an accumulation of proofs and arguments where one is sufficient. The superstitious man can not, of course, be convinced, for his belief does not rest upon evidence; and he is indignant that an attempt should be made to detract from the glory obtained by the Virgin Mary and the Church in this victory over the infidels. Had I attempted to prove that the feather which is now preserved with so much care in the Church of San Juan de Lateran at Rome did not fall from the wing of the angel Gabriel when he came to announce to Mary her conception, and that the whole history of that feather was a fable, notwithstanding it has received the attestations of so many of the Holy Fathers, I should be cursed for my impiety no more than I shall be for raising the question of the authenticity of the histories of the Conquest. With all these difficulties before me, I will venture to add one or two more reasons that have induced me to doubt the existence of those famous brigantines, which required a depth of twelve feet of water.
In support of the hypothesis that the street ditches, called canals, were independent of the Tezcuco for their supply, we have still the remains of an old Indian dike, which extended from near Iztapalapan, along the east part of the city, to Guadalupe or Tepeyaca, which must have been intended to shut off the Tezcuco when the water was high, and when it receded they probably opened a weir at the northern extremity, through which the waters of the city that had been discharged upon the flats of San Lazaro found an outlet.
The waters of the valley are now distributed in the best possible manner to favor evaporation; and yet so completely is this power taxed, that when, in 1629, a water-spout, bursting over the small river Guautitlan, had forced the waters of Zumpango over its barriers into the San Cristobal, and that again into the Tezcuco, the city was inundated to the depth of about three feet. Evaporation was unable to remove or materially lessen this new volume of water in a period of five years. This fully demonstrates that the average