, although afterwards I had some facility in it. In some cases, it would still be doubtful at the age of fourteen whether there was ability or not. In these cases, tentative methods would have to be continued for a while. But in most cases a decision could be made at fourteen. Some would definitely like the subjects and be good at them, others would dislike them and be bad at them. It would very seldom happen that a clever pupil disliked them or a stupid pupil liked them. What has been said about mathematics and science applies equally to the classics. Between twelve and fourteen, I should give just so much instruction in Latin as would suffice to show which boys and girls had a love of the subject and facility for it. I am assuming that at fourteen education should begin to be more or less specialized, according to the tastes and aptitudes of the pupil. The last years before this moment arrives should be spent in finding out what it will be best to teach in subsequent years.