If we take the wider range of labour in which women compete more directly with men in the labour market, it will be found that this practice of purchasing women introduces an unfair element in remuneration of labour. The introduction of the slave principle (the purchase of the human body) in cheapening women’s labour, has a formidable effect in depressing the wages of working-men. In all systems of industry carried on by slaves the cost of maintenance is, as a rule, the limit of expenditure, the equivalent of wages. Also in the industrial systems of so-called free industry, the maintenance of the labourer again forms a limit beyond which profit cannot be extracted, for no man will consent to labour for less wages than will keep him alive. But this is not the case in regard to women’s labour. As was proved a generation ago in France, and can be amply verified in other civilized countries, women’s wages are forced down below subsistence point. This important fact, with its cause, has evidently not been fully realized even by so close and impartial an observer as Mill. He says: ‘The wages at least of single women must be equal to their support, but need not be more than equal to it; the minimum in their case is the pittance absolutely requisite for the sustenance of one human being. Now, the lowest point to which the most superabundant competition can permanently depress the wages of a man is always somewhat more than this. The _ne plus ultra_ of low wages can hardly occur in any occupation which the person employed has to live by, except the occupation of a woman.’ Mill is evidently uncertain as to the causes of the under-payment of women in cases of equal efficiency with men, and is inclined to attribute it to injustice and to overcrowding in a few employments. He remarks: ‘When the efficiency is equal but the pay unequal, the only explanation that can be given is custom, which, making almost every woman an appendage of some man, enables men to take the lion’s share of whatever belongs to both.’