than blue-black on buff-white, and how much? Is white on
black not better than black on white, and, if so, in what exact degree?
Or is the real solution to be found in some other color contrast as yet
untried? The very mention of some of these possibilities shocks our
prejudices and stirs our conservatism to revolt in advance; yet, with or
against our will, we may be perfectly sure that the changes which
science finally pronounces imperative will be made.

Who can tell what is the normal length of line for legibility, or
whether there is one, and whether there is an ideal size of type, or
what it is? Are the newspapers, for instance, right as to length of line
and the books as to size of type, as many suppose? Has each size of type
a length of line normal to it? How is this affected by leading, or is
leading merely of imaginary value? Is large type desirable for the
schoolbooks of the youngest children, and may the type be made smaller,
down to a certain limit, without harm, as the children grow older, or is
there one ideal size for all ages? It is frankly recognized that in
certain works, like editions of the poets, legibility may properly be
sacrificed in some degree to beauty, and in certain reference works,
again, to economy of space; but we should like to know, as we do not now
with any exactness, what amount of legibility is surrendered.

It is easy, however, to see that one great battleground of controversy
in any suggested reforms must be the design of the type itself. Here,
fortunately, the English public starts with a great advantage. We have
thrown overboard our old black letter with its dazzling contrasts of
shading and its fussy ornament, and therefore can begin where the
Germans must some day leave off. We have no accents or other diacritical
marks, and in this respect are superior to the French also. We start
with a fairly extended and distinct letter like Caslon for our norm, but
even so the problem is in the highest degree complex and baffling.
First, accepting the traditional forms of the letters, we must determine
whether light or heavy, even or shaded, condensed or extended letters
are the more legible, and always in what proportion. We shall then be in
a position to decide the relative standing of the various commercial
types, if such we find, that fairly well meet the conditions. It will
also be obvious what changes can be introduced to improve the types that
stand highest. By and by the limit of i