The recent discussion of the departmental library system at the University of Chicago and the consequent restatement of the position of that university in reference to such libraries, together with the consideration of the problem in the annual reports of Dr. Canfield and Mr. Lane, have called up anew the question of the expediency of the system. Is the departmental library to be a permanent feature of the university library? Is the highest effectiveness of a library to be secured by a policy of decentralization? The public library has answered the question, finally, it seems, in the affirmative. Do the arguments which have induced the public librarian to establish branches and delivery stations apply in the case of the university library? Is the university library of the future to be housed in a single building, or is it to be scattered about in class rooms and laboratories? To my mind, there is no more important question of administration before those of us who are trying to render the university library an efficient instrument of instruction than this. With many librarians there is an element of necessity entering into the question. Mr. Lane is facing a condition where the library has altogether outgrown its building, and some place must be found where books can be stored and used. The situation is much the same in many other places. Shall the facilities of the library be enlarged by building or shall the books be transferred to the various departmental libraries? Mr. Lane, speaking for his own library, says of the latter alternative: "It would commit the library to an entirely different policy from what it has pursued hitherto, and such a change would be little short of a revolution for this library." At the University of Missouri we are expecting in the near future to begin the construction of a library building, but, before adopting any definite plans, we are trying to work out the problems that have just been stated, and to make ourselves reasonably sure that we are right before we go ahead. There are arguments enough on all sides of this question, of which Dr. Canfield says that it has not two sides only, but a dozen. We must premise that no two departments use their books in exactly the same way, and that, consequently, methods of administration must differ. It is generally for the advantage of all, for example, in a university where there is a law school, that the books on private law should be separated from the main collection and treated as a branch library. Similarly medicine, theology and possibly a few other subjects may be withdrawn and administered separately.