The scientific position that the first thing to do, in making an investigation, is to find out the facts, has only recently been taken in work of this kind. It has generally been assumed by those who have desired to better conditions of any sort that the existing conditions were well known to all. The fact is that no one person or group of persons is in a position to know all the conditions thoroughly and that the elementary task of ascertaining them and stating them is usually by no means easy. It is now generally recognized that we must have a Survey--an ascertainment and plain statement of the facts as they are--as a preliminary to action or even to discussion. It has seemed to your committee that the general feeling, shared by the educational and industrial worlds, that methods are not always efficiently adjusted to aims should find some place also in the library. We are spending large sums of public money, and investigations by "economy committees," "efficiency bureaus" and the like are taking place all around us. It will be well for us to take a step in advance of these and get for ourselves some sort of a birds-eye view of our work, from the standpoint of its possible lack of complete efficiency--adaptation of end to aim. In order to do this we must first have a survey, which we conceive to involve in this case a statement of just what libraries are trying to do and just how, in some minuteness of detail, they are trying to do it. Comparison and discussion of methods will naturally follow later. The method of taking up this matter was suggested by some very preliminary work done in the St. Louis public library. The head of each of the various branches and departments was asked to make a detailed written list of the various operations performed by the assistants in that particular department, dividing them into purely mechanical acts and those involving some thought or judgment. This in itself proved to be an interesting task and both information and stimulation resulted from it. Certain operations, common to the largest number of kinds of work, were then selected and tests were made, involving both speed of performance and efficiency of result. From a large number of such tests it is expected that some standardization of operations may result, or at any rate the cutting out of useless details and the saving of time for needed extensions of work. The object of an investigation of this kind is of course not to discover ways of making assistants work harder and faster but to find out whether the same amount of work, or more of it, may not be done with less effort.