^^Mf 
 
 T^i^JEN 
 
 
 Vi ^f^fxt^ 
 
 BV 
 
 OV"RT'>^ THOMAS 
 
 FROM THE SMITHSONIAN EEPORT FOR 1900, PAf^ES 70^-721 
 (AVITH PLATES l-lll). 
 
 (Xo. 153Ji.) 
 
 WASHINGTON: 
 
 GOVERNMENT PRINTING OFFICE, 
 
 19 04. , 
 
Qass ElAAk 
 
 Book II^_li_ 
 
BY 
 
 CYRUS THOMAS. 
 
 FROINI THE S.MITHSOXIAN REPORT FOR 1908, PAGES 705-721 
 (WITH PLATES I-HI). 
 
 (No. 1532.) 
 
 WASHINGTON: 
 
 GOVERNMENT PRINTING OFFICE. 
 
 1904. 
 
C\^ 
 
 '^ '^ 
 
 .y.Y 
 
\l 
 
 r- 
 
 CENTRAL AMERICAN HIEROGLYPHIC WRITING. 
 
 By Cyrus Thomas. 
 
 The Mayan tribes of Yucatan, Chitipas, Guatemala, and western 
 Honduras had reached at the time of the ""discover}' " the highest stage 
 of native culture found in North America, except possibly in political 
 organization, in which the ancient Mexicans, or Aztecs, excelled. 
 This advance is shown b}^ their architecture, as seen in the ruins of 
 stateh" stone structures found throughout the region indicated, by 
 their sculptures in stone and wood, b}' their complicated calendar 
 S3^stem, by their arithmetical computations, and, above all, by the 
 near approach they seem to have made to alphabetic writing, their 
 system falling apparently Imt a step behind that of the ancient 
 Eg3^ptians. They engraved their peculiar hieroglyphic characters on 
 stone tablets, on great sculptured monoliths, and on the walls and 
 lintels of their T)uildings, painted them on plastered surfaces and 
 on pottery, and wrote them in books. As most of these glyphs have 
 . rounded outlines, early authors imagined they resembled somewhat a 
 section of a pebble, and the term "calculiform characters" — from the 
 Latin calculus, "a pebble" — was for a time applied to them; but this 
 is no longer in use, the term "hieroglyph," or simply "glyph," 
 having -replaced it. Where inscribed on stone or wood (for they are 
 carved on lioth, but chiefly on the former) the}'^ are made to stand out 
 in low relief, as may be seen in plate i; but occasionally they were 
 scratched or incised on shells and pottery, in which cases the glyphs are 
 generally quite rude. 
 
 Inscriptions composed of these peculiar hieroglyphs have been 
 found in the ruins of temples and of other structures in the States 
 of Chiapas and Yucatan, Mexico, and in Guatemala and western 
 Honduras. They are found in difl^'erent situations, some of them on 
 stone slabs set in the inner side of the walls of the temples, one of 
 which, from Palenque, Chiapas, is among the collections of the Smith- 
 sonion Institution. A very extensive inscription is on the inside wall 
 of the structure at Palenque, named by Stephens the "Temple of 
 Inscriptions." At Copan, in western Honduras, and at Quirigua, in 
 eastern Guatemala, the more important ones are on the sides and 
 backs of the great stone statues which stood, and, in part, are yet 
 standing, in what the native priests considered sacred precincts. The 
 
 705 
 
706 • CENTEAL AMERICAN HIEEOGLTPHIC WRITUSTG. 
 
 lintels of the temple doors and, in a few instances, even the steps lead- 
 ing up to these edifices were utilized for this purpose. Casts and 
 excellent photographs of most of these inscriptions have been made, 
 thus bringing them in reach of students for investigation and study. 
 Most of the ruins are found covered with a heavy forest growth, which 
 has to be removed before exploration can be carried on. The present 
 condition of one of the ruins at Chichen Itza, in Yucatan, named by 
 Prof. W. H. Holmes the "Temple of Tables," is shown in plate ii, 
 where the growth has been partially removed. 
 
 The glyphs of the inscriptions, which were carved so as to stand out 
 in low relief, are, as seen in plate i, somewhat square in outline, varj^- 
 ing from 3i to -ii or 5 inches square. Each of these squares, which 
 are as a rule in straight columns or lines, constitutes a hieroglj^ph or 
 glj^ph, but they are usually composed of several elements or parts. 
 This characteristic, which can not be easity explained in words, will be 
 readil}^ understood by reference to plate i. Some of these elements, 
 as will be observed, consist of lines and dots, mostly at the left side 
 or on the top of the glyphs. These are of special importance and will 
 receive further notice. Some of the glyphs consist chiefly of an oval 
 figure surrounded by a rim, as in the Egyptian cartouch. These 
 inclosed characters, with probably a single exception, are symbols of 
 Maj^a days. It is bj'^ means of these day symbols and the month 
 sj^mbols, which are also given in the inscriptions, that students ascer- 
 tain that Maya people were the authors. Diego de Landa, a Spanish 
 bishop, who went to Yucatan as a missionary in 1540, when persons 
 were still living who could read the symbolic writing of the codices, 
 has preserved in his work (De las Cosas de Yucatan) the forms of 
 these symbols, each with its proper name attached, and this is the 
 initial point of later investigations. As these names are those of the 
 Mayan days and months, and the ruins are in the regions inhabited, 
 so far as known, only b}^ Mayan tribes, the remains as well as the 
 inscriptions are attributed to these tribes. 
 
 However, Maya scribes were not limited in their symbolic or hiero- 
 glyphic writing to stone or wood, but wrote or painted their characters 
 in manuscripts. Four examples of these manuscripts, or codices, as 
 they are usually termed, remain. These are the Codex Troanus and 
 Codex Cortesianus, thought by some authors to be parts of the same 
 book, which are at Madrid; the Codex Peresianus, which is in Paris, 
 and the Codex Dresdensis, the most important of the series, which is 
 in the Ro3^al Library at Dresden. 
 
 The first two strongly resemble each other, and were probably 
 written in Yucatan, as they follow the calendar system of that region. 
 The Codex Peresianus differs in some respects from all the others. 
 The Dresden codex, which is of chief importance in studying the 
 
Smithsonian Report, 1903. — Thomas. 
 R .^ T 
 
 Plate 
 
 T' V w 
 
 p,Q_ 1 -palenque Tablet (in Smithsonian Institution). 
 
CENTKAL AMEBIC AN HIEROGLYPHIC WRITING. 707 
 
 written glyphs, agrees closely with the temple inscription in essential 
 points, and was probably written in Chiapas or Guatemala. 
 
 These manuscripts are on a kind of paper made of the Maguey plant. 
 A description of one is substantially a description of all, though the 
 size and the number of pages var3^ The Troano codex, which will be 
 taken as an example, consists of a strip of mague}^ paper about 14 feet 
 long and 9 inches wide, both surfaces of which were tirst covered with a 
 white paint or varnish. The two-faces were then divided into spaces 
 about six inches wide by black or red lines across the strip, in which 
 spaces the characters and figures, in black, brown, red, and sometimes 
 blue, were painted. The strip was then folded back and forth, like a 
 pocket map, into 35 folds corresponding with the cross lines, repre- 
 senting, when pressed together, the appearance of an ordinary octavo 
 volume. The gl3q3hs and figures cover both sides of the paper, form- 
 ing TO pages, the writing and painting having been done apparently 
 after the folding, as the folds do not. interfere with it. A page is 
 shown in facsimile in plate iii. 
 
 The order in which this Avriting — if it may properly be so termed— 
 is to be read was for man}^ years a subject of discussion, some authors 
 contending for one direction, as from left to right, or from the top 
 downward, while some thought that the reading should be in the oppo- 
 site direction. The proper order in which the inscriptions and the 
 text, in part, of the manuscripts is to be read was first pointed out 
 by the writer in 1882. « 
 
 In the inscriptions, which usuall}' consists of two, four, or six col- 
 umns, the columns are to be taken by twos or pairs from left to right, 
 and the glyphs in each pair of columns are to be read from left to right 
 and from top to bottom, in the order of the letters in the diagram 
 (fig. 1). Where there is a single column the 
 reading is from the top downward, and in sin- 
 gle horizontal lines it is from left to right. 
 The order in which the glyphs in the codices are 
 to be taken, where there is a regular arrange- 
 ment, is substantiall}" the same. Although the 
 columns may consist of but two lines in depth 
 they are read in the order «, 5, c, d in the dia- 
 gram, at least in the Dresden, Troano, and Cortesian codices. In 
 the Dresden codex, however, the numeral and time series, some of 
 which are quite long, are in some cases to be read from right to left 
 b}'' lines across the page, the lines following one another from the 
 bottom upward. Usuall}- there are in the inscriptions, besides the 
 glyphs, figures of priests and deities, and symbolic representations. 
 A considerable portion of almost every page in the codices consists of 
 
 « Study of the Manuscript Troano, 
 
 a 
 
 h 
 
 c 
 
 d 
 
 e 
 
 f 
 
 (J 
 
 h 
 
708 
 
 CEISTTE^L AMERICAN SIEROGLYPHTC WRITING. 
 
 Fig. 2.— Symbols for number 20. 
 
 pictographic representations such as are seen in the spaces below the 
 text or lines of glyphs in plate in. 
 
 An important class of characters consists of those which as is now 
 known denote numbers. These are of two quite distinct tj^pes; one, 
 which is the usual form, found in both the inscriptions and the codices,, 
 
 but more abundantly in the 
 latter, consists chiefly of 
 dots and short lines. Thus . 
 (one dot) signifies 1 ; . . (two 
 dots) signify 2, and so on 
 up to 4; 5 is indicated by a 
 single short straight line, 
 thus — ; 10 by two similar 
 lines, and 1 5 by three similar 
 lines. To represent 6 the 
 Maya scribes used a straight line and one dot _^; for 7 a straight line 
 and two dots, and so on to 9. Eleven was denoted by two straight 
 lines and a dot; 12 by two straight lines and two dots, and so on 
 to 19, which was represented thus ==• The lines and rows of dots 
 are usually horizontal in the codices, the dots above as shown here, 
 but in the inscriptions, where the}^ are always attached to glj^phs, are 
 mostly perpendicular 
 and placed at the left 
 side, as at T 17 and U 
 17, plate I (the columns 
 in the figure being de- 
 noted by letters at the 
 top and the horizontal 
 lines by figures at the 
 side as in a reference 
 map). 
 
 The numeral sym- 
 bols of this type do 
 not appear to have 
 been used for a greater 
 number than 19, other 
 characters and relative 
 
 position also, as will be <^> ^^<II>^^<^^ <27^ ^^ <^> 
 shown, being used for 
 higher numbers. Line 
 and dot numerals of two colors are quite common in the codices, the one 
 class black, the other red; but the red characters are not used (except in 
 a single unexplained instance) to denote a number greater than 13, and 
 refer almost exclusively to the numbers given to the daj^s, as explained 
 on a subsequent page. This is one instance, at least, in which color 
 
 Fig. 3.— Symbols for 0, or full count. 
 
Smithsonian Report, 1903. — Thomas 
 
 i 
 
 Plate III. 
 
 Copy of Plate XXIX, Codex Troano 
 
 (Brasseur de Bourbourg's Edition) 
 
CEISTTEAL AMEKICAN HIEEOGLYPHIO WRITING. 709 
 
 has special significance in these native manuscripts and suggests the 
 probability that the different colors of the dots used to denote num- 
 bers in the Aztec codices in the time counts have a specific meaning, 
 though this has not as yet been determined. 
 
 The number 20 is represented by several different forms, as shown 
 in fig. 2. Those marked a, 5, c, d^ and e are found only in the codices; 
 those marked/", </, h, and i occur chiefly in the inscriptions and are 
 attached to the left side of the glyphs. Naught (0) is also represented 
 in the inscriptions by characters numbered 1 to 10 in fig. 3, those num- 
 bered 1 to 8 being placed at the left side or on top of the glyph when 
 used. Numbers 9 and 10 of the figure are used chiefly in double-face 
 characters, as those seen in flg. 6. Number 11 of fig. 3 shows the 
 characters for naught (0) used in the Dresden codex. The use of these 
 symbols for naught is interesting, as it manifests a very strict adher- 
 ence to mathematical steps in the representation of numbers, no blanks 
 being allowed. 
 
 The Maya scribes were capable of carrying their numeration • to a 
 high number, and this they did in the codices, not with new or dift'erent 
 symbols from those mentioned, but by relative position, on the same 
 principle that we denote higher numbers than the Arabic digits by the 
 position of these digits. Thus we increase the value of a number ten- 
 fold in our decimal system at each step to the left, while in the vigesi- 
 mal system, used b}^ the Maya scribes, the numbers increased twenty- 
 fold at each step, to indicate which thej'^ placed their digits, if we may 
 so call them, in a column increasing from the Ijottom upward, so that 
 a line and dot, mentioned above as denoting 6 if placed at the bottom, 
 as seen in the margin of the page, would denote 6, 
 but if placed one step upward would denote 120, or 
 6 by 20, and one step higher would, according to 
 their regular vigesimal system, be equal to 2,400, or 
 6 by 20 by 20, but in their time counts, which are 
 
 equal 2, 160 
 equal 120 
 equal 6 
 
 the only numeral series in the third place, or third order of units, 
 
 would be 6 by 20 by 18, making 2,160. The other steps upward 
 
 increase unif orml}^ twentyf old. As they rise as 
 
 high as the sixth step the value of the unit in the 
 
 several steps or orders of units would be as shown 
 
 in the column at the margin. As the day was the 
 
 primarj^ unit, a single dot in the sixth step or 
 
 order would denote 2,880,000 days. A single 
 
 dot in the flf th order would denote 144,000 days, 
 
 and two dots in that place would denote twice that amount; three dots, 
 
 three times that amount, and so on up to 19. This applies to each of 
 
 these orders, except that in the second, where 18 is the multiplier. 
 
 The highest number that can be inserted is 17. They are the same in 
 
 principle as our compound denominate numbers — as pounds, shillings, 
 
 6th orider 
 
 2,880,000 
 
 5th order 
 
 144,000 
 
 4th order 
 
 7,200 
 
 3cl order 
 
 360 
 
 2d order 
 
 20 
 
 1st order 
 
 1 
 
710 CENTEAL AMERICAlSr HIEEOGLYPHIC WEITTNG. 
 
 and pence — the highest number given in the pence place is 11, as 12 
 would be 1 shilling; and 19 the highest number to be given in the 
 shilling place, as 20 would be £1. These series, or units of the various 
 orders, can be reduced to the lowest denomination — which is days — in 
 flie same way that pounds, shillings, and pence are reduced to pence. 
 Some of the numeral series in the Dresden codex amount when 
 reduced to over 12,000,000 days. 
 
 As an example of their use of large numbers, one numeral series 
 from plate lxix of the Dresden codex is presented here, the num- 
 bers indicated by the numeral characters being placed at the left in 
 parentheses and the equivalents in days at the right. The names 
 placed at the extreme left (great cycle, cycle, etc.) are those adopted 
 'bj Mr, Goodman for the respective orders: 
 
 Days, 
 (great cycles) (4) . . . . equal 11,520,000 
 
 |:,, (cycles) (5) equal 720,000 
 
 % (katuns) (19) .... equal 136,800 
 
 (ahaus) (13) . . . equal 4,680 
 
 (chuens) (12) . . equal 240 
 
 (days) (8) ... equal 8 
 
 Total 12,381,728 
 
 That is to say, 4 great cycles (or 4 units of the sixth order or posi- 
 tion) equal 11,520,000 days; 5 cycles (or 5 units of the fifth order) 
 equal 720,000 days; 19 katuns (or 19 units of the fourth order) equal 
 136,000 days; 13 ahaus (or units of the third order) equal 4,680 days, 
 and so on. 
 
 The total amount expressed by this series is over 12,000,000 days. 
 This is a large number to be handled by a pre-Columbian native, yet 
 it can be demonstrated by actual count that the Maya scribe used this 
 number correctly in a calculation. 
 
 Writers of the present day have adopted the simple method of 
 expressing these numeral series thus (using the above example), posi- 
 tion indicating the orders of units 4-5-19-13-12-8, ascending toward 
 the left just as we may express .£4, 12 shillings, and 6 pence, thus — 
 4-12-6. 
 
 A knowledge of the Maya numeral system and method of counting 
 and expressing numbers, as given above, is absolutely necessary in the 
 attempt to decipher the glyphs. It is also necessar}^ to give here a 
 brief notice of the Maya calendar, as a knowledge thereof is another 
 requisite in deciphering. The process with the Maya glyphs, so far 
 as it has been carried, is wholly difi'erent from the method- pursued in 
 deciphering Egyptian hieroglyphs and the cuneiform inscriptions of 
 Assyria. There the phonetic value of the characters bei ng ascertained, 
 the combinations to form words can be followed and tested by the 
 
CENTRAL AMERICAN HIEROGLYPHIC WRITING. 
 
 711 
 
 result. In the Maya, with the few exceptions which will be mentioned 
 later, the gl3^phs, so far as determined, are to a large extent S3"mbols 
 (not phonetic characters), used to denote numbers, daj'^s, months, etc. 
 Hence the only means so far discovered b}^ which to test an interpre- 
 tation is the demonstrable relation of one character to another, thus: 
 Having- a S3mibol known to be that of Monday, another to be that for 
 7, another to be that for the month of March, and another for the 
 number 120, and finding them placed in an inscription in the order: 
 Monday, March 7, 120, and this followed b}^ two imperfect or unknown 
 
 characters and 5, thus ? ? 5, and having ascertained that the 
 
 intermediate numbers, as the 120 in this case, indicate the number of 
 
 Pop. 
 
 Yaxkin. 
 
 Chen. 
 
 Zac. 
 
 Ceh. 
 
 Mac. 
 
 Kankin. 
 
 Pax. 
 
 Kayab. Cumhu. 
 
 Fig. 4. — The symbols of the months. 
 
 Uayeb. 
 
 days from the first date to a second, we count 120 days from Monday, 
 March 7, which brings us to Tuesday, July 5. This gives us Tuesday 
 and July as the two unknown or doubtful characters of the terminal 
 date. Just as it is necessary, in the example given, to understand, in 
 part at least, our Gregorian calendar, so is it necessary to understand 
 the Maya calendar in attempting to decipher the Maya hieroglyphs. 
 
 The Maya years consisted uniformly of 365 days, no reference to or 
 evidence of bissextile years (corresponding to our leap year) having 
 been found in the codices or inscriptions. They were divided uni- 
 formly into 18 months of 20 days each, and a supplemental month of 
 5 days following the 18th. Each of these months had a name and a 
 
 SM 1903 16 
 
712 
 
 CENTRAL AMERICAN HIEROaLYPHIC WRITING. 
 
 symbol as shown in fig. 4. They alwaj^s followed one another in the 
 same order, the year uniformly beginning with Pop. The 20 days 
 were also named, each having its appropriate symbol as shown in fig. 
 5. The order in which they followed one another was uniform, though 
 the year did not always begin with the same day, the 5 in the supple- 
 mental month carrying the count forward 6 days each year. Although 
 the days had their month numbers, as 1, 2, etc., to 20, as we say the 
 
 Ik. 
 
 Eb. 
 
 Akbal. 
 
 
 
 Muluc. 
 
 Ben. 
 
 Chicclian. 
 
 Men. 
 
 Cimi. 
 
 Chuen. 
 
 Gib. 
 
 Caban. 
 
 Ezanab. 
 
 Cauac. 
 
 Ahau. 
 
 Imix. 
 
 Ahau. 
 
 Ahau. Ahau. 
 
 Fig. 5. — The symbols of the days. 
 
 Ahau. 
 
 Ahau. 
 
 fifth, sixth, and seventh day of the month, there was another number- 
 ing which applied to the daj^s only. This, however, was from 1 to 13, 
 beginning again with the unit. These numbers were prefixed to the 
 da3's and followed in regular succession, no day being without its num- 
 ber. It follows from this method that a day bearing both the same 
 name and the same number will not recur until 13 months have passed. 
 This gives a cycle or period of 260 days, which appears to have been 
 
CENTRAL AMEKICJ'N HIEROGLYPHIC WRITING. 718 
 
 more in use as a ceremonial or religious period, both among the Mayas 
 and Mexicans, than the secular year of 365 days. 
 
 The order of the daj'^s and their numbering passed on from month 
 to month and from year to year without a break or change in the reg- 
 ular succession. There is one series of 312 years in length in the 
 Dresden codex, in which there is not a break in the succession, nor 
 an indication of a bisextile jeRV. In the series given above, also 
 from the Dresden codex, which covers 31,059 years, 9 months, and 13 
 daj's, the date of the commencement and of the ending being given, 
 which calculation shows to be correct, this is evidence that there can 
 be no break or change in the succession of days, day numbers, or 
 months. In this regularit}^ of succession lies the possibility of deter- 
 mining the time series of the inscriptions and the codices. 
 
 In order to show what advance has been made in deciphering this 
 ancient American writing, it is necessary to present examples from 
 the codices and inscriptions that the reader may have the gl3'phs 
 referred to before him, for words alone can not describe them so as 
 to be understood. Beginning with the inscriptions, Avhich appear to 
 be older than the codices, attention is called to plate i, showing the 
 inscription on the Palenque tablet in the Smithsonian Institution. As 
 a means of identifjnng the individual glyphs, a letter is placed over 
 each column and a number at the side opposite each line, as in refer- 
 ence maps. R, S, T, U, V, W, X have been selected because they 
 are the letters used for these particular columns in Doctor Rau's 
 scheme.'* 
 
 The column R being separated from the others, and a single col- 
 umn, it must be read from the top downward. Passing by this, 
 attention is called to the other six, which are to be read two and two, 
 beginning with the two at the left, going from the top downward, 
 taking the glyphs alternately in the left of the two columns and then 
 in the right, thus: First glyph, S 1 then T 1; next, S 2, T 2; then S 3 
 and T 3, and so on to the bottom. Then columns U and V are to be 
 taken in the same order, and after these columns W and X. As it 
 would require a somewhat extended stud}^ of the subject to follow 
 out understandingly a complete explanation of the steps in the process 
 of decipherment,. an outline only of what has been accomplished in 
 this direction can be given. 
 
 Reading down columns S and T in this manner, the first glyph 
 which has been determined, or rather could be determined if unin- 
 jured, is T 2 (or the second in the T column), which, from the sur- 
 rounding border or band and the number attached is known to be the 
 symbol of a day, but on account of the imperfect markings or weath- 
 ering of the face, is not indentifiable with certainty. Here, however, 
 is an instance where a knowledge of the Maya calendar system 
 
 a Palenque Tablet, in Sm. Cont. Knowl., vol. xxii, p. 61. 
 
714 CENTRAL AMERICAN HIEROGLYPHIC WRITING. 
 
 becomes important, as it enables us to limit the investigation to one 
 of four out of twent}^ da^^s. As the next glyph which follows — that 
 is, S 3 (or the third in the S column) — is the symbol for the month 
 Pop (see fig. 4), the first month of the Maya year, and has attached 
 to the left the symbol for 20 (similar to that shown at i fig. 2), it is 
 evident that the day at T 2 is the 20th day of the month. As there 
 are but four da3^s (Ik, Manik, Eb, Caban — see fig. 4) in the calendar 
 system used in the inscriptions that can fall on the 20th, it is evident 
 that, it must be one of these. The reader will observe by inspecting 
 this glyph in the figure that there are two short perpendicular lines 
 and a dot at the left; these denote that it is the day 11 — ? 
 
 Passing on to S 10, T 10, we find another date, the glyph S 10, 
 being the symbol for the day 11 Lamat, and glyph T 10, the s3^mbol 
 for the month Xul with the numeral character for 6 at the left. For 
 the days mentioned reference can be made to fig. 5 and for the months 
 to fig. 4. However, for the illustration the names of the days and the 
 months are not essential, but are added here to avoid using blanks. 
 It will be observed, that above and below the little dot in the numeral 
 characters at the side of each of these glyphs is a little semicircle or 
 crescent. These, which might be mistaken for number dots, have no 
 significance, but are used to fill out the space or to guard the dot. 
 
 To be able to say that certain glyphs denote days, others months, 
 and others numbers is one step in the process of decipherment, but 
 the step is a comparatively short one unless their relation to one 
 another and the object of their introduction into the inscription can be 
 ascertained. This relation has been determined in part through inter- 
 mediate number series. For example, by passing on to glj^phs S 12 
 and T 12, we find the number series 9 daj^s, 3 chuens (or units of the 
 second order), and 13 ahaus (or units of the third order) — or 13-3-9 — 
 which, reduced to the lowest denomination, gives 4,749 days. Count- 
 ing this number of days, according to the Ma3^an calendar, from 13 
 Lamat, 6 Xul, the date given in S 10, T 10, and mentioned above, we 
 reach the date 2 Caban, 10 Xul, which is the date given two lines 
 below at S 14 and T 14. That is to say, the number in the numeral 
 series is the exact time included between the immediatelj^ preceding 
 and the immediately following date. This is proof positive that there 
 is a connection between the date at S 10 and T 10 and that at S 14, 
 T 14. Nor does the connection series end here. Glyph S 15 is a 
 short-number series of 123 days which connects the date 2 Caban, 10 
 Xul, of glyphs S 14, T 14, with the date 1 Ahau, 3 Zip of glyphs T 
 17 and U 1; or, omitting names, it connects the last preceding with 
 the next following date. 
 
 Dropping from consideration the names of the days and months, 
 introduced to avoid blanks or explanatory phrases, the important fact 
 
CENTRAL AMERICAN HIEROGLYPHIC WRITING. 
 
 715 
 
 remains that there is a connection between date glyphs that stand some 
 distance apart. 
 
 It has been stated above that the Maya writing- included two types 
 of numeral characters. One of these, consisting of dots and short 
 lines and the use of position, has been explained. The other type con- 
 sisted of face characters, some of which are shown in fig. 6. In order 
 
 Symbol oi' calendar rounds. 
 Fig. 6. — Face characters representing numbers. 
 
 to show how these are used, attention is called to fig. 7, which repre- 
 sents part of the inscription on the east side of Stela (or statue) F, at 
 Quirigua, Guatemala, as designated by Mr. Maudsla}^, from whose 
 great work (Biologia Centrali-Americana, "Archaeology"), part xii, 
 pi. 40, our figure is taken. 
 
716 
 
 CENTRAL AMERICAN HIEROGLYPHIC WRITING. 
 
 As seen in this illustration (fig. Y), there is at the top or beginning a 
 large quadruple glyph, below which follow, in the order of the num- 
 bers 1, 2, 3, 4, 5, and 6 at the sides, six double glyphs, each composed 
 
 of two faces. There are other 
 
 double-face glyphs below, but the 
 six will suffice for illustration. 
 Omitting from present consider- 
 ation the large character at the 
 top, attention is directed to these 
 six glyphs, from which we may 
 learn something of the mistakes 
 made in the attempts at decipher- 
 ment. A little more than a dec- 
 ade ago there was almost univer- 
 sal agreement among students of 
 the Ma}' a hieroglyphs that these 
 face characters, especially those 
 in commencing series, as shown 
 in fig. 7, were symbols of deities. 
 Subsequent investigation, how- 
 ever, has shorn them of their 
 sacred character and reduced 
 them to mere symbols represent- 
 ing numbers. The left face of 
 each of these six double glyphs 
 is one of the smaller numbers (1 
 to 19), which we have designated. 
 ' 'Ma3^afn digits ;" for instance, the 
 left face in glyph number 1 de- 
 notes 9; that in glyph 2 stands 
 for 16; that in 3 for 10; that in 
 4 for (naught), and that in 5 
 also stands for (naught). These 
 are the numbers prefixed respec- 
 tivel}^ to the symbols of the orders 
 of units in the inscription repre- 
 sented. The right face of num- 
 ber 1 denotes the cycle or fifth 
 order of units; adding the prefix 
 9, the double gl}" ph will signify 9 
 cycles or 9 units of the fifth order. 
 The right face of glyph 2 is the 
 katun or fourth order of units; that of 3 the ahau or third order of 
 units, etc. Glyph 6 is the day (1 Ahau) and glyph 12 the month 
 (3 Zip), forming together the terminal date of the series. 
 
 Fig. 7. 
 
 -Inscription on Stela F, Quirigua. 
 from Maudslay's plate. 
 
 Copy 
 
CEKTHAL AMiERICAJSr HIEROGLYPHIC WRITING. 717 
 
 Briefly stated, this series (fig. 7) and all those of like character are 
 made up of numbers and dates, and not of deities, as was formerl}^ 
 supposed. 
 
 The diflferences in these face characters, by which their respective 
 values are. determined, have not in every instance been so clearly 
 ascertained that the}^ can be determined by inspection alone. In the 
 left face of glyph 1 the circle of dots on the cheek forms the distin- 
 guishing characteristic for 9, but peculiar markings of others are less 
 distinct. The face characters representing the orders of units, as the 
 cycle, katun, etc., can be determined bj^ position alone. 
 
 The great quadruple glyph at the top is the symbol for the sixth 
 order of units (Goodman's "great cj^cle"), which seems to have repre- 
 sented the limit of Mayan time counts, although according to Doctor 
 Brinton their numeration in the regular Maya number system Was 
 carried a step higher; and Goodman intimates that their time counts 
 reached an additional step in the scale, amounting at the extreme to 
 280,800 years or 102,492,000 days. This large so-called "great cjxle 
 S3niibol," with the number characters and the immediately following- 
 date, form what Maudslay has termed an "initial scries," as the large 
 glyph is never found except at the commencement of an inscription. 
 
 The month symbol which helps to make up the date in this instance 
 is somewhat distant from the day sjnnbol, five compound glyphs inter- 
 vening; nevertheless there is numerical evidence that the two are con- 
 nected and that the date i's part of the "initial series." There is also 
 evidence that the initial series in the inscriptions at Copan, Quirigua, 
 and some other localities where the number of cycles is 9, as in this 
 case, start with the same date (4 Ahau 8 Cumhu), this date being 
 apparently the beginning of an era with the priests and scribes of 
 those sections. As this is but one instance of a number where the 
 count in these initial series gives the proper terminal date in the 
 inscription, the proof that they have been correctly interpreted seems 
 to be complete. This conclusion necessaril}^ carries with it the accept- 
 ance of the interpretation given the glyphs and also the calendar sys- 
 tem as above explained, though the native priests appear to have pur- 
 posely used characters which would be understood only by their own 
 class. 
 
 The codices, as will be seen b}" referring to plate iii, which is a 
 facsimile of three of the four divisions of plate xxix of the Codex 
 Troano, contain a much larger proportion of pictographic representa- 
 tion than the inscriptions. Besides the pictures there are two classes 
 of hieroglyphs; first, the ordinary numerals represented by dots and 
 short lines, which are of two colors, black and red. The latter, which 
 do not exceed 13 in value, are the numbers attached to or belonging to 
 the days — the day, where the symbol is omitted, as in the lines of the 
 alternate black and red numerals in this instance, being understood. 
 
718 CENTRAL AMERICAN HIEROGLYPHIC WRITING. 
 
 The column at the left side of the lowest division consists of the s3^mbols 
 of 5 days, which form the basis from which the count by the black and 
 red numerals is made. The dsij columns for the two upper divisions 
 are in a preceding plate, the line of numerals running through more 
 than one plate. 
 
 The two lines of black glyphs running across the upper part of each 
 division forms what may be termed the "text." These are read in this 
 instance by groups of four, considering them two short columns, as 
 those over the bird and personage in blue at the left side of the middle 
 division, the order being the same as a, h, c, cl in the diagram, fig, 1. 
 But little progress has as yet been made in deciphering this so-called 
 "text." 
 
 So far as the writer is aware, but three characters of the text of this 
 plate have been determined save in the manner mentioned below. 
 These are the symbols of three of the cardinal points, and are the first, 
 third, and fifth glyphs in the upper line of the upper division, counting 
 across from the left. 
 
 That the text in most instances contains reference to the figures 
 below is quite evident. This is shown in plate iii by the fact that some 
 feature of the pictures is represented by one of the four glyphs which 
 stand above it, as in the middle section_ the bird's head forms one of 
 the glyphs over the figure in which the bird appears; and likewise 
 the dog-like animal and worm in the same division are represented in 
 the glyphs of the text above. These would therefore seem to be 
 simple abbreviated pictographs or conventionalized figures and not in 
 any sense phonetic characters. In the lower division of the same figure 
 the three persons to the right are holding in their hands something like 
 the symbol for the day Ik (fig. 5) (which signifies "wind" in theMaj^a 
 language) ; the same symbol appears in the text above the head of each, 
 but its signification in these places is unknown. We may surmise if 
 we like, but the proof is wanting. 
 
 As the gij^phs in the middle division of the figure, on which the per- 
 sons and other forms are sitting, appear to represent something out of 
 which plants can grow and has the elements of the symbol of the da}^ 
 Caban (see fig. 5), it is possible they denote earth (cab in Maya signifies 
 "earth"). In the upper division the Kan-like s^^ubols (one of which 
 a bird is pecking and another is bitten by a little quadruped) prob- 
 ably represent grains of corn, supposed to be in the ground, the two 
 to the right throwing out sprouts. If this interpretation be correct, 
 this entire plate probably has reference to the cultivation of corn and 
 the dangers it is subject to. However, from what has been stated, the 
 reader can judge as to the portion of this codex that has been deter- 
 mined with certaint3^ which is but little, and as to what is as 3^et but 
 theoretical. Of the text proper, scarcely anj^thiug, as before stated, 
 has been absolutely determined. This failure to decipher is attribu- 
 
CENTRAL AMEEICAN HIEROGLYPHIC WRITING. 719 
 
 table in part to the fact that where the suggested signification may be 
 absohitely correct, no means, except where numbers come into play, 
 has been found to verify the conclusion. 
 
 From what has been stated and the examples presented from the 
 inscriptions and codices, it is apparent^ notwithstanding the number 
 of glj^phs whose signification has been ascertained, that practically no 
 progress has been made in determining the phonetic equivalents of these 
 characters. In other words, no satisfactory evidence has 3'et been 
 presented to show that any of these glyphs are phonetic, although 
 there is sufficient evidence that the language used was Maj^a. The 
 nearest approach to proof on this point is in regard to a few symbols, 
 such as that for the month Tzotz (fig. 4). The usual form of the 
 glyph is the conventionalized head of the leaf-nosed bat, and in one 
 instance (Stela D, Copan) the full form showing the wings and bod}" 
 is introduced as the symljol of the month. As tzotz is the Maya word 
 for bat, it is possible the word relates to the S3"mbol. Pop, the name of 
 another month, signifies in Ma3"a literally a mat, or rug, the refer- 
 ence being apparently to the structure, and the chief feature of the 
 symbol for the month consists usually of interlacing like basket 
 work. These and a few other instances of similar character consti- 
 tute the strongest indications of phoneticism that have been observed, 
 but as the elements of these gl3^phs found where the character can not 
 be determined b}^ other means furnish no aid in decipering them the 
 inference of phoneticism is doubtful. It is possible that some of the 
 characters are phonetic, yet it must be admitted that no satisfactory 
 proof thereof has yet been presented, although the author, with others, 
 thought but a few 3"ears ago that continued investigation would soon 
 produce this evidence. 
 
 The general purport of the inscriptions has not been ascertained 
 with certaint3^, 3^et the fact that half of them belong to the classes 
 heretofore described — the numeral symbols, calendar symbols, etc.— 
 leads to the conclusion that the3' contain little, if anything, relating to 
 the histor3^ of the tribes by whom they were made. 
 
 The indications that the Ma3^a priests, by whom these inscriptions 
 were doubtless designed, if not carved, recognized a prime or ruling 
 era from which a large portion of the initial time series are counted, 
 are so strong that most recent authorities who have devoted attention 
 to the subject have concluded to adopt the theory, at least tentatively. 
 We might hope that further research will prove that this has some 
 relation to Maya histor3^ were it not that the beginning was placed 
 about four thousand 3^ears prior to the time when the inscriptions 
 were made — a date so remote as to preclude the supposition that it 
 related to an3" noted event in the history of the tribes. 
 
720 CENTEAL AMERICAN HIEROGLYPHIC WRITING. 
 
 The progress made in deciphering the text of the codices is less than 
 that made in interpreting the inscriptions, as the number of numeral, 
 time, and other sj^mbols in the former which have been determined 
 is less in proportion to the whole than in the latter. However, this 
 proportion is limited to the text of the codices and does not include 
 the accompanying numeral and day series. Nevertheless, the aid fur- 
 nished by the figures which are introduced, together with the relation 
 a large portion of the time series bear to the text and figures, often 
 furnish some indication of the general purport of the plates, but all 
 attempts to give the details have thus far failed, from the lack of means 
 of verification.' Two or three of the plates of the Dresden codex are 
 devoted entirely to a single numeral series. These can be followed 
 throughout and the obliterated characters in most cases restored; in 
 fact, some of them seem to be little else than the steps of the calcula- 
 tion made by the original scribe. Possibly their relation to adjacent 
 series may yet be ascertained and their signification determined. 
 This has been accomplished in regard to the series running through 
 plates 46-50 of the Dresden codex. '^ 
 
 A brief answer to the question, What has the progress thus far 
 made in deciphering this hieroglyphic writing added to our knowledge 
 of the ancient history, life, and attainments of the Maya people? may 
 properly close this brief article. 
 
 That it has shown a greater advance in culture along particular lines 
 than was previously known is certainl}^ true. Much has been ascer- 
 tained from the remains of stone structures and the sculptured designs 
 thereon in regard to the advance of the Maj^as along certain lines of 
 art and their ability to form and to carry out comprehensive plans 
 and designs; but the study of the hieroglyphs has brought to light 
 evidence of mental capability and attainments of a higher grade in 
 some respects than has been shown elsewhere. It would be some- 
 what difficult for anyone at the present da}^, except a mathematician, 
 to calculate back 34,059 years 9 months and 13 days from a particular 
 day in the present year, using our Gregorian calendar, and determin- 
 ing the exact month, day of the month, and daj of the week that will 
 be reached. Yet this was accomplished by the Maya priests according 
 to their calendar and with their cumbersome vigesimal system. Not 
 only was it necessary to reduce the several orders of units (cvcles, 
 katuns, etc.) to the lowest denomination, but the sura had to be 
 changed into years, months, and days. The modern mathematician 
 has his books of tables, and his paper, ink, and pen and pencil, and a 
 numeral system that is simple and easy to handle. How did the 
 Mayan scribe solve the same problem with the means he had at hand? 
 The study of the glyphs has brought these facts and this question 
 before us. 
 
 «The Maya Year, by Cyrus Thomas, Bureau Am. Ethn., 1894. 
 
CENTEAL AMERICAN HIEROGLYPHIC WRITING. 721 
 
 The steps which have been made in decipherment have made it 
 evident that the Mayan priests had an understood era or a well-under- 
 stood point of departure in their time counts. They also indicate 
 that the inscriptions at Copan and Quirigua were carved in substan- 
 tially the same period, the range, judging by the terminal dates of 
 the initial series, being comprised in two hundred 3^ears. But the 
 attempts to connect the dates in the Maj^an inscriptions and codices 
 witli dates in the Gregorian calendar have failed, though greater suc- 
 cess has attended the efforts in this direction with the Aztec count. 
 Another fact made prominent by the study of these glyphs is the 
 uniformity in the system, art, and culture, along the lines indicated, 
 in Chiapas, Guatemala, western Honduras, and with slight exceptions 
 in Yucatan. The collection of hieroglyphs from the inscriptions of the 
 latter section are not sufficient to determine whether they follow the 
 Troano and Cortesian codices or the system of the inscriptions of 
 Chiapas and Guatemala. 
 
 The study of the inscriptions and codices has made it evident that 
 no adjustment between the Maya year and the solar year was made in 
 any way that appears in the record or interfered with the calendar 
 count. Although the efforts at interpretation have succeeded in few 
 if any instances in tracing the connection throughout long inscrip- 
 tions, they have made it evident that there is connection, or, in other 
 words, that these inscriptions (with possible exceptions) are continu-^ 
 ous records from the initial glj^^h to the end, though it may consist of 
 little else than number series and time counts. Both inscriptions and 
 codices evidentl}" relate ver}^ largely to ceremonies and priestly duties, 
 more particularly the latter. 
 
 Another result of the stud}' of the hierogl3"phs is the clear distinc- 
 tion it has established between the Maya and the Aztec symbolic 
 writings. 
 
 The Maya writing has been studied to a greater or less extent by 
 Leon de Rosny, Hyacinth de Charencey, and Brasseur de Bourbourg, 
 of France; P. Schellhas, E. Forstemann, and Eduard Seler, of Ger- 
 many; A. P. Maudslay, of England; Charles Rau, Edward Holden, 
 D. G. Brinton, J. T. Goodman, Marshall H. Saville, Cyrus Thomas, 
 G. B. Gordon, and C. P. Bowditch, in the United States. 
 
WUV Id'd 1SU4 
 
 ^9