that the evidence at our disposal is not sufficient to permit these questions to be answered with certainty. Much patient excavation, both in the Mexican valley and in the central Mayan region, is necessary before the origin of the pre-Spanish civilization can be traced with any degree of finality; but I think it is possible to put forward a working theory, which must of necessity be modified in accordance with future researches, but which may at least be of service to critics as an invitation to concentrate their minds upon the subject. It is solely with this idea that I have ventured to draw up a scheme of dating (see Appendix III), and I will deal with this first, pointing out its weaknesses with as much impartiality as possible.
First of all I should like to say that the dates are not intended to be more than approximate; even the records of Mexican history are not in absolute accordance, and contradictions become more frequent and serious in proportion with the remoteness of particular events from the date of the conquest. However, the history of the valley can be compiled with some degree of probability when the different accounts are studied from the point of view detailed on p. 19, though the dates of Toltec history, before the fall of the city, are in the highest degree problematical. The books of Chilan Balam give a chronology for the history of the Tutul Xiu which can be related to European time, though they do not agree exactly among themselves, and interpolations of a whole cycle appear to occur in places. The dating which I have adopted, which is in the main that of the book of Chilan Balam of Mani, is a conservative interpretation of these records, and may, I believe, be credited with reasonable probability. The dating of the monuments, the most important question of all, is unfortunately the most problematical, and the chain of evidence by which I seek to assign them a place in chronology has two weak links, one considerably weaker than the other. The first
