The coater is a species of wooden tray (of which the diagrams show the section), having a small slit in one of the bottom edges through which the emulsion passes in one even wave the whole width of the plate. The width of the coater is the same as that of the plate, though one six and a half inches wide can be used for either half or whole plates. [Illustration: Fig. 3.] I find the best way of making it, so as to get the slit an equal opening the whole length of it, is to put the back, bottom, and two sides together first, as in Fig. 3. Then by putting a piece of very thin paper (A B) on the angle piece when the front piece of wood is put tight down on the paper and fixed in its place, and the paper is drawn out, it will be found that the slit is very even. In one coater I made I had the slit a little too wide an opening, and to reduce it I glued a piece of muslin over it. This I found was a great improvement, as it not only acted as a strainer, but it checked and caused a more even flow of the emulsion over the plate. I varnished the wood and muslin (except over the slit) with black Japan. To coat the plates I put them close together in rows on the leveling-shelf, as shown below: [Illustration: Fig. 4.] A is a thin, narrow ledge of wood. B B B are thin pieces of wood, in the center of each of which is a small slot and thumb-screw. The plates are pressed against A by the pieces of wood, B, and the thumb-screws are then fastened. The plates are thus kept from slipping about. All this, of course, can be done in ordinary white light. The light is then made non-actinic; the melted emulsion is poured into the reservoir of the coater, which is put to the left hand edge of the outer row of plates. It is then lifted up on edge, as in Fig. 2, and drawn slowly over the row of plates, and so on until the whole of the rows are coated. Of course when not coating plates it is kept in a horizontal position, as in Fig 1. The emulsion on the plates is allowed to set without being disturbed; the shelf is then slipped into the drying-box until the plates are dry, so that they are not touched from the time they are coated until they are dry and ready for packing. I am at present engaged in making a modification of this coater to hold a much larger quantity of emulsion at one time, when a large number of plates require to be coated. It is something the shape of a flat teapot. [Illustration: Fig. 5.--A A is a piece of curved glass. B a piece of coarse ground flat glass, ground side uppermost, sliding in two grooves in the wooden side. C is the handle fixed to the wooden back.]