The designer of the book, therefore, first determines his imposition and then tackles the details of composition. The first principles of composition do not require much discussion since they necessarily follow from the conventions of alphabetical printing in the Roman letter accepted by those for whom we are printing. The matter is relatively simple. First, it is certain that the eye cannot read with ease any considerable number of words composed of letters embodying sharply contrasted thicks and thins; secondly, it is none the less certain that the eye cannot agreeably read a mass of words composed even in a rightly constructed letter, if the lines are beyond a certain length. The most expert reader's eye cannot seize more than a certain number of words in a given size except in a proportionate length of line. Thirdly, practice proves that the size of the letter must be related to the length of line. Respect for these principles will generally protect the reader from the risk of "doubling" (reading the same line twice). The average line of words which the reader's eye can conveniently seize is between ten and twelve. Nevertheless, the typographer, while exerting himself to the utmost to respect this ocular truth, is daily confronted with the fact that unavoidable conditions make it impossible for him to secure a type of the duly related size, and that he is driven to use a relatively small type. To obviate here the risk of "doubling," he consistently inserts proportionate leads through the matter, so opening the lines that the eye comfortably travels and returns from beginning to end and from end to beginning.
The practice of leading, denounced in certain quarters as essentially evil, is an inevitable necessity to a large proportion of printing; and the skilled typographer, making the best use of his material, makes in turn, wise use of leads. The orthodox high-brow view that leads produce in every instance an unhappy weak-looking effect will not survive a wide experience. On the contrary, it will be found that their absence may effectively ruin even a composition in large type, so that it is true to say that the intelligent use of leading distinguishes the expert from the inexpert printer. A slight differentiation of type-face may make the practice advisable. Clearly, while a letter of the size now under the reader's eye, with fairly long ascenders and descenders, would not require leading unless set to a measure of more than 3-1/2 in., there exist letters with short descenders designed rather to sustain leading by rule than by exception. Baskerville's is a type to which leading is invariably an advantage. The problem of determining the amount to be given is not to be settled by considering only the ascenders or the body of the type, because breadth of letter is also a factor to be reckoned with--some letters are narrow in respect to their height, while others are wide. A composition in a round, open, wide letter, chosen because it is rather loose (that is to say, the space between the letters is greater, or appears greater, by reason of the curves of the c, o, e, g), gains in consistency when there is a satisfactory lead between the lines. It is often argued that loose setting is not admirable in itself; to which it might be replied that the printer is generally bound to carry out the instructions of his customer; often to respect the wishes of an artist who may be illustrating the work; and, not seldom, committed by the publisher to a paper-size dictated by irrelevant considerations.