[195]The brigading was as follows:—1st Brigade (Hill), 5th, 9th, 38th; 2nd Brigade (Ferguson), 36th, 40th, 71st; 3rd Brigade (Nightingale), 29th, 82nd; 4th Brigade (Bowes), 6th, 32nd; 5th Brigade (C. Crawfurd), 50th, 91st; 6th Brigade (Fane), 45th, 5/60th, 2/95th. Before Vimiero the 45th and 50th changed places (see the narrative of Col. Leach of Fane’s Brigade). It is worth noting that six of these sixteen battalions, as also the 20th Light Dragoons, had just returned from the disheartening work of the Buenos Ayres expedition. They were the 5th, 36th, 38th, 40th, 45th, and 71st.

[196] Journal of a Soldier of the 71st Regiment (Edin. 1828), p. 47.

[197] Wellesley to Burrard, August 8 (Well. Disp., iv. 53).

[198] Napier, i. 197.

[199] According to the figures given by the Portuguese historian of the war, Da Luz Soriano, they stood as follows:—

Cavalry of the 6th, 11th, and 12th Regiments258sabres.
6th battalion of Cazadores562bayonets.
12th, 21st, and 24th line battalions1,514bayonets.

A few troopers of the Lisbon Police Guard, forty-one in all, according to Soriano, deserted Junot and joined the army before Vimiero. Landsheit of the 20th Light Dragoons mentions their arrival, and says that they were put in company with his regiment. This would give 2,375 as the total of the Portuguese whom Trant commanded.

[200]Well. Disp. (iv. 78) says 1,400, but in his narrative of Roliça Sir Arthur accounts for 1,600, 1,200 in his right and 400 in his centre column. As a middle figure between Wellesley and Soriano, 2,000 would probably be safe.

[201] Their allies did not think much of their looks. Col. Leslie describes them thus: ‘The poor fellows had little or no uniform, but were merely in white jackets, and large broad-brimmed hats turned up at one side, some having feathers and others none, so that they cut rather a grotesque appearance’ (p. 40).

[202] Delaborde’s numbers at the combat of Roliça have been the cause of much controversy. Wellesley in one of his dispatches estimated them at as much as 6,000 men; the unveracious Thiébault would reduce them as low as 1,900. But it is possible to arrive at something like the real figures.