gained as much as he lost in this way. Moreover, had he manoeuvred the Mexicans out of Churubusco, it would have been necessary to fight them elsewhere, when they would probably have been more ready to fight; the moral effect of this victory on both armies would not have been gained; and our military annals would not have contained this page. The moral effect on the Mexicans, however, was largely offset by pride in the stubborn resistance they had offered, and by the armistice that Scott immediately offered. One could not always determine just where firing, heard from a distance, was taking place. Probably for this reason we have inconsistent reports that make it impossible to determine precisely where and when the battle began. The Sixth Infantry, moving toward the bridge a considerable distance in advance of Worth himself, were said to have received the first fire from the convent (Hoffman: Ho. 60; 30, 1, p. 1076), but Scott reported that the attack upon the convent began some time before that upon the bridgehead. Stevens (I. I. Stevens, 198) supports him. The writer in "Apuntes" says that Worth was checked by ammunition wagons in the road, and that Santa Anna, seeing this, recalled Pérez to defend the bridge; but the wagons appear to have caused no such delay as this writer assumed. The rest of Santa Anna's force (which this writer says kept on towards Mexico) was mainly cavalry, and presumably this cavalry assisted in flanking Shields. Brev. Lieut. Col. C. F. Smith's battalion consisted of two companies from the Second Artillery, one from the Fifth Infantry and one from the Eighth Infantry (Sen. 1; 30, 1, p. 316).
The Fifteenth, but only one battalion (commanded by Capt. Wood) of the Twelfth Infantry was with Pierce, and a battery of mountain howitzers. Pierce, who had been thrown from his horse the day before, fainted and fell out. before coming into action, so Shields commanded both brigades. Lee was the engineer officer with Shields. Seeing the need of more troops, he went back to Scott and obtained the Mounted Rifles and a troop of the Second Dragoons, but these men did not reach the spot in time to fight. Scott has been criticized for not sending a stronger force in this direction; but in fact he did not even retain an escort, and the Rifles were Twiggs's reserve (Sen. 1; 30, 1, p. 309). The South Carolina regiment, commonly known as the "Palmettoes," was made up. of superior material. Men fit to be officers were in the ranks. Shields's movement was partly based on a misapprehension, for the Mexicans could retreat from Churubusco via Mexicaltzingo; but anyhow it was wise to aid the frontal attack on the bridge by applying pressure on the flank. Shields seems to have marched too far north to coöperate effectively with the attack upon the bridgehead. Presumably he did so in order to reach Santa Anna's rear. The combined effect of this movement and the outflanking of the Mexican left was to extend the American line enormously, and expose it to a (happily very improbable) counter-attack. At about three o'clock the Americans were in three sections, badly separated by distance or by the enemy, while the Mexicans, besides fighting behind strong defences, were all actually or virtually in touch one with another, and able to give mutual support.
Shields naturally overestimated the numbers opposed to him. Perhaps the Victoria and Hidalgo battalions from San Antonio passed along the highway to Mexico at this time. They would not fight. They thought hunger, sunburn and blistered feet bad enough. The Americans believed that they fought at least 32,000 men on Aug. 20 (Sen. 1; 30, 1, p. 313); but this was a great exaggeration. Rangel's brigade was in
