The question has naturally presented itself in most countries whether it is wise to attempt by law to develop banking on specified lines, recognizing given types of institutions and enacting codes of legislation to control them, or whether it is better to allow custom and practice to differentiate institutions one from another. In the United States we have sought for many years past to legislate in such a way as to define and to develop different types of institutions. Thus there have grown up the national banking system, organized under the National Bank Act; the state banking systems, substantially similar to the national system, but organized under state acts, which vary in detail; the trust companies, organized under state law and differentiated sharply from banks; the savings banks, usually established under legislation worked out in utmost detail; the private banking houses or investment houses, sometimes organized under general law or otherwise organized under legislation specifically intended for their governance; and a variety of other institutions such as rural credit associations, mortgage banks, building-loan associations, and the like. If it can be said that there has been an underlying theory directing this legislation, that theory would probably be held to specify that there should be limitation of functions, and that banking operations should be practiced only along narrow lines by institutions which are confined to a single type of business. This theory has much to sustain it. Credit in the modern world assumes an indefinitely complex form, and it is only a generalization from experience to say that institutions which assume many different kinds of risks are not usually able to judge any one of the risks very accurately or to provide carefully against it. The commercial bank which has a great many deposits and at the same time is lending heavily on real estate or long-term securities is always in danger of having its funds so "tied up" that it cannot free a sufficient amount of them for the purpose of meeting the demands of customers who may call for the redemption of deposits. On the other hand, the institution which is devoting itself chiefly to investment and fiduciary operations is hardly likely to be in a position to study commercial credit carefully unless it organizes a specially equipped department for that purpose. In a country where, as in the United States, the idea of free banking has been tenaciously adhered to, it was almost inevitable that there should be an attempt to protect the public against errors on the part of those who were undertaking the banking business, and one way of guaranteeing such protection was to limit the number of operations in which men of inexperience or poor judgment might engage.

In some foreign countries there has been a tendency to work along the same line partly as the result of law, but more largely as the outcome of custom, and lines of distinction have been drawn almost exactly similar to those which exist in the United States. Nevertheless, when the Federal Reserve Act was passed, it included a provision which allowed national banks to undertake fiduciary functions - that is to say, to parallel and compete with trust companies, and this bad example in legislation has been followed in a number of states. Acting on permission thus granted, more than 1,000 national banks applied for and received permission to engage in trust-company business. Some of them have taken over such business on a comparatively large scale, but experience is already demonstrating what was known before - that such a mixture of functions is a doubtful venture. When successfully undertaken it involves the co-ordinate organization of really separate institutions conducted under a single management, but adhering closely to the dictates of conservatism and experience in the management of their several kinds of business. In the main, it may fairly be said that the tendency of modern nations is away from such combination of functions and is toward specialization.