The Perez Codex was discovered just fifty years ago by Prof. LÉon de Rosny, while searching through the BibliothÈque ImpÉriale, Paris, in the hope of bringing to light some documents of interest for the then newly awakened study of Pre-Columbian America. It was found by him in a basket among a lot of old papers, black with dust and practically abandoned in a chimney corner. From a few words with the name Perez, written on a torn scrap of paper then around it but since lost, it received its name.

Being restored to its proper place in the Library, it was in 1864 photographed by order of M. Victor Duruy, Minister of Instruction, and a few copies issued without further explanatory notes than the printed wrappers. The number of copies is stated by Prof. de Rosny to have been very small; in Leclerc’s Bibl. AmÉr. (1878, No. 2290) it is given as only 10, and in Brasseur’s Bibl. Mex.-Guat. (page 95), as 50. A copy is in the library of the Bureau of Ethnology at Washington, and referred to in their publications as a most fortunate acquisition. I had the good fortune to secure a copy some ten years ago, and one other has recently appeared in a Leipzig catalog at a high price. Beyond these I have not traced any other copy.

In 1872 Prof. de Rosny published a reproduction, drawn by hand, which, as stated by him later, may be disregarded for practical purposes.7-*In 1887 he issued a facsimile edition in colors, 85 copies, which up to the present time has remained the only attempt to show the Codex in its proper colors, and has become exceedingly difficult to procure; so much so that it was only after seven years search that I was able to secure my own copy.8-*

In 1888 he reissued the Codex, uncolored, with the same letter-press, and in an edition of 100 copies. This has also become scarce.

Each of these three editions has its advantages and disadvantages. The colored edition of 1887, having been worked over by hand, in lithography, is defective in various places, both as regards the black of the figures and glyphs, and in the colors. Coloring exists on the original codex which was not reproduced at all in the edition, and the colors given are in many cases not exact. Thus on pages 19 and 20 two different reds are used for the backgrounds, whereas but one is found in the original; on pages 15, 16 the figures are a turquoise green, and on pages 17, 18 an olive green, the correct color for all four being turquoise green.

I have been able to find no inaccuracy in the 1888 edition, which is indeed stated in the introduction to be entirely by mechanical process, without hand intervention; but being reproduced by printer’s ink in black only, not only do the colors not appear, but the chromatic values are actually far inferior to the photographs of 1864. It was stated further by Prof. de Rosny that some features of the MS. had been lost by deterioration in the 25 years previous to his editions of 1887 and 1888, but this I have not been able to verify in any important point.

The photographs and the edition of 1888 are to all general purposes identical; but, notwithstanding that the photographs are steadily yellowing by age, the chromatic values are so far superior that I have continually come to find them the court of final decision in doubtful matters. In a very considerable number of instances a close examination of the photographs has suggested the presence of faint lines of color on glyphs or figures, which was entirely indistinguishable in both of the printed editions, and which was yet in every case confirmed, although sometimes with difficulty, by the examination of the original MS.

The proved value, as well as the scarcity, of these photographs was so great, that in 1905 I had my set photographed twice, by dry and wet plate processes, and a few copies printed after a careful comparison and selection of the two sets of plates. It is from these that the present edition has grown.9-*

The present edition, save for the photographs thus reproduced, having been entirely redrawn, and partly restored, it is fitting to detail just what has been done in this respect.

At the very beginning of my introduction to Maya studies the enormous burdens placed on research therein at every turn, bore upon me as upon every other student. The subject and its possibilities stimulate enthusiasm to the highest degree; the rewards of success are greater than those of any like problem today; and yet, fifty years since the present Codex was discovered, and thirty years since Dr. FÖrstemann’s unsurpassable edition of the Dresden Codex, the actual workers on the problem are the barest handful. A few scattered and obscure references amongst the volumes on volumes of Spanish writers, nearly all untranslated, most of them scarce or almost unprocurable, and many not even printed, make up the literature to be searched out. And a few points of decipherment won and safely fixed by the researchers, from Brasseur, de Rosny, Pousse, Brinton and others a generation ago, to Messrs. Bowditch, Seler, Goodman and a few others of today, are all we have—standing out in a wilderness of guesses by many writers, needless of naming.Of course the prime and absolute necessity of such a study is true facsimiles; but the task of using even these, taken as they must be from much defaced inscriptions and manuscripts, is too obvious for comment. So from the very first of my studies I began to cherish thoughts of the day when Maya could be printed with type, and classified indexes to the glyphs at hand. From one point of view such facilities can only be expected to come after decipherment; from another, in absence of bilingual keys, they are a necessity before that can be attained. So far as his work covers, a great deal has been done in this line by Mr. A. P. Maudslay in the field of the inscriptions.

At the very outset therefore I must enter acknowledgment of the assistance that I owe to the courtesy at that time of Prof. F. W. Putnam, of Peabody Museum, and Mr. Chas. P. Bowditch, in placing, with a freedom by no means universal among curators and researchers, their material at my disposal, with privilege of copying. I am safe to say that while I have reclassified the glyphs for my own use as my studies went on, yet without the copy which by Mr. Bowditch’s courtesy I was allowed to make of his card index to the glyphs of the three codices, as a start, this edition of the Perez Codex would not yet have reached daylight through the many other occupations among which Maya studies have had to take their chances.

At first it seemed possible to prepare a font of separate types for the various elements of the compound glyphs we find in the texts; but after having such a font made a number of years ago, and printing a couple of pages of the Dresden Codex, the result was unsatisfactory; it became evident that the proper Maya font of type must be both separate and composite, as is used in Chinese, and not separate only as we have for Egyptian. The type for the text cards of this edition have therefore been made this way.

As to the colored plates of the Codex herewith, it is evident that nothing whatever is gained by preserving the irregularities of the defaced parts of the Codex, while everything is to be gained by making all as clear and distinct as possible. The first step therefore was to have a set of photographed enlargements of two diameters, made direct from the 1864 issue. From these I made careful tracings, myself, of the black figure and glyph lines of the original, making at the same time the separate enlarged drawings from which the type were afterwards made. At this first drawing only the evident, the indisputable parts were drawn. The type forms were then classified, arranged in parallel columns, and compared. All was then gone over, and new points settled on the basis of the familiarity thus gained. It is a fair estimate to say that this process of checking and verifying was gone through, first to last, down to the final proof-reading of the printed sheets, some fifty times.

One most important fact was established by this process, and must be noted. In the Perez Codex at least, nothing is to be taken for granted, nothing charged to a careless scribe, and no variants regarded as being identical in value—with a very few exceptions, to which I shall advert later. Wherever there remains enough of any glyph to show its characteristic strokes, it can be regarded as safely indicated; whenever the strokes are not just those characteristic of any glyph, it cannot be inferred. Down to the very end of the various revisions I found myself able to add glyphs which at first seemed hopeless, and yet when once seen became clear and plain. Relying on the presence of the photographs to check the work, I have thus added a very considerable number to the glyphs at first apparent. In some cases, as in 6-b-11 and 17, and especially in 8-b-7, 8, 10, where glyphs were only partially erased, but no other instances of perfect glyphs existed to compare them with, I have let them alone, without attempting restoration. In short, I may have made some errors of eye, but I have guessed nothing.

In a very few places I have restored glyphs totally erased, relying on the parallelism of the passages. Such are some of the Ahau-numbers in the upper sections of pages 2 to 11, and in the central sections on those pages, the initial pairs of glyphs on pages 15 to 18-a, b, c, the first columns of pages 19 and 20, and a few day-signs on pages 21, 23 and 24. These glyphs are all necessitated by their different series, and hence can cause no confusions; while it seemed advantageous to have them before the eye. A fair instance of the procedure is shown on page 3-b-1, 3. The temptation was strong to put the usual hieroglyph glyph here as on all the other pages, but the slight variation in the lines left of glyph 3-b-3 forbade it.

The restoration will further be found a little bolder on the type-cards than in the colored plates, where I have in general only endeavored to reproduce what could be seen actually present. The glyphs restored on the upper part of page 7 would seem hopeless at first sight; but they are well-known and common forms, and the characteristic traces shown on the photographs belong to these and to no others known.

The cards of type-printed text, in parallel columns for convenience of study, are self-explanatory. Such an arrangement has from the first seemed to me indispensable for proper study and comparison. The paging of the de Rosny editions I have retained, except to change the practically blank page 1 to be page 25, since to number this as 1 is confusing. For the divisions and the numbering of the glyphs I have made my own arrangement. It is possible that section b on pages 2 to 11 should only go to the bottom line of the central figure, leaving section d to read clear across the page, and another section to be made to the left of the nearly erased figures at the bottom; but the chances as shown by the lining and arrangement of the columns seemed to favor it as I have given it. Only final decipherment can decide definitely.