Introduction to the Corpus

The Peabody Museum has had a long and distinguished history of research in Middle American archaeology and on the Maya, including their hiero­glyphic writing. Charles P. Bowditch, grandson of the famous navigator, became interested in Maya archaeology in the 1880s. In 1888, he visited Yucatan, engaging Edward H. Thompson to carry out archaeological sur­veys for the Peabody Museum. Subsequently, he planned and made possi­ble the Museum's excavation at Copan, Honduras.  He was an avid scholar of Maya glyphs and calendrics and published an important book and many articles on the subject.

Bowditch's most distinguished  protege, Alfred M. Tozzer, was intro­duced to the Maya area in 1902. In his early years, he carried out both ethnographic and archaeological investigations in the jungles, studying the Lacandon Maya of Chiapas at Bowditch's suggestion in the hope of finding residual knowledge of the hieroglyphic system. He mapped and explored such ruins as Tikal, Nakum, and Holmul. Perhaps his greatest contribution was the fact that he kept alive the study of Maya hieroglyphics in the United States, training eminent students in this field, including Sylvanus G. Mor­ley.

The Museum also has done much to bring the Maya and their culture to the broader attention of the public. In 1912, under Mr. Tozzer's direction, the Peabody prepared for the Boston Museum of Fine Arts an exhibition of Middle American art, mainly of the Maya, that was the first of its kind in America. Later, in 1940, another important exhibition was cooperatively held at the Fogg Museum, and in 1958, a number of significant Maya pieces were shown at the Museum of Fine Arts as part of a more general Primitive Arts exhibition cosponsored by the Peabody. In 1971, the Peabody Museum and the Center for Inter-American Relations jointly sponsored in New York City and then in Cambridge an interesting exhibition on the art of Maya hieroglyphic writing. It helped put the undeciphered Maya hieroglyphic language in proper perspective, thanks to the efforts of Ian Graham who organized the show and wrote the catalogue. Besides describing some of the methods and problems of the current study, and showing the beauty of some of the best original Maya monuments and inscriptions as self­ reliant works of art, the exhibition also touched on the tragic problem of the death and destruction that is ravaging so many Maya sites in the name of art.

If any field can be said to have a conscience, surely Ian Graham has been that for the area of research in Maya hieroglyphics. But more than just lamenting the havoc that looters have wreaked upon the jungle fastnesses which hold these precious documents, Ian Graham has almost single­ handedly sought to preserve and pass the hieroglyphic texts on to future generations of scholars. It is a monumental task, one which Charles P. Bowditch would have applauded, and we at the Museum are honored to help this immense project on its way. It is thus with deep satisfaction that we inaugurate the publication of the Corpus, hoping and knowing that in the years ahead other scholars in other lands will be able to carry forward major research in Maya hieroglyphics as a result of this work.

By choosing for this work the title Corpus, rather than Collection or even Sylloge, homage is done to August Boeckh, the  pioneer who initiated in 1828 the great Corpus Inscriptionum Graecarum; today, volumes in the series that stemmed directly from it still continue to appear, nearly one hundred and fifty years later.

Our title sounds an echo, too, of Sylvanus Morley, another tireless com­piler, who had without doubt the C.I.G. in mind when he called his mimeographed checklist of inscriptions the Corpuslnscriptionum Mayarum.

In the present work, Latin has been abandoned even for the title, but makes an appearance in the dedication: for I have borrowed a phrase of Boeckh's, one that he applied to F. A. Wolf, his revered mentor.

Although completion of this Corpus is anticipated in something less than a century, the enterprise is still grandly conceived, and honor is due to the man primarily responsible for its inception, Edgar H. Brenner. In 1968 Mr. Brenner, a lawyer and amateur of Maya studies, was able to persuade the Stella and Charles Guttman Foundation of New York City, of which he was a trustee, to commission a preliminary study that called for recommendations as to the form, content, and technical methods to be employed in compiling a Corpus, and an estimate of the magnitude of the undertaking.

The task of administering this pilot study was taken up by the Center for Inter-American Relations of New York City, which appointed an Advisory Committee for the purpose. In the course of its existence this committee was served by the following members: Dr. Ignacio Bernal, then director of the Instituto Nacional de Antropologfa e Historia, Mexico; Mr. Edgar H. Brenner, Stella and Charles Guttman Foundation; Mr. Stanton L. Catlin, Center for Inter-American Relations; Dr. Michael D. Coe, Yale University; Dr. Gordon F. Ekholm, American Museum of Natural History; Dr. Luis Lujan Munoz, director, Instituto Nacional de Antropologfa e Historia, Guatemala; Dr. Floyd Lounsbury, Yale University; Miss Tatiana Proskouriakoff and Dr. Gordon R. Willey, of the Peabody Museum, and Dr. Stephen Williams, its director.

It was then my good fortune to be asked by the committee to undertake this study. Having completed it, I delivered my report in September, 1969. Early in 1970 the question of an institutional base was resolved when the Peabody Museum offered to take on responsibility for the project. It was a most satisfactory arrangement, in view of the unrivaled photographic archives of Maya sculpture that are preserved in the Museum, incorporating as they do those of the old Department of Archaeology of the Carnegie Institution of Washington.

The Guttman Foundation having pledged further support, an application for matching funds was forwarded to the National Endowment for the Humanities. This was granted for the years 1971 and 1972. At this stage Dr. Eric von Euw joined the project and began his initiation into the twin mysteries of hieroglyphics and the management of mules.

And now, as publication of the Corpus begins, there are features of its design and execution that I wish to describe. Those meaning to make use of the work may be anxious to know what its scope will be, how it is or­ganized, and what techniques and standards have been adopted. These are not necessarily self-evident.

This can be stated quite briefly. First, as to the media in which Maya texts are preserved, all of them except codices and pottery vessels are taken to fall within the purview of the Corpus. Thus, inscriptions on jade, shell, bone, wood, stucco, and painted walls will be covered, in addition to those carved in stone, which form the largest category.

Second, no precise geographical limits have been set; any object that carries writing of predominantly Maya character comes within the scope of the Corpus.

Third, where hieroglyphs are accompanied by representational art, or accompany it, the entire design will be recorded. In a very few cases an object that bears no inscription will be included: one member of a set of lintels, for example, when other members of the set do carry inscriptions.

We do not propose to offer any commentary on the inscriptions, since to embark on even the barest analysis would seriously hold back progress of fieldwork and publication. Other factors weighed in this decision have been the increase in bulk and in price that a commentary would impose on these volumes, and the reflection that a large proportion of this additional matter inevitably would grow obsolete and  become an encumbrance in a still vital work of reference.

It is not our intention, either, to include within these volumes any account of the minor excavations that are occasionally necessary in recording a monument, nor will there be notes on architecture or sherd collections gathered during our work at the various sites. These will appear in another place.

A suggestion made by Mr. Brenner at an early stage was for a loose-leaf Corpus. This would allow the user to bring together all the pages concerning a single site, some of which might, for various reasons, have had to be issued over a span of years, scattered therefore among several volumes. The possibility also of extracting pages out of their binders for use in a particular study would be of further help to the user, by enabling him to eliminate from his desk an unwieldy pile of books.

It was agreed that this idea was likely to offer real benefits, especially to individual owners of the work (for we rather doubt that librarians will warm to a loose-leaf publication in several volumes). Thus the Corpus is designed for alternative forms of use. As issued, the pages are bound; but the binding can be cut off and the leaves kept in spring binders if this is preferred.

The decision to offer this option has imposed certain features on the design. The most important is the allocation of two pages, the recto and verso of one leaf, to each monument or other object. In some cases more than one leaf will have to be assigned to a monument. Had only a perma­nent binding been in prospect, then it would have been logical to use pairs of facing pages, placing the photographs on one and the drawings on the other. As it is, it will be seen that in most of the layouts a photograph and the drawing corresponding to it are printed on the same page, as it is most important to permit easy comparison of the two.

The large page size was dictated by the decision to accommodate the photographs of all monuments, save the most exceptionally tall stelae, at a fixed scale of 1:10 without resorting to foldouts; also by the need, just mentioned, to fit pairs of drawings and photographs side by side on the page - again, with some unavoidable exceptions.

At the head of each page two numbers will be found, printed in Arabic numerals and separated by a colon. These will be the numbers to cite in references. Here it is opportune to suggest that in all references to the Corpus, the compilers' names be omitted and the abbreviation CMHI be used, followed directly by the complete page number, and without date.

The method by which the owner of loose leaves arranges them is for him to decide: sites can be grouped alphabetically, by geographical areas, or by any other criterion. For this mode of use, each leaf is identified by the code printed above the page number; this consists of the abbreviated site name, followed either by Roman numerals for the preliminary matter, map, site plan, etc., or by the monument or object designation. The method of finding a loose leaf from a volume-and-page reference is described in a note below.¹

In composition, many of these volumes will be somewhat variegated: unavoidably so, unless the beginning of publication were to be delayed until the entire field program was completed. Ideally, each volume would consist wholly of monuments from one large site or would provide com­plete coverage of smaller sites within a well-defined area. Not many vol­umes, perhaps, will approach this ideal, but in none shall we veer to the opposite extreme of offering a miscellany of material that chances to be ready.

All volumes with the exception of the present one will be issued in three parts, or fascicles. By this means we hope to avoid the delays that might otherwise be engendered by the need to accumulate enough material from one geographical area to complete a whole volume; flexibility in planning field operations is in the same degree preserved. For our plan is to concen­trate, in each of these volumes, upon one of five sectors of the Maya area, namely Yucatan, Central Lowlands, Chiapas and Usumacinta River, High­lands and Pacific Coast, and Lower Motagua Drainage. Volumes devoted to these areas will not appear in any regular sequence. Within a volume, the coverage of each fascicle will be more narrowly confined to subareas such as the Rio de la Pasion drainage, or northwestern Yucatan.

One modest service which the Corpus may perform is to bring uniformity to the nomenclature of glyph-blocks, and monuments. I would have added: site names too, if there had been any hope of settling at one stroke all of those more controversial questions such as the name most properly applied to a site (Moral or Morales; Benque Viejo or Xunantunich), or even the best spelling.² As to glyph-blocks, the general policy will be to letter and number each one, as far as possible in conformity with existing nomenclature; an abbreviated form of reference to all monument types and sites will also be suggested.³ We hope that this compact terminology may be found convenient for use in tabulations. The names of all sites with inscriptions on stone have been reduced to three-letter codes; these and abbreviations for monument types are listed in appendix A. A monument can thus be specified by citing the code in capital letters followed by a colon, next the suggested abbreviation for the monument type followed by a period, and lastly the number of the monument. This may be followed, after a comma, by the glyph-block designation; for example, YAX:Lnt.48,Al-B2.

A listing of sites and their codes will be found in appendix A. It will be reprinted with any necessary additions in every third volume, together with an index covering that volume and all those preceding. Objects lacking provenience are numbered in a separate class, "Collections," and such pieces will be placed in fascicles that seem appropriate to them on the basis of likely origin.

Although the scale of reproduction for the principal photograph of the entire object will be with very few exceptions 1:10, and usually the same for the drawing, for sculpture having very fine detail, or glyphs of small size, the scale of the drawing will be increased. For the epigrapher, one of the advantages of line drawings is that clear photocopies of them are easily made which may then be cut up and rearranged at will; they are less satisfactory, however, when the glyph-blocks come out less than one or one-and-a-half centimeters in height. As an extreme case, the hieroglyphs on Lintel 2 from Piedras Negras may be cited, which if reproduced at 1:10 would measure 2.8 by 3.4 millimeters.

Since the beginning of the project three guiding principles have been regarded as dominant: accuracy, clarity, and comprehensiveness. It soon became apparent that the difficulties standing in the way of meeting these objectives are very unequal. Comprehensive coverage depends on continu­ing financial support more than on any other factor; clarity can be ensured through the use of good photographic technique and the preparation of line drawings; the crucial difficulty is presented by the third objective, accura­cy. Here one is compelled to recognize that there is no possibility of attaining perfect accuracy in transcribing our material, and this fact should be stressed. But accuracy does remain the goal that one constantly strives to approach.⁴

As the first step in this approach, the decision was made to eschew, as much too perilous, all attempts to restore worn or missing areas of sculpture. This is best left to the reader. Yet it would be a mistake to suppose that by confining one's attention to that which still exists, the process of selection or discrimination has been eliminated. Far from it: just as correct delineation of surviving traces of relief in a sculpture is an obvious requirement, so too is omission of all elements that owe their existence to quirks of erosion. Here discrimination is called for, and the recurring necessity of distinguishing between the authentic and the accidental, or spurious, soon forces those engaged in recording weathered inscriptions to acknowledge their near-illiteracy in hieroglyphics as a se­vere handicap. Study of faint lines on a pitted surface, or a pattern of eroded hollows, summons up half unconsciously a review of known glyphs before the mind's eye in the quest of an identification. All too often there is failure to recognize any known design, least of all one that lends reassurance by being known to fit the context. And yet there remains always the possibility that in the light of future research the same blurred details may provide a reading that will seem by then quite obvious. It is therefore imperative to record all such weathered hieroglyphs as effectively as possible.⁵

The certain knowledge that our record will be flawed by many errors makes us, we confess, uncomfortably conscious of responsibilities. But we hasten to disclaim definitive status for these drawings; they are no more than convenient guides, and the epigrapher is not obliged in the last resort to depend on them. The photographs and drawings published here are not presented on a take-it-or-leave-it basis; instead, like the tip of an iceberg, they are supported by a submerged mass of unpublished data that is accumulating in the Peabody Museum, available to those who have specific points to clarify. We regard formation of this archive as a goal no less important than that of publishing printed volumes.

The scheme of measurements given for stelae and the abbreviated notation employed in stating all dimensions are explained in appendix B. The scope of this information may be thought meager; scarcely more ambitious in­ deed than that of Wordsworth, in his report on a small excavation in The Thorn (1798 edition): "I've measured it from side to side /'Tis three feet long, and two feet wide."

Since the location of most Maya sites is not known with any precision at the time of writing, it does not seem appropriate at this stage to publish a map of the whole Maya area marking the sites. However, during their visits to each site the compilers invariably attempt to establish its true location, and this will be shown on the map to be found in the introductory pages for that site.⁶ These maps are all drawn at the scale of 1:125,000; possibly this scale will seem outlandish to professional cartographers, but it has the advantage of facilitating the transfer of a given location onto the official maps of Mexico, Guatemala, and Belize. These are either at 1:250,000 (Guatemala and Belize) or 1:500,000 (Mexico), so that it is only necessary to measure the intercepts of the site from the given meridian and parallel of latitude, and halve or quarter them as the case may be, to plot it on those smaller-scale maps.

A scale of 1:2,000 has been chosen for site plans. Whenever accurate plans are already available they will be republished, sometimes redrafted, and if necessary simplified or reduced in coverage. But in most cases plans will have been drawn from compass-and-tape surveys made expressly by the compilers, supplemented occasionally by data from plans published by other workers. Readers do not, perhaps, require warning that a high degree of accuracy should not be expected in such plans, as the primary purpose of our undertaking-gathering inscriptions before  they wilt-precludes us from devoting weeks or months to the careful survey of large sites. The most obvious reflection of these economies in effort shows in the substitution of true contours by form-lines that merely give an impression of the terrain.

The symbols employed in maps and plans are given in appendix B.

The number of references to other works that will be made in these volumes is expected to be quite small. This being so, and in view of the progressive and necessarily disordered sequence of publication and the alternative systems of binding that are proposed, the problem of where to locate bibliographic references that occur on pages devoted to monuments is best solved by placing them on the same page, either in the text itself, or as footnotes. On these pages mention of Maler, Morley, and Maudslay will usually not be supported by bibliographic references when the passages concerned are to be found in the Peabody Museum Memoirs, the Inscriptions of Peten, and the Biologia Centrali-Americana, in the section of the work devoted to the monument concerned, that is to say, in the most obvious place.

1. It is recommended that those who have opted for loose-leaf usage keep a separate binder for the tables of contents and indexes of all volumes and the most recently issued list of site-codes. If a volume-and-page reference is to be looked up, it will be simple to identify the site and object from the tables of contents and then to find the page in whichever binder the material from that site has been assembled.

At Chichen Itza, the majority of inscriptions are designated by descriptive names-Casa Colora­da, Temple of the Four Lintels, etc. - and in varying degrees the same is true of some other sites. These do not fall into any natural order, as do monuments at sites where only stelae and altars are to be found, all in numbered series. If, then, we are to avoid confusion and waste of time in searching for a particular page, a filing order must be established for the monuments of every site. This is provided by the Register of Inscriptions that comes at the end of the introductory matter for each site.

2. In choosing between alternative names or spellings, preference has usually been given to the more commonly accepted form, rather than versions that might be pedantically correct. Thus, Seibal with an s is retained, as established in the literature too firmly for change at this late date. In any case, by the rules of Spanish orthography this spelling ought not to be objectionable for a word of Caribbean origin, and it is in fact employed in the place names Seiba Mocha, Cuba, and Seiba Playa, Yucatan.

Accents have been omitted from all Maya place names. In many of these names h has been retained for aspirates rather than the more guttural Spanish jota, unless well-established usage dictates otherwise, as in Kaminaljuyu. The only substantial change concerns the site previously known as Yaltitud. This mistakenly Hispanicized version has been discarded in favor of Yaltutu, a name that can be found for several localities in Peten (apparently it refers to an abundance of freshwater shellfish at those places).

In the designation of glyph-blocks some difficulties arise. One of them is presented by fragmentary sculpture in which the beginning of the text is missing: the normal scheme of letters and numbers cannot be applied, starting from the first surviving glyph, because subsequent discovery of another fragment is likely to make necessary an entirely new designation, and this is at all costs to be avoided. The policy to be followed in the Corpus for fragmentary texts of this kind that do not already have an established designation will be to add the letter p (for provisional) as a prefix to each column-letter or row-number, if the possibility exists of preceding columns or rows having been lost. Thus, on a fragment that lacks its original left-hand edge, the numbering will begin with pA1, pB1, etc. If there is doubt only about the original number of rows, as with a full-width lower fragment of an all-glyphic panel, the p will precede the number only, in this fashion: Ap1, Bp1, etc. By this notation the provisional nature of the designation is made clear.

If a revised, but still provisional, glyph-block designation is called for upon the discovery of a further fragment which still does not supply the beginning of the text, it can be distinguished by the use of the letter q, instead of p.

Another difficulty lies in the implication that an alphabetic and numerical series indicates the correct order of reading. Indeed, more often than not it does. Morley's very serviceable usage was to look for the opening date, begin the lettering with the column in which this statement commenced, and proceed from left to right. However, an isolated panel of glyphs may be a source of difficulty. Sometimes it can be recognized as a continuation of a column that has merely been interrupted by some nonhieroglyphic element of design; in other cases it clearly does not form part of the main text, but could be a caption relating to an ancillary figure or some other kind of subsidiary matter.

To provide for this situation, a scheme has been devised by Linton Satterthwaite, which is being applied where necessary to monuments at Tikal. The main text receives the conventional designa­tion. Subsidiary panels or groups are identified by letters from the end of the alphabet, and glyphs within them are designated in the ordinary way, except that the column-letters A, B, C, etc., are prefixed with the letter given to that group, in lower case. An illustration is provided by Altar 5 at Tikal. Two groups not labeled by Morley (a row and a panel) lie within the encircling glyphic band. These are now designated Panel Y and Panel Z, respectively, and they contain the glyph-blocks yA1-yA4, and zA1, zA2.

Well-conceived though this system of nomenclature clearly is, the compilers of the Corpus have decided not to adopt it when it falls to them to label glyph-blocks. The additional complexity, and the danger of erroneously incorporating a group into, or divorcing it from, the main text seem to us to outweigh the advantages. Nevertheless, designations will be applied in such a way that they correspond as far as possible with the apparent order of reading. Users of the Corpus are urged to regard these designations in the first instance as mere locus-indicators of the blocks in illustrations and then, in each particular case, to decide whether--or to what extent--they are valid guides to the order of reading the contents of the blocks.

Unfortunately, codes conforming with any uniform system of site designation (Rowe 1971) could not be employed, principally because the only justification for the codes in this context--their brevity--would be lost, and also because in many cases insufficient geographical data are available. The codes devised here have the great advantage of being derived from the common names of sites, and so are usually easy to recognize. There are a few instances in which "unnatural" code letters have had to be chosen to avoid confusion among similar site names. However, one ambiguous code, YAX, has been allowed to stand for Yaxchilan, in spite of the possibility of mistaking it for Yaxha, on the grounds that the great volume and importance of the sculpture at Yaxchilan justifies the choice for it of the most obvious code.

4. The various techniques and problems of photography will not be discussed here, chiefly because they are not static but evolve from one year to the next. Indeed, I will take the opportunity to express my hope that an improvement will be noticed in the quality of photographs from volume 3 onward, compared with those taken by me for volume 2; if so, this can be attributed to a change in lighting technique. A discussion of the whole subject is contemplated for publication elsewhere. For the moment it should be enough to state that photographs are generally taken by artificial light, arranged to fall from various directions in successive exposures, the lens axis always being kept perpendicular to the sculptured surlace.

About the technique of drawing there is perhaps less that can be said, and there is little to be expected in the way of technical developments during the progress of the Corpus. But there are comments to make that do have some bearing upon the interpretation of these drawings.

Our drawings are made on Mylar polyester film and are based on tracings from photographs printed at the scale of 1:4 (or larger, when the drawing is to be reproduced at a scale greater than 1:10). For details to be incorporated in the rendering, heavy reliance is placed on pencil drawings that have been made in the presence of the original; these field drawings are always checked over at night with electric light arranged to fall on the stone at raking angles and from various directions in turn, so as to bring up the faintest remains of relief. Figure 1 is presented here to show how different features of a carved surface are revealed by lighting from different directions; also to point up the need to insure a strictly frontal camera position so as to avoid distortion.

Apart from field drawings and photographs, other resources that may be utilized in the rendering include plastercasts, these being especially valuable in the case of a lintel in situ or a sculpture having a curved surface (e.g., Naranjo Stela 35), and old photographs that may show sculptural detail or even whole fragments now lost.

Broadly speaking, there are two ways of representing relief with pen or pencil on paper. One technique aims at lifelike effect, simulating the texture of stone and the play of light upon it by skillful use of stipple or hachure; the other is a line drawing that does not pretend to be more than schematic, suggesting neither plasticity and tactile value nor illumination from any particular direction. This is the technique adopted for the present work. Besides demanding less artistic talent, the line drawing has the advantage of being able to convey more information in a given area. This is important because in the corpus of Maya sculpture there are pieces so finely carved as to make realistic rendering at a much reduced scale impracticable, if detail is not to be lost. Even in a line drawing, there is difficulty with crosshatched areas in which the lines may be spaced less than two millimeters apart on the original or half a millimeter at drafting scale, and one fifth of a millimeter at reproduction scale.

In the common sort of Maya bas-relief, the convention appears to be that figures stand against a featureless wall, or a clear sky, this being represented by a recessed surface. The background of hieroglyphs, although they maybe ona raised panel or isolated in a frame, is likewise plain. Where a glyph is pierced with an opening, this is recessed to the general background level. Thus, a line drawing can be rendered more intelligible if the background is shown stippled. In the Corpus the density of stippling is meant to indicate the probability that no feature was ever carved in that area. For example, if in the area surrounding a headdress part of the plain background has been badly pitted, a less dense stippling will be applied there, as the former existence of some detail, such as a pair of incised glyphs or a feather, cannot be ruled out entirely. Additionally, it is a useful feature of the stippled background that it can be made to show the eroded outline of a glyph or other element with any degree of vagueness desired.

Occasionally, difficulties are met with in the application of stippling. The commonest and most serious doubt arises over small recessed areas: do they represent holes in the object portrayed or merely recessed features in it? Where there is uncertainty the stippling is best left out, and the area can be shown as recessed in the following manner: a close look at any stippled background in this work will show that the line enclosing it, which represents a projecting rim of relief, is strengthened with closely spaced dots along the low, or stippled, side. Equally, in the absence of stippling, a similar row of dots touching a line on one side signifies that on that side the surface recedes. This notation, however, will be employed only where doubt is likely to arise over the interpretation of a passage.

As far as possible, the stylistic character of the original is preserved in the line drawing, a conscious effort having been made to suppress any tendency to smooth out erratic curves or eliminate other signs of poor workmanship. However, line drawings are deficient in one important respect, for they do not convey the particular quality of roundness, or lack of it, in the relief. To give an example, the cross section of the cartouche or rim of a glyph is often as demonstrated in figure 2. Certain sculptors, however, rejecting the use of an incised inner line, worked the stone with evident care into another cross section (figure 3). Differentiating between these two in a drawing at small scale whatever the technique of drawing, would be a test of skill for any draftsman, but in a line drawing the task is plainly impossible. Therefore, information about the quality of relief has to be sought in the photographs (see also note 5).

To represent a line whose course or even existence is uncertain we use a dotted line, with the degree of uncertainty expressed by the spacing of the dots. Solid lines are used only for features about whose form there is no doubt. On the face of it, this is a simple enough principle; but one that can scarcely be maintained rigorously. The draftsman (to his dismay) is repeatedly confronted by evidence of his own overconfidence, the most chastening experience being to finish drawing a rather well-preserved sculpture and then discover an early photograph showing the piece in pristine condition. Many a solid line is found to have minor, but possibly significant, defects.

In order to distinguish between lines that come to a definite end and others that are interrupted by erosion or breakage, the latter are terminated in the drawing by a dot, which may not be noticeable without the use of a magnifying glass.

A minor problem in representing relief sculpture by line drawings is presented by the displacement of lines that erosion can cause. While it would be pretentious to claim that lines are drawn to correspond with any precisely determined point on the shoulder of the relief, it may be said that the point is about halfway up the slope, or a little more. Erosion can cause displacement in either direction. If the raised areas have been preferentially eroded, leaving the recessed background unaffected--this may be the result of flaking or other processes not understood; Naranjo Stela 23 provides a good example--then the foot of the relief will remain perfectly clear, justifying even a solid line in the drawing; but this line will have been displaced outwards (b in figure 4). In contrast, the acid decomposition products of mosses and humus attack stone more evenly, with a consequent shift inwards of the apparent line (c in figure 4). In consequence, a lone feather or other narrow element will show quite distinctly, but it will have suffered considerable attenuation. Perhaps even more noticeable is the gap that develops between elements such as a glyphic mainsign and affixes that once made contact with it.

Dots, represented by drilled holes in sculpture, are a common element in Maya designs and in weathered stone are easily confused with erosion pits. Undoubted dots, then, are represented in these drawings by heavy spots, and doubtful ones by three small dots, whereas a circlet of uncertain authenticity is represented by five or more dots in a circle. An effort is made always to depict the correct number of dots or beads in figurative art and in hieroglyphic affixes such as T-32 or T-36. The number of lines, however, in a crosshatched area is not so carefully reproduced, as being less likely to have significance.

5. Glyphs carved in a fine-grained stone, with interior detail lightly carved or incised, weather down to fairly clear outlines enclosing totally blank interiors. These present no problem of representation. The difficult case is the glyph deeply carved in a stone of coarse and uneven composition. Here, if a significant element is to be recognized among the lumps and hollows largely caused by erosion, that glyph must almost have been expected for that position in the text. Glyph combinations such as Emblem Glyphs, Imix-comb-Imix glyphs, and members of known clauses can often be spotted when neighboring glyphs no worse eroded have to be shown blank in the drawing. In such cases the better prepared epigrapher of the future is likely to feel betrayed by the publication of only a single photograph and a line drawing.

The solution seems to lie in three-dimensional photography. In the first fascicle to be issued, volume 2, part 1, stereophotographs are supplied only for the glyph-panels on the front of Stelae 22 and 23 from Naranjo, but in later volumes they will be more liberally provided. These stereophoto­ graphs are meant to be examined with the standard pocket viewer manufactured for use with aerial photographs.

It has been suggested that if stereophotographs are to be taken, then the further step ought to be made up to full photogrammetry. This seems to me a doubtful proposition: the cost of stereo-plotting is very great, and it is not clear in the present context what benefits the technique has to offer--even though claims have been made for it in connection with a problem not encountered in Maya sculpture, the separation of palimpsest inscriptions (Silva and Fernando 1971).

6. Determining the latitude and longitude of ruins in the forest by means of transit and chronometer, as was attempted on Carnegie Institution expeditions in the 1920s and 1930s, is attended with certain difficulties. In those days a source of serious error lay in the chronometer, but this has now been almost eliminated: wristwatches of previously unattainable accuracy are available, and they can be set by time signals received over a portable radio. But another impediment remains. The altitude of the Pole Star is so low in the latitudes of Maya sites that observations are possible only if quite a long clearing exists, extending north and south, or if there is a tall pyramid cleared of trees round the top. In the course of a season's formal excavation at a site, one can sometimes arrange without too much difficulty for one of these conditions to be met, but clearing trees is seldom possible during a visit made for the sole purpose of recording sculpture.

Another method has therefore to be used. Generally the only one that is feasible is to make a ground traverse from a known station, and to correlate the plot later with stereoscopic aerial photographs. Along the trail, compass readings to the nearest five degrees are taken every minute, together with notes of prominent natural features: swamps, escarpments, streams, etc. Traverses recorded on the outward journey can be helpful in navigating towards a particular goal, but they are likely to be less accurate than those taken on the way back, owing to the repeated interruptions of pace that occur when obstructions are met and are cleared away by machete work.

After the traverse has been plotted to a scale close to that of the aerial photographs, it will have to be bent a little and stretched here and there so that the streams crossed and escarpments skirted seem to correspond with features seen in the aerial photographs. Then the ruins themselves can usually be pinpointed. The practiced eye is often aided in recognizing ruins by the look of the vegetation around them: the dark patches of breadnut trees (ramón), and on the tops of high mounds the pale glistening of fig trees.

The technique has been found quite successful in Peten, failing only in karst terrain where there are scarcely any features other than hills, and these quite indistinguishable in their abundance, either on the ground or in stereoscopic aerial view.