TH0MA6.1 LANDA'S LETTER CHARACTERS. 143 



elements; hence it is, that while we find it impossible to decipher the man- 

 uscripts by using them, yet we find such frequent resemblances as to com- 

 pel us to admit a fundamental relationship. This theory I think is borne 

 out by his attempt at explaining their method of spelling, which does not 

 correspond with anything to be found either in the Troano Manuscript or 

 the Dresden Codex, nor with his previous statement, which I have quoted. 

 Moreover, his own language, taken in connection with his alphabet, implies 

 that the natives with whom he was consulting found it impossible to recom- 

 bine the elements he had picked out so as to form words. 



This I believe to be the true explanation of his letter characters and 

 the only one that will enable us to understand why it is impossible to read 

 the manuscripts by means of them, and yet finding them so often agreeing 

 with the characters we meet with in these works. 



The day-charactei's we know he found in their books, as we see abun- 

 dant evidence of this in those yet in existence. 



Although the month-characters appear to be wanting in the Troano 

 Manuscript they are to be found repeatedly in the Dresden Codex, not 

 always in the form given by Landa, j'et substantially the same, and accom- 

 panied by numerals or other particulars by which we can readily determine 

 them. We have, then, as our only positive guide to start with, in our 

 attempts at deciphering the written characters of the Maya manuscripts, the 

 day and month symbols, and with no assurance that these are phonetic. 



If there are any phonetic elements in this writing they must be discov- 

 ered with but little reliance on Landa's letter characters. 



As it is extremely doubtful whether the day and month characters in 

 all cases correspond with the modern names applied to them, we must bring 

 every other possible test to bear in determining the meaning and corre- 

 sponding word. We have also to proceed upon the assumption that the 

 language of the Maimscript is the same as that found in the Maya lexicons 

 which have been given to the world, when it is possible that it is in a dialect 

 of the Maya varying from that in the lexicons. 



As the safest basis on which to found my arguments, I select a few 

 characters, the meaning of which, I think, can be ascertained with satis- 

 factory certainty without having to decide whether they are phonetic oi- not. 



