The political economist, therefore, may recommend that all such institutions be subjected to the control of the legislature, in order to prevent endowments from being wasted by providing teachers for whom there are no congregations, and that they be also subjected to periodical revision, in order to accommodate the supply of instruction to the demand. He may proceed to consider the different forms of endowments, by tithes, by land, by rent-charges, and by the investment of money. He may show how the first is an obstacle to all improvement, and the second to improvement by the landlord; how the third diminishes with the progress of wealth, and the fourth may perish with the fund on which it is secured. And he may propose remedies for these different inconveniences. If he go further than this, he wanders from the art of wealth into the art of government. I have introduced this rather long illustration, not only as an example of the different modes in which the _art_ of Political Economy must be treated, according to the definition with which the teacher sets out, but also as a specimen of the extent and variety of the details into which he must enter, even if he adopt the less extensive definition. But this is not all. I have already remarked that all the practical arts draw their principles from sciences. If, however, the teacher of an art were to attempt to teach also the different sciences on which it is founded, his treatise would want unity of subject, and be inconveniently long. He generally, therefore, assumes his scientific principles as established, and refers to them as well known. The teacher of the art of medicine merely alludes to the facts which form the sciences of anatomy and chemistry; the teacher of rhetoric assumes that his pupil is acquainted with the science of logic and with that of grammar. Many of the sciences and of the arts which are subservient to the art of Political Economy, may be thus treated. The political economist, for instance, assumes that protection from domestic or foreign violence or fraud, is essential to any considerable production or accumulation of wealth, and he considers the means by which the expense of providing this protection may be best supported; but he does not inquire what are the necessary legal and military institutions. He leaves these to be pointed out by the arts of war and of penal and civil jurisprudence, and by the sciences on which those arts depend.