were hand-lettered and not taken from metal originals.
‘There is a fate about this,’ Colvin said after I had explained my mission. ‘We have here in the Museum six original drawings of Petrarch’s Triumphs, attributed by some to Fra Lippo Lippi and certainly belonging to his school, which have never been reproduced. They are exactly the right size for the format which you have determined upon, and if you can have the reproductions made here at the Museum the drawings are at your disposal.’
I made arrangements with Emery Walker, the designer of the Doves type and famous as an engraver, to etch these six plates on steel, and the reproductions of the originals were extraordinarily exact. Those made for the parchment edition looked as if drawn on ivory.
Parchment was required for the specially illuminated copies which were to form a feature of the edition, and before leaving America I had been told that the Roman was the best. I naturally assumed that I should find this in Rome, but investigation developed the fact that Roman parchment is prepared in Florence. Following this lead, I examined the skins sold by Floren-tine dealers, but Dr. Biagi assured me that the best grade was not Roman but Florentine, and that Florentine parchment is produced in Issoudun, France. It seemed a far cry to seek out Italian skins in France, but to Issoudun I went. In the meantime I learned that there was a still better grade prepared in Brentford, England—this, in fact, being where William Morris procured the parchment for his Kelmscott publications.
At Brentford I secured my skins; and here I learned something that interested me exceedingly. Owing to the oil which remains in the parchment after it has been prepared for use, the difficulty in printing is as great as if on glass. To obviate this, the concern at Brentford, in preparing parchment for the Kelmscott volumes, filled in the pores of the skins with chalk, producing an artificial surface. The process of time must operate adversely upon this extraneous substance, and the question naturally arises as to whether eventually, in the Kelmscott parchment volumes, the chalk surface will flake off in spots, producing blemishes which can never be repaired.
For my own purposes I purchased the skins in their original condition, and overcame the difficulty in printing by a treatment of the ink which, after much experiment, enabled me to secure as fine results upon the parchment as if printing upon handmade paper.
The volumes were to, be printed in the two humanistic colors, black and blue. In the original manuscript volumes this blue is a most unusual shade, the hand-letterer having prepared his own ink by grinding lapis lazuli, in which there is no red. By artificial light the lines written in blue cannot be distinguished from the black. To reproduce the same effect in the printed volume, I secured in Florence a limited quantity of lapis lazuli, and by special arrangement with the Italian Government had it crushed into powder at the royal mint. This powder I brought home to America, and I arranged with a leading manufacturer to produce what I believe to be the first printing-ink mixed exactly as the scribes of the fifteenth century used to prepare their pigments.
The months required to produce the Triumphs represented a period alternating in anxiety and satisfaction. The greatest difficulty came in impressing upon the typesetter the fact that the various characters of these letters could not be used with mathematical precision, but that the change should come