ral of
contractions, and used an excess of 'tied' letters, which, by the way,
are very useful to the compositor. So I entirely eschewed contractions,
except for the '&,' and had very few tied letters, in fact none but the
absolutely necessary ones. Keeping my end steadily in view, I designed a
black-letter type which I think I may claim to be as readable as a Roman
one, and to say the truth I prefer it to the Roman. This type is of the
size called Great Primer (the Roman type is of 'English' size); but
later on I was driven by the necessities of the Chaucer (a
double-columned book) to get a smaller Gothic type of Pica size.

The punches for all these types, I may mention, were cut for me with
great intelligence and skill by Mr. E. P. Prince, and render my designs
most satisfactorily.

Now as to the spacing: First, the 'face' of the letter should be as
nearly conterminous with the 'body' as possible, so as to avoid undue
whites between the letters. Next, the lateral spaces between the words
should be (a) no more than is necessary to distinguish clearly the
division into words, and (b) should be as nearly equal as possible.
Modern printers, even the best, pay very little heed to these two
essentials of seemly composition, and the inferior ones run riot in
licentious spacing, thereby producing, inter alia, those ugly rivers of
lines running about the page which are such a blemish to decent
printing. Third, the whites between the lines should not be excessive;
the modern practice of 'leading' should be used as little as possible,
and never without some definite reason, such as marking some special
piece of printing. The only leading I have allowed myself is in some
cases a 'thin' lead between the lines of my Gothic pica type: in the
Chaucer and the double-columned books I have used a 'hair' lead, and not
even this in the 16mo books. Lastly, but by no means least, comes the
position of the printed matter on the page. This should always leave the
inner margin the narrowest, the top somewhat wider, the outside
(fore-edge) wider still, and the bottom widest of all. This rule is
never departed from in mediaeval books, written or printed. Modern
printers systematically transgress against it; thus apparently
contradicting the fact that the unit of a book is not one page, but a
pair of pages. A friend, the librarian of one of our most important
private libraries, tells me that after careful testing he has come to
the conclusion t