This theory, which has been advancing by rapid strides in recent years, and is now supported by a great quantity of carefully-collected evidence, requires more serious consideration. The evidence of Sir T. Brassey was chiefly, though by no means wholly, derived from branches of industry where muscular strength was an important element, as in road-making, railway-making, and mining; or from the building trades where machinery does not play a chief part in directing the pace and character of productive effort. It would not be unreasonable to expect that the quantitative relation between work and wages might be closer in industries where freely expended muscular labour played a more prominent part than in industries where machinery was a dominating factor, and where most of the work consisted in tending machinery. It might well be the case that it would pay to provide a high standard of physical consumption to navvies, but that it would not pay to the same extent to give high wages to factory operatives, or even to other classes of workers less subject to the strain of heavy muscular work. In so far as the tendency of modern production is to relieve man more and more of this rough muscular work, it might happen that the true economy favoured high wages only in those kinds of work which were tending to occupy a subordinate place in the industry of the future. The earlier facts, which associated high wages with high productivity, low wages with low productivity, in textile factories and ironworks, were of a fragmentary character, and, considered as evidence of a causal connection between high wages and high productivity, were vitiated by the wide differences in the development of machinery and industrial method in the cases compared. In recent years the labours of many trained economists, some of them with close practical knowledge of the industrial arts, have collected and tabulated a vast amount of evidence upon the subject. A large number of American economists, among them General F.A. Walker, Mr. Gunton, Mr. Schoenhof, Mr. Gould, Mr. E. Atkinson, have made close researches into the relation between work and wages in America and in the chief industrial countries of Europe. A too patent advocacy of tariff reform or a shorter working day has in some cases prevented the statistics collected from receiving adequate attention, but there is no reason to doubt the substantial accuracy of the research.