right or left by the action of the levers they hold these levers in certain designated positions, nor do they permit them to occupy any other. In this way what is termed the interlocking is effected. The apparatus, though complicated, is simplicity itself compared with a clock or a locomotive. The complication, also, such as it is, arises from the fact that each situation is a problem by itself, and as such has to be studied out and provided for separately. This, however, is a difficulty affecting the manufacturer rather than the operator. To the latter the apparatus presents no difficulty which a fairly intelligent mechanic cannot easily master; while for the former the highly complicated nature of the problem may, perhaps, best be inferred from the example given by Mr. Barry, the simplest that can offer, that of an ordinary junction where a double-track branch-road connects with its double-track main line. There would in this case be of necessity two switch levers and four signal levers, which would admit of sixty-four possible combinations. "The signal might be arranged in any of sixteen ways, and the points might occupy any of four positions, irrespective of the position of the signals. Of the sixty-four combinations thus possible only thirteen are safe, and the rest are such as might lure an engine-driver into danger." [18] A sufficiently popular description of this apparatus also, illustrated by cuts, will be found in Barry's excellent little treatise on _Railway Appliances_, already referred to, published by Longmans & Co. as one of their series of text-books of science. Originally the locking bar was worked through the direct action of certain locks, as they were called, between which the levers when moved played to and fro. These locks were mere bars or plates of iron, some with inclined sides, and others with sides indented or notched. At one end they were secured on a pivot to a fixed bar opposite to and parallel with the movable locking bar, while their other ends were made fast to the locking bar; whence it necessarily followed that, as certain of the levers were pushed to and fro between them, the action of these levers on the inclined sides of the locks could by a skilful combination be made to throw other levers into the notches and indentations of other locks, thus securing them in certain positions, and making it impossible for them to be in any other positions.