but with the _word_. The mode in which this is done corresponds
precisely to that of the rebus. It is a simple method, readily
suggesting itself. In the middle ages it was much in vogue in Europe for
the same purpose for which it was chiefly employed in Mexico at the same
time--the writing of proper names. For example, the English family
Bolton was known in heraldry by a _tun_ transfixed by a _bolt_.
Precisely so the Mexican emperor Ixcoatl is mentioned in the Aztec
manuscripts under the figure of a serpent _coatl_, pierced by obsidian
knives _ixtli_, and Moquauhzoma by a mouse-trap _montli_, an eagle
_quauhtli_, a lancet _zo_, and a hand _maitl_. As a syllable could be
expressed by any object whose name commenced with it, as few words can
be given the form of a rebus without some change, as the figures
sometimes represent their full phonetic value, sometimes only that of
their initial sound, and as universally the attention of the artist was
directed less to the sound than to the idea, the didactic painting of
the Mexicans, whatever it might have been to them, is a sealed book to
us, and must remain so in great part. Moreover, it is entirely
undetermined whether it should be read from the first to the last page,
or _vice versa_, whether from right to left or from left to right, from
bottom to top or from top to bottom, around the edges of the page toward
the centre, or each line in the opposite direction from the preceding
one. There are good authorities for all these methods,[12-1] and they
may all be correct, for there is no evidence that any fixed rule had
been laid down in this respect.

Immense masses of such documents were stored in the imperial archives of
ancient Mexico. Torquemada asserts that five cities alone yielded to the
Spanish governor on one requisition no less than sixteen thousand
volumes or scrolls! Every leaf was destroyed. Indeed, so thorough and
wholesale was the destruction of these memorials now so precious in our
eyes that hardly enough remain to whet the wits of antiquaries. In the
libraries of Paris, Dresden, Pesth, and the Vatican are, however, a
sufficient number to make us despair of deciphering them had we for
comparison all which the Spaniards destroyed.

Beyond all others the Mayas, resident on the peninsula of Yucatan, would
seem to have approached nearest a true phonetic system. They had a
regular and well understood alphabet of twenty seven elementary sounds,
the letters of which