'The entrepreneurs', he holds forth, 'are essentially civil servants of economy. By the institution of property, they are once and for all entrusted with the nation's means of production. If they set them to work and strain all their energies in the process, they do but their duty, since capital--let me repeat--exists entirely for the sake of production.' And a further, factual argument: 'Or would you have them (the entrepreneurs) turn acute attacks of suffering into a chronic state by working persistently and from the first with fewer forces than are given by the means of production; are they to pay for a less severe form of the evil with its permanent duration? Even if we were silly enough to give them this advice, they would not be able to follow it. How could the entrepreneurs of the world recognise the limits beyond which the market would cease to be healthy? They engage in production without knowing the one of the other, they are producing in the most distant corners of the earth for a market hundreds of miles away, they produce with such vast forces that a month's production may already overstep the limit. How could production--so divided and yet so powerful--conceivably estimate in good time what will be enough? Where, for instance, are the organisations, the up-to-date statistical bureaux and the like to help them in this task? What is worse, the price alone, its rise and fall, indicates the position of the market, and this is not like a barometer which predicts the temperature of the market, but more like a thermometer which only registers it. If the price falls, the limit has been passed already, and the evil is with us.'[256] These thrusts, obviously aimed at Sismondi, exhibit quite fundamental differences between the two opponents. If Engels then says in his _Anti-Duehring_ that Sismondi first explained the crises as resulting from under-consumption, and that Rodbertus borrowed this view from him, he is not strictly accurate. All that Rodbertus and Sismondi have in common is their opposition against the classical school and the general explanation of crises as the result of the distribution of incomes. Even in this connection Rodbertus mounts his own particular hobby horse: over-production is not caused by the low level of working class incomes, nor yet, as Sismondi maintains, by the capitalists' limited capacity for consumption, but solely by the fact that with a growing productivity of labour, the workers' income, in terms of value, represents an ever smaller share of the product. Rodbertus takes pains to convince the opposition that it is not the small volume of the workers' share which causes the crises.