Figures 2 and 3 represent the interior sides of the mould. For the purpose of clearer illustration, the half of the mould, Figure 2, is shown reversed, or upside down; but when this half is connected with its mate, the two halves appear as they do in Figure 1. These two halves differ from each other only in a few minor features. They are so constructed that, when joined, the sides which determine the body of the types are in exact parallel, and at a-fixed and unalterable distance from each other. In Figure 2, the ridges which make the nicks are noticeable; in Figure 3 the cast type is shown as it appears before it is thrown from the mould, with jet attached.[12] Although the two sides of the mould are fixed so as to be immovable in the direction which determines the body of the type, they have great freedom of motion and nicety of adjustment in the direction which determines its width. They can be brought close together, so as to make a hair space, or can be fixed wide apart, so as to cast a three-em quadrat, but they always slide on broad and solid bearings, between guides which keep them from getting out of square. In the construction of the mould and adjustment of the matrices, every care is taken to insure exactness of body. The illustration on page 52 may be again referred to as an example of the necessity for minute accuracy. We there see that the feasibility of typography depends upon the geometrical exactness of its tools, and that types are of no practical use, if they cannot be readily combined and interchanged. The casting or founding of types, in a mould constructed like that of the engraving, is now accomplished by a complex machine, the invention of Mr. David Bruce, Jr., of New-York city, and by him patented in the year 1838. Before this date [p059] all types were cast by hand, from a hand-mould, and by a process which received no noticeable improvement for two centuries. The following illustration, taken from an engraving published by an early English type-founder,[13] can be offered as a substantially correct representation of the method of casting which was practised by all type-founders in the first quarter of this century. [Illustration: Type-Casting as Practised in 1683. [From Moxon.]]