In what way is the ratio of exchange fixed between the precious metals and other commodities? There is a common notion that the function of the precious metals as money involves some peculiar relation between them and commodities in regard to the adjustment of their ratio of exchange. It is said, or assumed, that every portion of the precious metals not wanted for any other purpose, becomes money;" that the ratio of exchange with other articles rises or falls as there is less or more more "money," the fall or rise being proportionate to the change in the quantity of "money ;" and that this money use is also so much the preponderant use, that nothing else is material in settling the ratio of exchange between the precious metals and other commodities. It would not be going too far to say that this notion is at the root of the bimetallic theory, so far as bimetallism is based on any consistent and substantial theory. Supplemented by the NEW SERIES.-VOL. LI., No. 1.

com

Old Series vols. plete

further assumption that gold and silver are not only interchangeable as " "" but money, that each.can be made to take the place of the other, by legislation, at a given ratio, which will continue to operate until one or the other is wholly displaced, what is known as the quantitative theory of money is really the basis of the whole bimetallic structure. Holding myself the view that there is "a" relation between the quantity of the precious metals and prices, I have been frequently claimed by bimetallists as going a long way with them. I avail myself, therefore, of the present opportunity to give an answer to the question as to how the ratio of exchange between the precious metals and other commodities is fixed. I hold most fully to the view that it is fixed in no other way than is any other ratio, viz. by supply and demand, and by the cost of production of the last margin of supply necessary to meet the last margin of demand. There is "a relation be

tween the quantity of money and prices, but it is rather one in which prices assist in determining the quantity of the precious metals to be used as money, and not one in which prices are themselves determined by that quantity. There are some complicated elements in the problem; but this is the substantial result. In no respect, therefore, do I go any way with the bimetallist, not even the fraction of an inch; and, apart from the interest of the present problem itself, I am the more ready to avail myself of the opportunity of discussing it, because it enables me to explain how different is my own view of the relation between the quantity of money and prices from that quantitative theory of money which, with its supplement as above described, is the foundation of the bimetallic theory itself.

The precious metals, it is admitted on all sides, have an extensive non monetary use. They are merchandise as well as inoney. But few people perhaps realize that probably this non-monetary use is preponderant over the monetary use itself. The assumption to the contrary is, in fact, made by bimetallists and others as if there could be no question of it. The vast etores of coin in existence and circulating in people's hands are pointed to, and the use being assumed to be entirely "monetary," this monetary use is considered to be overwhelming. What is the annual production, it is said, of the precious metals as compared with the enormous mass of money?

But the mass of the precious metals in an uncoined form must be enormous. In the form of plate and ornaments there is endless gold and silver. The belief to the contrary appears to be due to an impression that only a small proportion of the wealth of modern societies is in plate and ornaments that the days when people kept their wealth in this form are past. What seems to be forgotten is that the wealth of modern societies is itself such that while the proportion of that wealth kept in plate and ornaments is indefinitely less than it was, yet the amount so kept may be large in proportion to the amount of the precious metals themselves. The two proportions are entirely distinct and unconnected. If, however, it is considered for a moment how indestructible are articles composed of the precious metals,

In

how steady is the conversion of the precious metals into such forms, though the amount annually converted may not be large, and for how many centuries the accumulation of objects of value has been going on, it will be seen that in most of the rich countries of the world the plate and ornaments must be a large mass. It is a moderate estimate that in this country alone-in plate, in watches, in jewelry and ornaments-there cannot be less than 50,000,000l. worth of gold, which represents no more than an accumulation of half a million per annum for a century, not to speak of the older accumulations at all. There is probably an equal amount of silver, though the fashion of solid silver plate has for many years died out. England, then, the stock of gold held for non-monetary uses is probably not far short of, if not equal to, the stock held for monetary uses, which is probably little more than about 60,000,0007. The stock of silver, again, in England held for nonmonetary uses must be two or three times the stock of silver money, which is little over 20,000,000l. This is not a statistical paper, but the figures may illustrate what the facts are throughout the rest of the civilized and semi-civilized world, where greater taste for ornaments may compensate to some extent the smaller wealth of the people compared with England. It is enough for the present purpose to indicate that there must be an enormous mass of gold and silver in existence and used for non-monetary purposes.

The demand for non-monetary purposes on the annual production is also preponderant in the case of gold, and very large in the case of silver. About two-thirds of the gold annually produced is taken for the arts; and if the consumption of India is included, as being either for simple hoarding, or for the arts, and in no case for the purpose of circulating money, then the demand for gold for non-monetary purposes appears almost equal to the entire annual production. The normal demand for money proper it is almost impossible to state, owing to the amount of recoinage and other difficulties; but it may be doubted whether the annual addition to circulating gold money in normal years can be anything nearly so great. Of silver, apart from India, about a fourth or fifth of the annual production is consumed in the arts; but if the Indian consumption

is included, as being mostly of a nonmonetary kind, about half the annual production of silver may be considered as required for non-monetary uses. Not only, therefore, is the non-monetary stock of the precious metals enormous, but the preponderating demands falling on the annual production are also non-monetary.

And both as regards this mass of the precious metals in existence not used as money, and the demands on the annual production for non-monetary uses, the same conditions as to the ratio of exchange of the metals with other articles must exist, as exist for those other articles themselves in their exchanges with each other, unless in the case of gold and silver their use as money should alter the conditions. The proposition is self-evident. The precious metals, so far, are ordinary merchandise and nothing more.

What I have next to point out is that as regards even the monetary uses of the precious metals, there are different uses. There is "money" and "money." The precious metals, when used for one kind of monetary purpose, may remain obviously under the ordinary conditions of merchandise, although not obviously so when used for another kind of monetary purpose. It is convenient in any case to make distinctions, and to look at the matter in detail instead of speaking of the monetary use in a vague and general way. Three principal kinds of employment of the precious metals for monetary purposes are apparently to be distinguished: 1. They are employed for token or quasitoken coinage, i.e. for the retail payments of society. 2. They are employed as reserves in banks, or other hands, forming the guarantee of paper money and checks, and thus becoming the instrument of the wholesale payments of society. With this employment may be included the use of the precious metals as an instrument of international remittance. 3. They are used as a means of hoarding. Only the first two of these employments can be spoken of, I believe, as a proper monetary employment. At any rate, although the precious metals, when hoarded, whether in coin or in bullion, may be considered as potential money, they are clearly not money in circulation, and a distinction must be made between the use for money in circulation and the use for hoarding.

We may begin with the hoarding use.

When in demand for hoarding, the precious metals, although they may be in the form of coin, remain mere merchandise. They are subject to the same laws respecting their ratio of exchange as diamonds, valuable pictures, or any other valuable object which may be hoarded. The motives of hoarding, and the price to be paid in meal or in malt" for the hoards, determine the demand; and the price to be obtained, which acts upon both the existing hoards as well as upon the annual production, when any new hoard is in question, determines the supply. The possibility of using the hoards as money, especially when in the form of coin, may be an element in their value; but it is only one element out of many, and for this purpose, accordingly, the precious metals are practically merchandise only.

Equally with the stock of the precious metals for purely non-monetary uses, the stocks of the precious metals hoarded are very large. The military and quasimilitary chests of military Governments like France, Germany, and Russia: the accumulations of the precious metals in those countries, far beyond any strict requirement of monetary circulation: are enormous. The United States, again, has accumulated both gold and silver in its Government vaults far beyond any ordinary monetary necessity. There are large private hoards besides all over the world, but especially in India, where both gold and silver are largely hoarded. What. ever the motives may be which determine these hoards, the hoards themselves are not money in circulation in any form, and the demand to replenish them is not a demand for money,' " and the supply of these demands is not a supply of

money, " which can help to make any such relation between the quantity of money and prices as the quantitative theory of money, and with it the bimetallic the

which is in one sense a monetary use of the precious metals, is to be included, for the purposes of the present discussion, in the category of merchandise uses where gold and silver are subject to the same conditions as regards their ratio of exchange with other articles as are those other articles themselves.

We come then to the more special monetary uses of the precious metals as above defined. And here again we find that as regards the most important of these in respect of quantity, viz. the use of the precious metals as token money, or quasitoken money, the demand for them must also be viewed as an ordinary merchandise demand. The point is so important as to excuse a somewhat full exposition.

As regards all kinds of token money, then, I have to put forward the proposition that the general economic circumstances of a community of which the range of prices of staple articles is an important part, but still only a part, determine in ordinary circumstances the quantity of the precious metals used as money in circulation in that form. The range of incomes seems even more important in this connection than prices ordinarily so called; but the two are interconnected, and incomes are a part of "prices," using the word in its most general sense. Further, the quantities of the different kinds of the precious metals so used as money may be considered as a fixed amount of each kind per head of the population, or rather an amount oscillating between fixed limits according to the seasons and the ebb and flow of credit. The amounts at any rate do not vary proportionately with small or ordinary fluctuations of prices, though they are liable to great changes with changes of magnitude in economic circumstances, including, in such changes of magnitude, great changes in the range of incomes and the range of prices of staple commodities.

Take first the case of copper or nickel money, which is all the better for illustration because copper and nickel, though used for token money, are not precious metals. Apparently, then, for a comcommunity of given numbers in a certain state of civilization and economic development, only a definite amount of such small money is required, whatever the range of prices may be. The same (or nearly the same) copper money will do the work

which copper has to do in such a community at almost any range of prices. The statistics of copper coinage show that it is a machine whose size is increased automatically as population increases-more rapidly, perhaps, in good times (when prices rise) than in bad times (when prices fall), but not in such a way at any time as to make any proportion between the changes in quantity and the changes in prices.

What is true of copper money is true of silver money in a country like England. This money being wanted for small change, the quantity in use varies only as copper money does, and from similar causes. The determining factor is a custom and habit of the people, which requires so much silver money per head. At a point, no doubt, silver might tend to go out of use, and copper come in on the one side in place of it, and gold on the other; but the limits of change are ap parently very wide.

Mutatis mutandis, it is obvious, the same remarks must apply to that part of the gold money in a country like England which is either explicitly token money or which, though standard money and unlimited legal tender, is really used as a kind of small change only—that is, the whole stock of gold coin in a country like England which is neither held as reserve in the Bank of England nor hoarded, the banking system reducing the uses of gold coin in circulation to those really of small change only. The amount of such small change must be viewed as strictly regulated by the habits and customs of the people, remaining at the same chronic amount with given habits and customs, and not changing-or, at any rate, not changing greatly-according to the ordinary fluctuations of prices.

The point when stated is so obvious as to seem hardly worth laboring; but it may be pointed out that the analogy of the circulation of paper money of small denominations in a country where the quantities of such paper in circulation are exactly ascertainable, quite supports the conclusion, although so little is known of the circulation of gold itself that it cannot be directly proved. The paper (and this is true of inconvertible as well as convertible paper) is very nearly a fixed quantity per head in such countries, or rather a quantity varying between fixed points ac