We have already seen that differences in the turnover period create differences in the annual rate of surplus-value, even when the total amount of surplus-value produced each year stays exactly the same.

But there is also, necessarily, a difference in how the surplus-value gets capitalized — in accumulation — and so, even with the rate of surplus-value staying the same, a difference in the mass of surplus-value produced during the year.

Take capital A, from the example in the previous chapter. It has a steady, recurring income, so — apart from the turnover period at the very start of the business — it covers its own owner's spending during the year out of the surplus-value it is currently producing, and has no need to advance anything from a fund of its own. B is different. B produces exactly as much surplus-value in the same stretches of time as A does. But that surplus-value has not yet turned into money, so it cannot be spent — not on the owner's own consumption, and not productively either. As far as the owner's own consumption is concerned, it is being spent before it has actually been realized. A fund for it has to be advanced.

One part of productive capital that is hard to classify — the extra capital needed to repair and maintain fixed capital — now looks different too, in a new light.

For A, this part of the capital — wholly, or for the most part — is not advanced at the start of production at all. It does not need to be available in advance, or even to exist yet. It arises out of the business itself, through surplus-value being turned directly into capital, that is, applied directly as capital. Part of the surplus-value that is not only produced but also realized in money at intervals during the year can cover the outlay needed for repairs and the like. In this way, part of the capital needed to keep the business running at its original scale gets generated by the business itself, while it runs, simply by capitalizing part of the surplus-value. For capitalist B, this is impossible. For him, that part of the capital has to form part of the capital he advanced at the outset. In both cases this part of the capital will show up in the capitalist's books as advanced capital — which it is, since, on our assumption, it forms part of the productive capital needed to run the business at the given scale. But it makes a huge difference which fund it is advanced from. For B, it really is part of the capital that had to be advanced, or held ready, from the start. For A, by contrast, it is a part of surplus-value applied as capital. This second case shows us that not only accumulated capital, but even part of the originally advanced capital, can be nothing but capitalized surplus-value.

Once credit enters the picture, the relation between originally advanced capital and capitalized surplus-value gets tangled further still. Say A borrows part of the productive capital he uses to start the business, or to keep it running during the year, from banker C. A does not have enough capital of his own from the start to run the business. Banker C lends him a sum made up of nothing but surplus-value that the industrialists D, E, F and others have deposited with him. From A's own point of view, this is not yet accumulated capital. But in fact, as far as D, E, F and the others are concerned, A is nothing but an agent who capitalizes the surplus-value they have appropriated.

We saw in Volume 1, Chapter 22, that accumulation — the turning of surplus-value into capital — is, in its real content, reproduction on an extended scale, whether that extension takes the form of adding new factories to the old ones, or of expanding, more intensively, the scale on which the business already runs.

The scale of production can grow in small doses. Part of the surplus-value can go toward improvements that either simply raise the productive power of the labour already employed, or also let that labour be worked more intensively at the same time. Or, where the working day is not legally limited, a small extra outlay of circulating capital — on materials and on wages — is enough to expand the scale of production without any increase in fixed capital: the fixed capital's daily hours of use are simply stretched longer, while its turnover period shortens to match. Or, when market conditions are favourable, the capitalized surplus-value may allow speculation in raw materials — operations the originally advanced capital would not have stretched to cover.

Still, it is clear that where a larger number of turnover periods brings more frequent realization of surplus-value within the year, there will be stretches when neither the working day can be lengthened nor individual improvements introduced. Expanding the whole business proportionally, meanwhile, is only possible within certain wider or narrower limits — limits set partly by the business's whole layout, its buildings for instance, and partly, as in agriculture, by how far the wage fund can stretch. And expanding the whole business, on top of that, calls for an amount of extra capital that only several years of accumulating surplus-value can supply.

So alongside the real accumulation — the actual turning of surplus-value into productive capital, with the corresponding reproduction on an extended scale — a second process runs: accumulating money, gathering part of the surplus-value into a hoard of latent money-capital — money that is capital only in waiting, doing nothing until it is large enough to be set to work as extra, active capital.

That is how things look from the standpoint of the individual capitalist. But as capitalist production develops, the credit system develops right alongside it. Money capital that a capitalist cannot yet use in his own business gets used by others, who pay him interest for it. For him, it functions as money capital in the specific sense — a sort of capital distinct from productive capital, but only in this sense: it works as capital in someone else's hands, not his own. It is clear that as surplus-value gets realized more often, and as the scale on which it is produced keeps rising, the proportion grows in which new money capital — money working as capital — gets thrown onto the money market, and from there gets absorbed again, at least for the most part, into expanded production.

The simplest form this extra, latent money capital can take is a hoard. It is possible that this hoard is extra gold or silver, obtained directly or indirectly through exchange with the countries that produce the precious metals — and only in this way does a country's stock of hoarded money actually grow in absolute terms. It is also possible — and this is the more common case — that the hoard is nothing but money withdrawn from domestic circulation, which has taken the form of a hoard in the hands of individual capitalists. It is possible, further, that this latent money capital consists merely of tokens of value — leaving credit-money aside here — or even of nothing more than claims, legal titles established by documents, that capitalists hold against other people. Whatever form this extra money capital takes, in every one of these cases, so far as it is capital still to come, it represents nothing whatsoever but extra legal titles, held in reserve, that capitalists hold on society's future, additional, annual production — never a stock of wealth already there.

"The mass of really accumulated wealth, considered by its size, is entirely insignificant compared with the productive powers of the society it belongs to, whatever that society's stage of civilization — or even just compared with that same society's actual consumption over a mere few years. So insignificant, that lawmakers and political economists ought to be giving their main attention to the productive powers and their future free development, not — as they have done up to now — to the mere accumulated wealth that catches the eye. By far the largest part of so-called accumulated wealth is only nominal, and does not consist of real objects — ships, houses, cotton goods, land improvements — but of mere legal titles: claims on society's future annual productive powers, titles produced and made permanent by the expedients and institutions of insecurity ... The use of such articles — accumulations of physical things, real wealth — as a mere means of letting their owners appropriate the wealth that society's future productive powers have yet to create: this use would gradually be taken from them by the natural laws of distribution, without any need for force; and with the help of co-operative labour, it would be taken from them within a few years." (William Thompson, Inquiry into the Principles of the Distribution of Wealth, London 1850, p. 453 — the book itself first appeared in 1824.)

"It is little considered, and by most people not even suspected, how extremely small a proportion — whether by size or by effect — society's actual accumulations bear to human productive powers, or even to the ordinary consumption of a single generation over just a few years. The reason is obvious enough, but the effect is very harmful. Wealth that is consumed each year vanishes with its use; it stands before the eye only for a moment, and makes its impression only while it is being enjoyed or used up. But the part of wealth that is only slowly consumed — furniture, machines, buildings — stands before our eyes from childhood to old age, lasting monuments to human effort. By virtue of owning this fixed, durable, slowly consumed part of public wealth — the land and raw materials, the tools worked with, the buildings that shelter the work — the owners of these things control, to their own advantage, the annual productive powers of every truly productive worker in society, however insignificant those objects may be next to the constantly recurring products of that labour. The population of Britain and Ireland is 20 million; the average consumption of each person, man, woman and child, is probably about £20 — together a wealth of about £400 million, the yearly product of labour consumed. The total accumulated capital of these countries, by the best estimate, does not exceed £1,200 million, or three times the annual product of labour; divided equally, that is £60 of capital per head. What matters here is the ratio, more than the more or less exact absolute size of these estimated sums. The interest on this whole capital would be enough to keep the whole population at their present standard of living for about two months of the year, and the whole accumulated capital itself — could buyers be found for it — would support them without any work at all for three whole years. At the end of which time, with no houses, clothes or food, they would have to starve, or else become the slaves of whoever had supported them through those three years. As three years stands to the lifetime of a healthy generation — say 40 years — so the size and importance of the real wealth, the accumulated capital even of the richest country, stands to its productive power, to the productive powers of a single human generation: not to what they could produce under sensible arrangements offering equal security, and above all with co-operative labour, but to what they actually, absolutely produce under the poor and discouraging shifts and dodges of insecurity ... And in order to preserve and perpetuate this seemingly enormous mass of existing capital — or rather the command and monopoly over the products of annual labour that it buys, in its present state of forced division — the whole frightful machinery, and the vices, crimes and sufferings of insecurity, must be kept going and perpetuated. Nothing can be accumulated until necessary wants are first satisfied, and the great stream of human inclination flows toward enjoyment; hence the comparatively insignificant amount of society's real wealth at any given moment. It is an endless cycle of production and consumption. Within this immense mass of annual production and consumption, the handful that is really accumulated would hardly be missed — and yet it is that handful of accumulation, not the mass of productive power, that has drawn the main attention. But this handful has been seized by a few, and turned into the instrument for appropriating the constantly recurring annual products of the labour of the great mass. Hence the decisive importance of such an instrument to this small number ... About a third of the nation's annual product is now taken from the producers under the name of public burdens, and consumed unproductively by people who give no equivalent for it — none, at least, that counts as such to the producers ... The eye of the crowd looks on astonished at the accumulated masses, especially when they are concentrated in a few hands. But the masses produced every year, like the endless, uncountable waves of a mighty river, roll past and lose themselves in the forgotten ocean of consumption. And yet it is this endless consumption that all enjoyment depends on — indeed the very existence of the whole human race. The size and distribution of this annual product ought above all else to be made the object of consideration. Real accumulation is of thoroughly secondary importance, and owes what importance it has almost entirely to its influence on the distribution of the annual product ... Real accumulation and distribution are here" (in Thompson's book) "always considered with reference and in subordination to productive power. In almost every other system, productive power has been considered with reference and in subordination to accumulation, and to perpetuating the existing way of distributing wealth. Next to preserving this existing way of distribution, the recurring misery or well-being of the whole human race has not been thought worth a single glance. Perpetuating the results of force, fraud and chance — that is what has been called security; and to preserve this false security, all the productive powers of the human race have been mercilessly sacrificed." (Same work, pp. 440–443.)

For reproduction, only two normal cases are possible — leaving aside disruptions that hold back even reproduction on the same scale as before.

Either reproduction takes place on the same scale as before.

Or capitalization of surplus-value takes place — accumulation.

Under simple reproduction, the capitalists take the surplus-value that gets produced and cashed in each year — or with several turnovers within the year — and consume it individually — that is, unproductively — themselves. This consumption doesn't go back into production, it's just spent.

The value of the total product breaks into three parts: surplus-value, the value that replaces the variable capital reproduced in it, and the value that replaces the constant capital used up in making it. Splitting the value this way changes absolutely nothing about how much of the product there is, or what it's worth, as it constantly flows into circulation and constantly flows back out again, to be used either as means of production or as means of consumption. Set the constant-capital part aside, and the only thing this split affects is how the annual product is divided between workers and capitalists.

Even simple reproduction, then, requires that part of the surplus-value constantly exist in money form — not in product — because it's only in that form that it can be turned into product when it comes time to consume it. This turning of surplus-value out of its original commodity form and into money still needs examining. To keep things simple, assume the simplest version of the problem: that only metallic money circulates — money that is itself a real equivalent, not a stand-in for one.

By the laws already worked out for simple commodity circulation (Volume 1, Chapter 3), the stock of metallic money in a country has to do more than just circulate the commodities. It has to be enough to absorb swings in the pace of circulation, changes in commodity prices, and shifts in how much money is used as a means of payment rather than as a plain medium of circulation. The split between money sitting as hoard and money actually circulating keeps changing, but the total — hoard plus circulating — always equals the whole money stock on hand. This stock, this mass of precious metal, is a social treasure built up gradually over time. As part of it wears away through use, it has to be replaced every year, like any other product. In reality this happens by trading part of the country's annual product, directly or indirectly, for the product of the countries that mine gold and silver. But because that trade crosses borders, it hides how simple the process really is. So, to reduce the problem to its plainest and clearest form, assume — as a simplifying assumption — that the gold and silver are mined within the very country under consideration, so that gold and silver production is one branch of that country's total social production.

Leaving aside gold or silver made into luxury articles, the least that must be produced each year is enough to replace the wear the money-metals suffer from circulating. And further: if the total value of the commodities produced and circulated each year grows, gold and silver production has to grow too — unless that growth in value, and the extra money it would otherwise take to circulate it (and to build the hoards that go with it), is offset instead by money circulating faster, or by money being used more as a means of payment, that is, by more purchases and sales cancelling each other out without any actual money changing hands.

So every year, part of society's labour-power and part of its means of production has to go into producing gold and silver.

Take the capitalists who run gold and silver production — who, since we're assuming simple reproduction, produce only within the limits of the average yearly wear-and-tear and the consumption that wear creates. They consume their whole surplus-value every year, capitalizing none of it, and throw it straight into circulation in money form, since for gold, money is the form it already comes in out of the ground - not a form it has to be turned into by being sold.

The same holds for wages — the money form in which variable capital is advanced. Here too, that money is replaced not by selling a product and turning it into money, but by a product whose natural form is money from the very start.

The same is true, finally, of the part of the precious-metal output equal in value to the constant capital used up along the way — both the circulating constant capital and the part of the fixed constant capital consumed during the year.

Look at the circuit — or turnover — of capital invested in precious-metal production, first in the form Money – Commodities … Production … More Money. Where the commodities bought with the first money are not just labour-power and materials but also fixed capital, of which only part of the value gets used up in production, then it's clear: the resulting More Money — the product — has to equal the variable capital laid out in wages, plus the circulating constant capital laid out in materials, plus the value-portion of the fixed capital worn away, plus the surplus-value. If the result were smaller than that, with gold's general value unchanged, the mine would be running at a loss. Or, if this were true across the board, gold's value would rise relative to commodities whose own value hadn't changed — meaning commodity prices would fall, and the money sum laid out at the start of the circuit would in future be smaller.

Look now just at the circulating part of the capital advanced at the start of that circuit. A fixed sum of money is advanced, thrown into circulation to pay for labour-power and buy materials. But the circuit of this capital does not pull that money back out of circulation in order to throw it in again. The product, in its very natural form, already is money — it doesn't need to be exchanged, doesn't need to pass through circulation, to become money. It leaves the production process and enters circulation not as commodity-capital that still has to turn into money-capital, but already as money-capital, ready to turn back into productive capital — that is, to buy fresh labour-power and materials all over again. The money form of the circulating capital used up in labour-power and materials is replaced not by selling the product, but by the product's own natural form. So it isn't replaced by pulling the same value back out of circulation in money form — it's replaced by extra, newly produced money.

Take this circulating capital as 500 pounds, with a turnover period of 5 weeks: a working period of 4 weeks, and a circulation period of only 1 week. From the start, money for the full 5 weeks has to be advanced — part of it held as a stock of materials, part of it kept on hand to be paid out gradually as wages. By the start of the 6th week, 400 pounds have flowed back and 100 pounds have been freed up. This keeps repeating. As before, for part of each turnover, 100 pounds sit in this freed-up state — but, just like the other 400 pounds, this 100 pounds is extra, newly produced money. Here there are 10 turnovers a year, so the year's product comes to 5,000 pounds of gold. (The circulation period here doesn't come from the time it takes to turn the commodity into money — it comes from the time it takes to turn money into the elements of production.)

With any other 500-pound capital turning over under the same conditions, the money form that keeps reappearing is a transformed shape of the commodity-capital produced — thrown into circulation every 4 weeks, and getting its money form back each time only because it's sold, that is, because the very sum of money it started out as gets periodically pulled back out of circulation. Here it's the opposite: in every turnover period, a new, extra mass of 500 pounds of money gets thrown into circulation straight out of the production process itself, constantly drawing materials and labour-power out of circulation in exchange. This money thrown into circulation is never pulled back out again by this capital's circuit — instead it keeps being added to, by fresh masses of gold newly produced.

Take the variable part of this circulating capital, set as before at 100 pounds. In ordinary commodity production, with ten turnovers a year, that 100 pounds would be enough to keep paying the labour-power. Here, in gold production, the same sum is enough too — but the 100 pounds that flows back every 5 weeks to pay the labour-power isn't a transformed shape of what the workers produced. It's simply part of their own constantly-renewed product. The gold producer pays his workers directly with part of the very gold they themselves produced. So the 1,000 pounds laid out each year in labour-power, and thrown into circulation by the workers who receive it, never comes back to its starting point by way of circulation.

Now for the fixed capital. Setting the business up in the first place requires laying out a larger sum of money-capital, which gets thrown into circulation. Like all fixed capital, it flows back only bit by bit, over several years. But it flows back as an immediate piece of the product itself — gold — not by way of selling the product and thereby, so to speak, gilding it. So it gradually gets its money form back not by pulling money out of circulation, but by piling up the matching part of the product. The money-capital restored this way is not a sum of money gradually withdrawn from circulation to balance out the sum originally thrown in for the fixed capital. It is an extra mass of money.

Finally, the surplus-value. It too equals part of the new gold product thrown into circulation each turnover period — and, on our assumption, it gets spent unproductively, laid out for means of subsistence and luxury goods.

But on our assumption, this whole year's gold production — which constantly draws labour-power and materials out of the market, without drawing any money out of it, while constantly supplying it with extra money — does nothing more than replace the money worn away over the year. It simply keeps the social money stock at full strength: a stock that exists constantly, though in shifting proportions, in the two forms of hoard and money actually in circulation.

By the law of commodity circulation, the money stock has to equal the money needed for circulation, plus a reserve of idle money that swells when trade contracts and drains away when it expands, above all to build up the reserve funds needed for making payments. What has to be paid in money — where payments aren't simply cancelled against each other — is the value of the commodities. That part of this value is surplus-value, meaning it cost the seller nothing to produce, changes absolutely nothing about that. Suppose all the producers owned their own means of production outright, so that circulation happened directly between these producers themselves. Setting aside the constant part of their capital, you could still divide their annual surplus-product into two parts, by analogy with the capitalist case: one part, a, which simply replaces what they need to live on, and another part, b, which they partly consume as luxuries and partly use to expand production. a then stands in for variable capital, b for surplus-value. But this division would have no effect at all on how much money is needed to circulate their total product. Other things being equal, the value of the commodities in circulation would be exactly the same, and so would the money needed for it. Given the same split of turnover periods, they would need the same money reserves too — the same part of their capital sitting constantly in money form — since, on this assumption, their production would still be commodity production, just as before. So the fact that part of the commodities' value is surplus-value changes absolutely nothing about the amount of money the business needs to run on.

An opponent of Tooke's, who holds to the form money-capital-more money, asks him how the capitalist manages to keep pulling more money out of circulation than he puts into it. Let's be clear about what's being asked here. This is not about where surplus-value comes from. That is the one real mystery, and from the capitalist standpoint it explains itself: the sum of value applied wouldn't be capital at all unless it grew by a surplus-value. Since it is assumed, from the start, to be capital, the surplus-value is simply taken for granted.

So the question is not: where does the surplus-value come from? It is: where does the money come from to turn it into cash?

But in bourgeois economics the existence of surplus-value goes without saying. So it isn't just assumed on its own — one thing more gets assumed along with it: that part of the mass of commodities thrown into circulation consists of surplus product, and so represents a value that the capitalist did not put into circulation as part of his capital. In other words, the capitalist throws a surplus over his capital into circulation together with his product, and then pulls that same surplus back out again.

The commodity-capital the capitalist throws into circulation is worth more (where this extra value comes from is neither explained nor understood, but from this same standpoint it's simply a fact) than the productive capital — labour-power plus means of production — that he withdrew from circulation to make it. Given this, it's clear why not only capitalist A but also B, C, D, and so on can each constantly pull more value out of circulation, by exchanging his commodity, than the value of the capital he originally advanced and keeps advancing again. A, B, C, D and the rest constantly throw a greater commodity-value into circulation, in the form of commodity-capital — this happens in as many different ways as there are capitals operating independently — than the value they withdraw from circulation in the form of productive capital. Which means: what they draw out of circulation as productive capital, and what they throw back in as surplus commodity-value, are two reciprocal shares of one division. Each capitalist must withdraw from circulation a value equal to the productive capital he advances, and just as constantly split off a further sum — the surplus of the commodity's value over the value of what went into producing it — which he likewise throws into circulation in commodity form.

But the commodity-capital has to be turned into money before it can turn back into productive capital, and before the surplus-value locked inside it can be spent. Where does this money come from? At first sight the question looks difficult, and neither Tooke nor anyone else has yet answered it.

Take the circulating capital advanced in money form — 500 pounds — whatever its turnover period may be, and let it stand for the whole circulating capital of society, that is, of the capitalist class. Let the surplus-value be 100 pounds. How, then, can the whole capitalist class constantly pull 600 pounds out of circulation, when it constantly puts in only 500?

Once the 500 pounds of money-capital has been turned into productive capital, that productive capital turns, in the course of the production process, into a commodity-value of 600 pounds. So what is now in circulation is not just a commodity-value of 500 pounds, equal to the money-capital originally advanced, but that plus a newly produced surplus-value of 100 pounds.

This extra 100 pounds of surplus-value is thrown into circulation in commodity form. There is no doubt about that. But that same operation does not supply the extra money needed to circulate this extra commodity-value.

The difficulty must not be talked away with plausible evasions.

For example: as for the constant circulating capital, it's clear that not everyone lays it out at the same moment. While capitalist A sells his commodity — so that, for him, the capital he advanced takes on money form — the reverse happens for the buyer B: his capital, which existed in money form, takes on the form of his means of production, the very things A has just produced. Through this one act, by which A gives his produced commodity-capital back its money form, B gives his own capital back its productive form, turning it from money form into means of production and labour-power; the same sum of money functions in this two-sided process just as it does in any simple act of W - G (commodity for money).

On the other hand, when A turns the money back into means of production, he buys from C, and C then pays B with it, and so on. That, it seems, would explain the whole process. But —

All the laws laid down about the quantity of money needed for the circulation of commodities are not changed in any way by the capitalist character of the production process.

So when it is said that the circulating capital of society that has to be advanced in money form comes to 500 pounds, this figure already assumes two things at once: that this is the sum advanced at any one moment, and also that this same sum sets more than 500 pounds' worth of productive capital in motion, because it serves in turn as the money-fund for one productive capital after another. This way of explaining things, in other words, already assumes the very money whose existence it's supposed to explain.

It might further be said: capitalist A produces goods that capitalist B consumes individually, unproductively. B's money, then, turns A's commodity-capital into cash, and so the same sum of money serves both to turn B's surplus-value into cash and to circulate A's constant circulating capital. But here the very question at issue is assumed even more directly. Namely: where does B get this money to cover his own spending in the first place? How did he himself turn this part of his product's surplus-value into cash?

It might further be said: the part of the circulating variable capital that A constantly advances to his workers keeps flowing back to him out of circulation, and only a shifting portion of it is tied up with him at any moment for paying wages. But some time passes between paying it out and its flowing back, and during that time the money paid out as wages can, among other things, also serve to turn surplus-value into cash. — But we know, first, that the longer this time is, the larger the stock of money capitalist A must constantly keep in reserve. Second, the worker spends the money and buys commodities with it, and so turns into cash, to that extent, the surplus-value locked inside those commodities. So the same money advanced in the form of variable capital also serves, to that extent, to turn surplus-value into cash. Without going any deeper into this question here, this much can be said: the consumption of the whole capitalist class, and of the unproductive people who depend on it, keeps pace with the consumption of the working class; so, at the same time as the workers throw money into circulation, the capitalists must also throw money into circulation to spend their surplus-value as revenue — and that same money must, for the same reason, be drawn back out of circulation. The explanation just given would only shrink the amount of money needed for this, not get rid of the need itself.

Finally, it might be said: a large quantity of money is constantly thrown into circulation for the first outlay on fixed capital, money that flows back to whoever put it in only gradually, bit by bit, over years. Can't this sum be enough to turn the surplus-value into cash? — The answer is that the 500 pounds (a sum that also covers the hoarding needed for reserve funds) may already include the use of part of that money as fixed capital — if not by the person who put it in, then by someone else. Besides, the sum spent on acquiring the goods that serve as fixed capital already assumes that the surplus-value locked inside those commodities has also been paid — and that is exactly the question: where does that money come from?

The general answer has already been given: if a mass of commodities worth x times 1,000 pounds has to circulate, it makes absolutely no difference to the amount of money needed for that circulation whether the value of this mass of commodities contains surplus-value or not, whether the commodities were produced capitalistically or not. So the problem, as such, does not exist. Given everything else — the speed at which money circulates, and so on — a fixed sum of money is required to circulate a commodity-value of x times 1,000 pounds, completely independent of how much or how little of that value falls to the immediate producers of these commodities. To the extent that any problem exists here, it is the same as the general problem: where does the sum of money needed to circulate a country's commodities come from?

All the same, from the standpoint of capitalist production, there does seem to be a special problem here. The reason is that here it is the capitalist who appears as the starting point from which the money is thrown into circulation. The money the worker spends on his means of subsistence already exists beforehand as the money form of variable capital, and so is originally thrown into circulation by the capitalist, as a means of buying or paying for labour-power. Besides this, the capitalist throws into circulation the money that for him originally forms the money form of his constant capital, both fixed and circulating; he spends it as a means of buying or paying for means of labour and materials of production. But beyond this, the capitalist does not appear as a starting point for the mass of money in circulation at all. Now, there are only two starting points: the capitalist and the worker. Every third party must either get money from these two classes in return for services, or, to the extent they get it without giving anything in return, they are co-owners of the surplus-value, in the form of rent, interest, and so on. That the surplus-value does not stay entirely in the industrial capitalist's pocket, but has to be shared by him with other people, has nothing to do with the question at hand. The question is how he turns his surplus-value into cash, not how the cash he gets for it is later divided up. So for our purposes the capitalist still counts as the sole owner of the surplus-value. As for the worker, it has already been said that he is only the secondary starting point, while the capitalist is the primary starting point, of the money the worker throws into circulation. The money first advanced as variable capital is already on its second circuit when the worker spends it to pay for his means of subsistence.

The capitalist class, then, remains the sole starting point of the circulation of money. If it needs 400 pounds to pay for means of production and 100 pounds to pay for labour-power, it throws 500 pounds into circulation. But the surplus-value locked inside the product, at a rate of surplus-value of 100 per cent, equals a value of 100 pounds. How, then, can it constantly pull 600 pounds out of circulation, when it constantly puts in only 500? Nothing comes from nothing. The capitalist class as a whole cannot pull out of circulation anything that was not put into it beforehand.

We are leaving aside here the fact that the sum of 400 pounds may well be enough, turning over ten times, to circulate means of production worth 4,000 pounds and labour worth 1,000 pounds, with the remaining 100 pounds likewise enough to circulate a surplus-value of 1,000 pounds. This ratio between the sum of money and the commodity-value it circulates makes no difference to the matter. The problem stays exactly the same. If the same pieces of money did not circulate several times over, 5,000 pounds would have to be thrown into circulation as capital, and 1,000 pounds would be needed to turn the surplus-value into cash. The question remains where this latter money comes from, whether it is 1,000 pounds or 100 pounds. Either way, it is a surplus over and above the money-capital thrown into circulation.

In fact, paradoxical as it looks at first sight, it is the capitalist class itself that throws into circulation the money used to realize the surplus-value locked inside the commodities. But note well: it throws this money in not as money advanced, that is, not as capital. It expends it as a means of purchase for its own individual consumption. So this money is not advanced by the capitalist class, even though the class is the starting point of its circulation.

Take a single capitalist starting out in business — a farmer, say. In the first year he lays out a money capital of 5,000 pounds: 4,000 for means of production and 1,000 for labour-power. The rate of surplus-value is 100%, so he pockets 1,000 pounds of surplus-value. That 5,000 pounds is all the money he lays out as capital. But he also has to live, and he gets no money in until the year ends. Say his own consumption costs 1,000 pounds. He needs that money too, and he might say he has to "advance" it for the first year. But this advancing has only subjective meaning — that is how it feels to him. In fact it means only this: for that first year he has to pay for his own living out of his own pocket, instead of out of the free production of his workers. He is not laying this money out as capital. He spends it, hands it over for an equivalent in means of subsistence, which he consumes. That value has left him in money, gone into circulation, and pulled commodity-values out of it. Those commodity-values he has consumed. So he no longer stands in any relation to their value. The money he paid with still exists as a piece of the money circulating in society. But the value of that money he has withdrawn from circulation in the form of products, and once the products he consumed are used up, their value is gone with them. As far as that money's value goes, it is finished for him. At the end of the year he throws a commodity-value of 6,000 pounds into circulation and sells it. That brings back to him: first, his advanced money capital of 5,000 pounds; second, his surplus-value of 1,000 pounds, now turned into money. He advanced 5,000 pounds as capital and threw it into circulation, and he draws 6,000 pounds back out of it — 5,000 for capital, 1,000 for surplus-value. That last 1,000 pounds is turned into money by means of the very money he himself threw into circulation, not as a capitalist but as a consumer — money he did not advance but spent. It now comes back to him as the money form of the surplus-value he produced. From now on this happens every year. But from the second year on, the 1,000 pounds he spends is, every time, the transformed money form of the surplus-value he himself produced. He spends it every year, and it flows back to him every year just the same.

If his capital turned over more often within the year, that would change nothing about the substance of it. What it would change is the length of time, and therefore the size of the sum, he would have to throw into circulation for his own consumption on top of his advanced money capital.

This money is not thrown into circulation by the capitalist as capital. But it is part of what makes him a capitalist that he can live, until his surplus-value flows back to him, off the means he already has in his possession.

In this case it was assumed that the sum of money the capitalist throws into circulation for his own consumption, up to the first return of his capital, is exactly equal to the surplus-value he has produced and therefore has to turn into money. For the individual capitalist this is plainly an arbitrary assumption. But it must hold true for the capitalist class as a whole, once simple reproduction is assumed. It only spells out what that assumption already says: that the whole of the surplus-value — but only the surplus-value, no part of the original capital stock — gets consumed unproductively.

It was assumed above that the total output of precious metals — set at 500 pounds — is only enough to replace the money that gets worn out.

The gold-producing capitalists own their whole product in gold — the part that replaces constant capital, the part that replaces variable capital, and the part that consists of surplus-value, all of it. So part of society's surplus-value consists of gold from the start, not of some other product that only turns into gold later, inside circulation. It exists as gold from the outset and is thrown into circulation to draw products out of it. The same holds here for wages, for variable capital, and for replacing the constant capital advanced. So if one part of the capitalist class throws a commodity-value into circulation that is larger — larger by the amount of the surplus-value — than the money capital it advanced, then another part of the capitalists throws a larger money-value into circulation — again larger by the amount of their surplus-value — than the commodity-value they are constantly drawing out of circulation to produce that gold. If one part of the capitalists is constantly pumping more money out of circulation than it pumps in, the gold-producing part is constantly pumping more money in than it draws out in means of production.

Now although part of this 500-pound product of gold is the gold-producers' own surplus-value, the whole sum is meant only to replace the money that the circulation of commodities needs. How much of it turns the surplus-value in commodities into money, and how much turns their other components of value into money, makes no difference here.

If you move gold production out of the country and into other countries, that changes absolutely nothing about the substance of it. Part of the social labour-power and social means of production in country A gets turned into a product — say, linen worth 500 pounds — which is exported to country B to buy gold there. The productive capital used this way in A throws no more commodities, as distinct from money, onto A's market than if it had been used directly in gold production. This product of A's takes the shape of 500 pounds of gold and enters A's circulation only as money. The part of society's surplus-value this product contains exists directly as money, and for country A it never exists in any other form. Although for the capitalists who produce the gold only part of the product is surplus-value and another part stands for the replacement of their capital, how much of this gold — beyond the circulating constant capital — replaces variable capital and how much represents surplus-value depends entirely on the respective shares that wages and surplus-value form of the value of the circulating commodities. The part that forms surplus-value gets divided among the various members of the capitalist class. It is constantly spent by them on their own consumption and taken back in again by selling new product — indeed it is exactly this buying and selling that makes the money needed to turn the surplus-value into money circulate among them at all. Even so, part of society's surplus-value, though in shifting portions, sits in the pockets of the capitalists in the form of money — just as part of the wage sits, for at least part of the week, in the pockets of the workers in the form of money. And this part is not limited by the part of the gold-product that originally forms the gold-producing capitalists' own surplus-value. It is limited, as already said, by the proportion in which that 500-pound product is divided up between capitalists and workers generally, and by the proportion in which the commodity-value that has to circulate is made up of surplus-value and the other components of value.

However, the part of the surplus-value that doesn't exist in other commodities but alongside them, in money, consists of part of the annually produced gold only to the extent that part of that year's gold output circulates to realize surplus-value. The other part of the money — the part that sits, in shifting portions, as the money form of their surplus-value in the hands of the capitalist class — is not an element of the gold produced that year at all, but of the masses of money accumulated earlier in the country.

On our assumption, the annual gold output of 500 pounds is only just enough to replace the money that gets used up each year.

So let us keep our eye only on this 500 pounds, and set aside the part of the annually produced mass of commodities that circulates by means of money accumulated in earlier years. Then the surplus-value produced in commodity form already finds money waiting for it in circulation to turn it into money — and it finds that money simply because, on the other side, surplus-value is being produced annually in the form of gold. The same holds for the other parts of the 500-pound gold-product that replace the advanced money capital.

There are two things to note here.

First, it follows that the surplus-value the capitalists spend in money — and likewise the variable and other productive capital they advance in money — is in fact the product of the workers, namely the workers employed in gold production. They newly produce both the part of the gold-product that gets "advanced" to them as wages and the part of the gold-product in which the gold-producing capitalists' surplus-value directly presents itself. As for the part of the gold-product that only replaces the constant capital-value advanced for its production, it too reappears in gold form — in some product, generally — only through the workers' annual labour. At the start of the business it was originally handed over by the capitalist in money that was not newly produced but formed part of society's circulating money-mass. But to the extent that it gets replaced by new product, by additional gold, it is the workers' annual product. Here too the capitalist's "advance" turns out to be only a form — one that comes from the fact that the worker owns neither his own means of production nor, during production, the means of subsistence produced by other workers.

But second, as for the mass of money that exists independently of this annual 500-pound replacement — partly as a hoard, partly as money in circulation — things must stand with it exactly as they still stand, every year, with this 500 pounds; that is, it must originally have arisen the same way this 500 pounds still does. We come back to this point at the end of this subsection. First, a few more remarks.

We already saw, when we looked at turnover, that if everything else stays the same, a change in how long the turnover period is changes how much money capital is needed to keep production running at the same scale. So the way money circulates has to be elastic enough to stretch and shrink along with that change.

Now suppose everything else stays fixed too - the length of the working day, its intensity, its productive power - but the split of the value produced between wages and surplus-value changes, so that either wages rise and surplus-value falls, or the other way round. This shift does not touch the amount of money in circulation at all. It can happen without any expansion or contraction of the money that is circulating. Take the case where wages rise generally and - under the assumptions we are making - the rate of surplus-value falls generally, while, also by assumption, the value of the mass of goods in circulation does not change. In this case the money capital that must be advanced as variable capital does grow - that is, the amount of money serving in that role grows. But surplus-value shrinks by exactly as much as that money grows, and so does the amount of money needed to turn that surplus-value into cash. The total sum of money needed to turn the value of the goods into cash is untouched by all this, just as that value itself is untouched. The cost price of the goods rises for the individual capitalist, but their social price of production stays the same. What changes is only the ratio in which that price of production - leaving the constant part of value aside - splits into wages and profit.

But, the objection runs, a bigger outlay of variable money capital - the value of money being assumed unchanged, of course - simply means more money in workers' hands. From this, it is said, follows greater demand from workers for goods. And from that follows a rise in the price of goods. Or else it is put this way: if wages rise, capitalists raise the prices of their goods. Either way, on this view, a general rise in wages causes goods to rise in price. So, whichever of the two explanations one prefers, a larger amount of money must be needed to circulate the goods.

Take the first version first. A rise in wages will chiefly make workers demand more of the goods they need directly. To a smaller degree it will also increase their demand for luxuries, or create demand for things that were previously outside what they could buy at all. The sudden, larger-scale demand for necessities will certainly push their price up for a while. The result: more of society's capital goes into producing necessities, and less into producing luxuries - because luxuries fall in price, since capitalists now have less surplus-value and so demand less of them. But wherever workers themselves buy luxuries, the wage rise, to that extent, does not push up the price of necessities at all - it simply changes who is buying the luxuries. More luxury goods now go to workers' consumption, and proportionally fewer to capitalists' consumption. That is all there is to it. After some to-and-fro, the same total value of goods circulates as before. As for the momentary swings, their only real effect is to throw idle money capital into domestic circulation - money that had until then been looking for an outlet in speculation on the stock exchange or abroad.

Now the second version. If it were within the power of capitalist producers to raise the prices of their goods at will, they could do this, and would do it, even without any rise in wages. Wages would then never rise while the prices of goods were falling. And the capitalist class would never resist the trade unions at all - since they could always do, in every case, what they now do only as an exception, under particular, local circumstances: use every wage rise as an occasion to push up the price of goods by far more than the wage rise itself, and pocket a bigger profit for it.

The claim that capitalists can raise the price of luxuries because demand for them has fallen - because capitalists' own reduced income has cut their demand - would be a strikingly original way to apply the law of supply and demand. Set aside the pure displacement of buyers, workers instead of capitalists - and to the extent that this displacement happens, workers' demand does not push up the price of necessities, because whatever part of their extra wages workers spend on luxuries, they cannot also spend on necessities. Apart from that, the price of luxuries falls because demand for them has fallen. As a result, capital is withdrawn from producing them, until the supply shrinks down to match their changed role in the process of social production. With this reduced

production, luxury prices rise back to their normal level, their value otherwise being unchanged. For as long as this contraction, this process of adjustment, continues, the production of necessities keeps drawing in - at their now-higher prices - exactly as much capital as is withdrawn from the other branch of production, until demand is satisfied. Then balance returns, and the end result of the whole process is that society's capital, and with it its money capital, is now divided between producing necessities and producing luxuries in a changed proportion.

The whole objection is a scare-shot fired by the capitalists and their economic yes-men.

The facts that supply the pretext for this scare-shot are of three kinds.

First: it is a general law of money circulation that when the sum of the prices of the goods in circulation rises - whether that rise happens for the same mass of goods or for a larger one - then, other things equal, the amount of circulating money grows. Here, effect is being mistaken for cause. Wages rise together with the rising price of necessities - but only in exceptional cases, and even then only partly keeping pace with it. Their rise is the consequence of goods rising in price, not its cause.

Second: a partial or local rise in wages - that is, a rise in only a few branches of production - can produce a local rise in the price of what those branches make. But even this depends on many conditions: for instance, that wages there were not abnormally low to begin with, so that the rate of profit was not abnormally high; that the market for these goods does not shrink because of the price rise, so that no prior cut in supply is needed to make the price rise stick; and so on.

Third: with a general rise in wages, the price of goods rises in branches of industry where variable capital predominates, but it falls, for that very reason, in branches where constant - or rather fixed - capital predominates.

It already emerged when we looked at simple commodity circulation that although the money-form taken on by any given quantity of goods, within circulation, is only fleeting, the money that disappears from one person's hand in a commodity's transformation necessarily turns up in someone else's. So it is not just that goods are all round exchanged for, or replace, one another - this replacement is also mediated and accompanied, all round, by money settling somewhere. As Volume One put it: 'The replacement of one commodity by another leaves the money-commodity sticking in a third pair of hands at the same time. Circulation is constantly sweating out money.' The very same fact, on the basis of capitalist commodity production, shows up as this: a part of capital constantly exists in the form of money capital, and a part of surplus-value likewise constantly sits, in money form, in the hands of its owners.

Apart from this, the circuit of money - that is, money's return to its starting point, so far as this forms one moment of capital's turnover - is a quite different phenomenon from the circulation of money, indeed the opposite one. Circulation expresses money's steady movement away from its starting point, through a series of hands. Even so, a faster turnover does, by that very fact, bring a faster circulation along with it.

Take variable capital first. Say a money capital of 500 pounds turns over ten times a year in the form of variable capital. Then clearly this slice of the circulating money moves a sum ten times its own size - 5,000 pounds - over the year. It passes between capitalist and worker ten times a year: the worker is paid, and pays out, ten times a year with the very same slice of the circulating money. If, at the same scale of production, this variable capital turned over only once a year, then only a single movement of 5,000 pounds would take place.

Further: say the constant part of the circulating capital is 1,000 pounds. If the capital turns over ten times, the capitalist sells his goods ten times a year, and so also sells the constant circulating part of their value ten times. The same slice of circulating money, 1,000 pounds, passes ten times a year out of the hands of its owners into the capitalist's hands. That is ten changes of hands for this money, from one person to another.

Second: the capitalist also buys means of production ten times a year - again, ten more movements of money from one hand to another. With 1,000 pounds in money, the industrial capitalist sells goods worth 10,000 pounds, and buys goods worth 10,000 pounds again. By moving twenty times, that 1,000 pounds in money has circulated a stock of goods worth 20,000 pounds.

Finally, with faster turnover, the part of the money that realizes surplus-value also moves faster.

But the reverse does not hold. A faster movement of money does not necessarily bring with it a faster turnover of capital, and so a faster turnover of money - that is, it does not necessarily mean the process of reproduction is shortened and renewed more quickly.

Money moves faster whenever a larger mass of transactions is carried out with the same amount of money. This can happen even with the same reproduction periods for capital, simply because the technical arrangements for moving money have changed. Further, the mass of transactions in which money moves can grow without expressing any real turnover of goods at all - think of speculative dealing on the stock exchange. On the other hand, movements of money can disappear altogether. For instance, where the farmer is himself the landowner, no money moves between tenant and landowner; where the industrial capitalist himself owns the capital, none moves between him and a lender.

As for how a country first builds up a hoard of money, and how a few people come to seize it, there is no need to go further into that here.

The capitalist mode of production has wage-labour as its basis, and with it, paying the worker in money, and in general turning payment in kind into payment in money. This can only develop on a large scale, and take deep root, where a country already has a mass of money big enough for circulation, and for the hoard-formation - reserve funds and the like - that circulation requires. This is a historical precondition. It should not be understood as meaning that a sufficient mass of hoarded money forms first, and only then does capitalist production begin. Rather, capitalist production develops together with the development of its own conditions, and one of those conditions is an adequate supply of precious metals. That is why the increased supply of precious metals since the sixteenth century forms an essential moment in the history of how capitalist production developed. But so far as it concerns the further supply of money-material that capitalist production goes on needing, the picture is this: on one side, surplus-value in the form of a product is thrown into circulation without the money needed to turn it into cash; on the other side, surplus-value in the form of gold is thrown in without any product having first been turned into money at all.

The extra goods that need to be turned into money find the sum of money they need already waiting for them - because, on the other side, extra gold and silver is thrown into circulation not through exchange but through production itself, and that gold and silver in turn needs to be turned into goods.

When accumulation takes the form of reproduction on an enlarged scale, it clearly poses no new problem for the circulation of money.

Take first the additional money capital required for the function of the growing productive capital. It is supplied by the part of the realized surplus-value that the capitalists throw into circulation as money capital, instead of as the money-form of revenue. The money is already in the capitalists' hands. Only its use is different.

But now, as a result of this additional productive capital, an additional mass of commodities — its product — gets thrown into circulation. Together with this additional mass of commodities, part of the additional money needed to realize it was thrown into circulation too, to the extent that the value of this commodity mass equals the value of the productive capital used up in producing it. This additional sum of money was advanced precisely as additional money capital, and so flows back to the capitalist as his capital turns over. The same question comes up again as before: where does the additional money come from to realize the additional surplus-value, which now exists in the form of commodities?

The general answer is again the same. The total price of the circulating mass of commodities has risen — not because the prices of a given mass of commodities have gone up, but because the mass of commodities now circulating is larger than before, without any fall in prices offsetting that. The additional money needed to circulate this larger, more valuable mass of commodities has to be found either through greater economizing of the circulating money — say, by offsetting payments against each other, or by speeding up how fast the same coins change hands — or through converting money out of its hoard-form into its circulating form. This second source covers more than idle money capital coming into use as a means of buying or paying. It also covers money capital that is already serving as a reserve fund, which, even while performing that reserve function for its owner, is at the same time actively circulating for society (as with bank deposits, which are constantly being lent out) — so that it performs a double function. And it covers the economizing of reserve funds of coin that would otherwise sit idle.

In an earlier book of his own, Marx put it this way: for money to keep flowing as coin, the coin must keep condensing back into money. The constant circulation of coin depends on its constantly coming to rest — in larger or smaller portions — in reserve funds of coin, funds that arise out of circulation just as much as they make it possible, and whose formation, distribution, dissolution and re-formation never stop changing: their existence is constantly disappearing, and their disappearing is constantly there. Adam Smith expressed this endless conversion of coin into money and money into coin by saying that every owner of a commodity must always keep in stock, alongside the particular commodity he sells, a certain quantity of the universal commodity he buys with. We saw that in the circuit commodity-money-commodity, the second link, money-commodity, keeps splitting into a series of purchases carried out not all at once but one after another over time, so that one portion of the money circulates as coin while another sits idle as money. Money here is really just coin held in suspension, and the individual pieces making up the circulating mass of coin keep changing which of the two forms they are in. This first conversion of the means of circulation into money is therefore only a technical moment within the circulation of money itself. Here 'coin', as against 'money', is used for money in its function purely as a means of circulation, as distinct from its other functions.

Only if none of the means already mentioned — economizing the circulating money, or drawing hoarded money back into use — suffice, must additional gold production take place. Or, what comes to the same thing, part of the additional product must be exchanged, directly or indirectly, for gold, the product of the gold-producing countries.

The whole sum of labour-power and social means of production spent every year on producing gold and silver, as instruments of circulation, is a heavy item among the faux frais — the incidental running costs — of the capitalist mode of production, and indeed of any mode of production founded on producing for exchange. It withdraws from social use a corresponding sum of possible additional means of production and consumption — that is, of real wealth. To the extent that the costs of this expensive circulation-machinery are cut, while the scale of production stays the same or expands by a given amount, the productive power of social labour is increased by just that much. So to the extent that the aids developed along with the credit system have this effect, they directly increase capitalist wealth — whether because a large part of the social process of production and labour is carried out without any real money intervening at all, or because the capacity of the money that really is functioning is increased.

This also settles the silly question of whether capitalist production, at its present scale, would be possible without the credit system — looking at it purely from this angle, that is, with nothing but metallic circulation. Clearly it would not. It would instead have run up against limits set by the scale of precious-metal production. On the other hand, one should not build up mystical notions about the productive power of the credit system, in so far as it makes money capital available or sets it free. Taking this further belongs elsewhere.

We now need to look at the case where what happens is not real accumulation — that is, not an immediate expansion of the scale of production — but rather where part of the realized surplus-value is piled up for a longer or shorter time as a money reserve fund, to be turned into productive capital only later.

Where the money that piles up as this reserve fund is additional money, the matter is self-evident: it can only be part of the surplus gold brought in from the gold-producing countries. And it should be noted here that the national product given up in exchange for this gold no longer exists in the country. It has been sent abroad in exchange for gold.

But suppose instead that the same total mass of money remains in the country as before. Then the money that has been piling up, and keeps piling up, has flowed out of circulation; only its function has changed. Out of circulating money it has gradually turned into latent money capital, forming itself in that state.

The money piled up here is the money-form of a commodity that has been sold — specifically, of the part of its value that represents surplus-value for its owner. (We are assuming here that the credit system does not exist.) The capitalist who has piled up this money has, to that extent, sold without buying.

If we picture this piling-up of money happening only on a partial scale, there is nothing here that needs explaining. One group of capitalists holds back part of the money it got from selling its product, without taking any product off the market in return. Another group, meanwhile — apart from the money capital it constantly needs back for running production — turns its money entirely into product. Part of the product thrown onto the market as the bearer of surplus-value consists of means of production, or of the real elements of variable capital, namely necessary means of subsistence. So it can serve at once to expand production. None of this assumes that one group of capitalists piles up money capital while the other consumes the whole of its surplus-value. It assumes only that one group carries out its accumulation in money-form, forming latent money capital, while the other really accumulates — that is, expands the scale of production, really extends its productive capital. The money mass on hand remains sufficient for the needs of circulation, even if first one group piles up money while the other expands production, and then the other way round. Money piling up on one side can, moreover, happen even without cash, simply through the piling-up of claims on debt.

But the difficulty arises once we assume not partial, but general, accumulation of money capital across the whole capitalist class. Outside this class there is, on our assumption — the general and exclusive rule of capitalist production — no other class at all except the working class. Everything the working class buys is equal to the sum of its wages, equal to the sum of the variable capital advanced by the capitalist class as a whole. This money flows back to the capitalists when they sell their product to the working class, and their variable capital thereby regains its money-form. Let the sum of variable capital be x times £100 — meaning the sum not of the variable capital advanced over the year, but of the variable capital actually used. How much or how little money, depending on the speed of turnover, is advanced to cover this variable-capital value during the year makes no difference to the question at hand. With this x times £100 of capital, the capitalist class buys a certain quantity of labour-power, or pays wages to a certain number of workers — first transaction. The workers use the same sum to buy a quantity of commodities from the capitalists, and by that the sum of x times £100 flows back into the capitalists' hands — second transaction. And this repeats endlessly. So the sum of x times £100 can never enable the working class to buy the part of the product that represents constant capital — let alone the part that represents the capitalist class's surplus-value. With their x times £100, the workers can only ever buy a portion of the value of the social product equal to the value-portion made up by the variable capital advanced.

Leaving aside the case where this piling-up of money on every side expresses nothing more than the distribution of newly imported precious metal — in whatever proportion — among the various individual capitalists: how, then, is the capitalist class as a whole supposed to accumulate money at all?

The whole capitalist class would have to sell part of its product without buying again in return. That they all hold a certain fund of money, which they throw into circulation as a means of circulating their own consumption, and of which a certain part flows back to each of them again, is nothing mysterious at all. But this money fund then exists precisely as a circulation fund, formed by turning surplus-value into money — never as latent money capital.

If we look at how the matter actually plays out in reality, the latent money capital piled up for later use consists of:

1. Deposits in banks — and the sum of money the bank actually has at its disposal is comparatively small. What is piled up here is only nominally money capital. What is really piled up are claims to money, which are only convertible into money (to the extent they ever are converted) because a balance holds between the money withdrawn and the money deposited. What actually sits in the bank's hands as money is, relatively speaking, only a small sum.

2. Government bonds. These are not capital at all — merely claims to debt on the nation's annual product.

3. Shares. Provided there is no swindle involved, these are titles of ownership in real capital belonging to a corporation, and a claim on the surplus-value flowing from it each year.

In each of these three cases — deposits, government bonds, shares — there is no piling-up of money at all. What appears on one side as an accumulation of money capital appears on the other as a constant, real expenditure of money. Whether the money is spent by the person it belongs to, or by others who owe it to him, makes no difference to the matter.

On the basis of capitalist production, forming a hoard as such is never the purpose — it is always only the result: either of a stoppage in circulation, where larger sums of money than usual take on the form of a hoard; or of accumulations brought about by the turnover of capital; or, finally, a hoard is simply money capital forming itself, for the time being in latent form, destined to function later as productive capital.

So when, on one side, part of the surplus-value realized in money is withdrawn from circulation and piled up as a hoard, at the very same time another part of the surplus-value is constantly being turned into productive capital. Except for the case of distributing additional precious metal among the capitalist class, the piling-up of money never happens at every point at once.

Exactly the same holds for the part of the annual product that represents surplus-value in commodity-form as holds for the rest of the annual product. Circulating it requires a certain sum of money. This sum of money belongs to the capitalist class just as much as the annual mass of commodities representing surplus-value does. It is originally thrown into circulation by the capitalist class itself, and it is constantly redistributed among them anew through circulation itself. As with the circulation of coin generally, part of this mass sits idle at constantly shifting points while another part keeps circulating. Whether part of this piling-up is deliberate, meant to form money capital, makes no difference to the matter.

We have left aside here the chance events of circulation, through which one capitalist grabs hold of a piece of another's surplus-value, or even of his capital, giving rise to a one-sided accumulation and centralization of both money capital and productive capital. So, for example, part of the surplus-value that A piles up as money capital, having seized it in this way, may be a piece of B's surplus-value that never flows back to him.