Subscriptions came in but slowly, and it was not till 1836 that the works were commenced. The first stone of the abutment on the Leigh woods or Somersetshire side of the river was laid on August 27 by the Marquis of Northampton, President of the British Association, which was then holding its meeting in Bristol.[30] The span of the bridge is greater than that of Mr. Brunel’s design for the second competition, but much less than the spans of the earlier designs, to which he had given the preference.[31] On this point, as well as on the question of site, he had to conform to the wishes of the trustees.[32] The span approved of by them necessitated the building of a very large abutment on the Leigh woods side, the height of which, from the surface of the rock to the level of the roadway, is 110 feet. Above the roadway, the tower to carry the chains is built to a height of 86 feet. On the Clifton side, the base of the tower is formed by one of the boldest of the range of St. Vincent’s rocks, which here rise almost perpendicularly to a height of 230 feet above high water, and consequently a very small abutment was required. The tower on this side is 3 feet higher than that on the Leigh woods side, and the roadway has a general inclination of about 1 in 233. Mr. Brunel thought that if the roadway were level, it would have the appearance of falling towards Clifton, owing to the ground there being precipitous, while on the Leigh woods side it is sloping. He intended, in the construction of the bridge, to have followed out the ideas embodied in his report of 1829, and would have preferred to have had only one chain on each side of the bridge, and that much stronger than was usually adopted; but, in deference to public opinion, he put two chains, though he doubted if they would expand equally. ‘A rigid platform would in some degree prevent the unequal distribution of load thus caused, but he endeavoured to lessen the effect of unequal expansion by arranging a stirrup at the top of each suspending rod, so as to hold equally at all times on both chains, and thus to cause each to sustain its proportion of the load.’ The road platform was to have had beneath it ‘a complete system of triangular bracing, which would render it very stiff.’ In order to lessen the action of wind on the bridge, he brought down the main chains in the centre nearly to the level of the platform, and intended to apply the system of brace chains at a small angle to check vibration. There were, moreover, to be two curved chains lying horizontally, and attached underneath the platform, so as to resist the lateral action of the wind.[33] He here introduced movable saddles to carry the chains on the top of the towers, with rollers running on perfectly flat and horizontal roller beds.[34] By this arrangement no pressure except a vertical one could come on the towers.