Hence to maintain a minimum speed of 40 miles per hour with locomotive power on a line with long gradients of 1 in 40 involved on those parts of the line a wasted power of nearly 4 times that usefully employed; and if a minimum limit of 60 miles per hour were contemplated, a locomotive of the most powerful class in existence three years subsequent to Mr. Brunel’s report advising the adoption of the Atmospheric System would only have been able to take a single carriage up an incline of 1 in 52. So heavily at high speeds on steep gradients is the performance of a locomotive taxed by the resistance due to its own dead weight.[75] * * * * * A comparison has now to be made between the cost of power as developed by a locomotive and as developed by a stationary engine. From the well-known experiments made for the information of the Gauge Commissioners in December 1845, taking the high speed trials as the basis of calculation, it appears that 4·5 lbs. of coke per horse-power per hour may be taken as the average consumption of the engine.[76] It will be well, however, to allow for the improvement which was at the time anticipated in locomotive working, and to assume an expenditure of 4 lbs. of coke per indicated horse-power per hour, as representing the case then for the locomotive engine. Coke may be taken to have at that time cost 21_s._ a ton, or ·0094_s._ per lb. Moreover, a careful analysis of the Great Western Railway half-yearly reports, for 1844 and 1845, shows that for every shilling expended in coke, 1·44 shillings were expended on the average in wages, oil and waste, repairs, etc. Putting the results together, it appears that for each single indicated horse-power delivered by a high-speed locomotive, the cost per hour was 0·0915_s._ or 1·098_d._; that is to say, about 1-1/10_d._ per hour. Let this now be compared with the cost per horse-power per hour at which the best Cornish pumping engines had long been known to perform the work. This comparison is manifestly a rational one--with reference to the kindred employment of engine power in atmospheric pumping-engines. The performances of nearly all the pumping-engines in Cornwall were for many years so systematically and exactly reported, and the reports of each were so critically scrutinised by the rival makers, that the data they supply may be relied on without hesitation. It was well known that the best of the engines continuously performed useful work with a consumption of coal at the rate of 2·33 lbs. per delivered horse-power per hour, or, counting coal at 16_s._ per ton (a fair price on the South Devon), at the cost of ·2_d._, or one-fifth of a penny per horse-power per hour.