be blamed somewhat for his excessive caution in giving way so rapidly before St. Cyr’s advance: but if we remember the quality of his troops and the inevitable result of a battle, it is hard to censure him overmuch. Meanwhile Garcia Conde, whose movements were most happy and adroit, reinforced the garrison of Gerona up to its original strength of 5,000 bayonets, by making over to Alvarez four whole battalions and some picked companies from other corps, and prepared to leave the town with the rest of his division and the vast drove of mules, whose burden had been discharged into the magazines. If he had dedicated his whole force to strengthening the garrison, the additional troops would have eaten up in a few days all the provisions that the convoy had brought in[46]. Accordingly he started off at two a.m. on September 4 with some 1,200 men, by the upland path that leads past the hermitage of Los Angeles: St. Cyr had just placed Pino’s troops from the covering army to guard the heights to the south-east of Gerona, but Garcia Conde, warned by the peasants of their exact position, slipped between the posts and got off to Hostalrich with a loss of no more than fifty men[47]. [46] The reinforcements left behind by Garcia Conde consisted of two battalions of Baza (one of Reding’s old Granadan regiments), with 1,368 bayonets, two Catalan ‘tercios,’ 1st and 2nd of Talarn, with 716 bayonets, and select companies of 1st of Granada (Iliberia), 2nd of Vich, and Voluntarios de Tarragona--in all apparently about 2,707 men. The table on p. 375 of Arteche’s vol. vii seems to err in crediting the Cervera ‘tercio’ to Garcia: this had come in on Aug. 17, as described on p. 35. On the other hand the company of Voluntarios de Tarragona should be credited to him. [47] St. Cyr tells a story to the effect that he had placed Mazzuchelli’s brigade of Pino’s division in ambush behind the hill of Palau to intercept Garcia Conde, and that the Spaniards would have marched right into the trap on Sept. 3, if the Italians had not been stupid enough to sound the _réveil_ at dawn, and so warn the enemy of their existence. But the Spanish accounts of Minali and Claros are quite different (see Arteche, vii. 377); they are to the effect that Garcia Conde had intended to start _at dusk_ on the third, but, hearing firing on the side of Palau, deferred his exit and took another road. If he was starting at 7 or 8 o’clock at night on the third, he cannot have been warned by the morning bugles at 4 o’clock on the previous morning. See St. Cyr, p. 234, and Napier, ii. 245, for the French story, which the latter takes over whole from the former. Belmas and Vacani do not give the tale, though they have a full narration of the escape of Garcia Conde.