The establishment of more or less personal relations of confidence between library assistant and reader takes longer and is less complete when the sole intermediary is written language. It is always harder and requires more time to become intimate by letter than by personal intercourse. In the former case the contact is purely mental, in the latter it is affected by personal appearance and conduct, by facial expression and manner. All this is one of the chief factors in the success of the open shelf. But the advantages are not all on the side of the direct personal contact, as the correspondence schools have been astute enough to find out. In the first place, _litera scripta manet_; one may read the same written communication several times, whereas the same spoken communication is of and for the moment. Then the very fact that the written message is purely intellectual and has no physical accompaniments may lend force to its intellectual appeal, when that appeal has once gained a foothold. When this is the case the writer may take his time and may plan his campaign of influence more carefully than the speaker. The effect of trivial circumstances, of unfavorable personal elements, of momentary moods, is obviated. It may be, then, that if personal relations between librarian and reader can be set up through the written word, there may be something of this kind even in long-distance, closed-shelf circulation. This relation may be lacking, even when the circulation is at short range. It is usually lacking at the closed-shelf delivery desk, necessarily so in a rush, although at quieter times there is no good reason why it should not exist. I know that it sometimes does exist under these conditions, though a counter between two human beings, whether in a store, an office or a library, is not conducive to relations of confidence. It may even be lacking in the open-shelf room, when assistants on floor duty have not the proper spirit and a due conception of their own responsibilities and opportunities. It may exist at long range. But does it? I can answer for only one library; but I have no reason to believe that our experience is by any means exceptional. Here are some instances, reported at my request from our own Station Department by Miss Elsie Miller, the department chief: