Accumulation in Dpt. II

[Victus Mortium]

After completing Marx's second volume of Capital, I have a few problems with his theory that I need clarified.
In Dpt. I, he explains how accumulation of constant capital occurs. He describes how the surplus value is slowly hoarded in I in order to buy future constant capital. He explains the origin of this money by separating I into two parts, A and B. A is hoarding (selling without buying) and B is in the process of accumulating (buying without selling with the money previously hoarded). In this way he explains the origin of the hoarded money in I.
My question is, then, why, in II is the accumulation and hoarding not possible in the same way. When he speaks of accumulation of Dpt. II, he explains that the money used to buy accumulated constant and variable capital must come from the consumption of the capitalists (because the other surplus is already being capitalized). Why is this necessary? Couldn't II accumulate in the same way as I? Couldn't there be subdepartments?

For example, exchange between I(v) and II(c+s) could still take place, couldn't it? I mean, assuming that II was divided into A and B, with A hoarding money, and B spending it. Suppose that I(v) = 1000 and IIA(c+s) = 1000 and IIB(c+s) = 1000. B therefore buys 1000 of the MoP produced by I, providing I with 1000 in money. I then buys 1000 worth of commodities from IIA. A has then hoarded, while B has spent its money. So, why is it necessary for money to come from the surplus of II in order to perform accumulation?

[gilhyle]

I understand the question you are asking as being can people not save (i.e. hoard) for consumption - patently they do so why does Marx's schema not allow for that ?

I think the answer to that is that Marx constructed the schema to exclude this source of consumption.

Then the question is whether it is legitimate for him to have done so. If the schemas were meant to be a description of what actually happens then this would be a problem. But that is not what the schemas are. They are a description of the logic of the relationships between c, s and v.