“For the part of the constant capital which is exchanged in kind, the position is the same as it was. The same quantity of iron, timber and coal as before will be exchanged in kind in order to replace the iron, timber and coal that has been used up, and in this transaction the higher prices will balance each other.” (p 195)
In other words, if we view the situation in terms of c + v + s, the value of c has risen, but it is the quantity not the value (as measured by the historic price) which must be reproduced and “which is exchanged in kind”. In Marx's words here (and he uses the same terminology in Capital III, in analysing social reproduction) it is replaced in kind. So, although the value here of iron, wood, and machines rises, they have to be physically reproduced in kind. Their value rises, and is passed on into the value of output, but the amount of value that must, therefore, be set aside for their physical replacement rises by an identical amount.
That is clear, for that part of the product which physically replaces itself directly. For example, if the value of corn rises, because more labour-time is required for its production, this does not change the quantity of that corn that must be set aside, as seed, to replace the seed used in production. It does mean, therefore, that the current reproduction cost of that seed, used as constant capital, is thereby raised relative to its historic cost. As Marx puts it, this constant capital is retrospectively revalued, in line with its current reproduction cost. It is irrelevant how much labour-time was used to produce that seed at some point in the past. It is how much labour-time that must be set aside now that is decisive.
The same thing is true of the coal taken from the output of the coal mine to directly replace the coal used to fuel its steam engines. Suppose a coal mine uses ten tons of coal to fuel these engines, and the coal it set aside from last year's production had required one hundred hours of labour to produce. The mine this year year produces one hundred tons of coal, but requires 1200 hours of labour to produce it, meaning productivity has fallen.
On this basis, to produce the ten tons of coal required to reproduce the constant capital, the ten tons required for the steam engines would require not 100 but 120 hours. The value of the coal used as constant capital would thereby be retrospectively revalued to reflect the 120 hours of labour currently required for its reproduction, as opposed to the 100 hours of labour actually expended upon the production of the consumed coal.
But, the same thing applies to that portion which is not directly replaced in kind. If the value of wood, iron or machines rises, so that more labour-time is required to replace the constant capital of the coal producer, this will be reflected as described above, in an increase in the value of the coal itself. The coal producer will sell all of their coal at this higher price, but a portion of the proceeds of that coal, sold to the consumer goods sector (whether used for consumption or production by the buyers from that sector) will now go not to fund purchases of consumption goods, by the coal producer, but will go to fund their purchase of the required quantity of wood, iron and machines, at their now higher price, which reflects the additional labour employed in those industries, due to the lower productivity.
In other words, whatever the historic cost of the wood, iron and machines used in current production, a greater proportion of current coal production will be required to cover their reproduction.
“But the surplus of coal which now forms a part of the constant capital of the coal producer and does not enter into this exchange in kind is, as before, exchanged for revenue (in the case given above, in part not only for wages but also for profit); this revenue, however, instead of going to the former consumers, accrues to the producers in whose spheres of production a greater quantity of labour is used, that is, the number of labourers has increased.” (p 195)