of wages and salaries, may well be given an equal measure of importance. Further, when wages and salaries do advance the advance is apt to be very general, while a considerable and increasing variety of commodities and services are sold at customary prices, susceptible to little or no change during a short period of years. But space lacks for a consideration in detail of this most difficult subject. Enough has perhaps been said to indicate the fragmentary character of the analysis which Professor Fisher has brought forward.

The second assumption, that price changes of a disturbing character would not occur in the absence of the initial stimulus of an increase in the supply of money, would seem to be equally untenable. It would seem entirely possible, tho no doubt less probable, that through the increase of the velocity of the circulation of money and deposits, and the expansion of deposits upon a stationary cash foundation, an increase in prices might occur sufficient to give rise to all the troubles which culminate in a commercial crisis. Further, it may be noted that the possible range of price fluctuation, from causes other than changes in the money supply, is apt to be particularly wide in countries in which general conditions and the temperament of the people favor enterprise and speculative activity. In such countries deposits fall far below the customary ratio to cash reserves in periods of trade reaction, and there is therefore this slack to be taken up in the periods of trade activity, in addition to that expansion of deposits beyond the customary ratio which usually precedes the outbreak of a panic.

The potent influence of the increase of the supply of gold, during the last fifteen years, may perhaps explain Professor Fisher's failure to give specific attention to transition changes in velocities and deposits independent of changes in the supply of money. This incompleteness in the analysis is found conspicuously in the chapter devoted to the historical verification of the quantity theory. As an account of price movements during the last five hundred years, the chapter is excellent. It is easy to show that pronounced price changes

of long duration have been accompanied by corresponding changes in the supply of money. But the chapter contains also a summary account of crises, which leaves the impression that all commercial crises have been preceded by an increase in the supply of money, either specie, government paper, or bank notes. But this is a conclusion to which there are notable exceptions, the most striking of which is the crisis of 1873 in the United States. Professor Fisher does not, indeed, regard it as an exception. He calls attention to the increase in bank note circulation between 1868 and 1873, from $295,000,000 to $341,000,000, but he does not mention that in these years an almost equivalent volume of 3 per cent certificates were retired. These certificates did not circulate as money, but as they were used for bank reserves, their retirement involved a contraction of the available money supply. Again, the increase in the stock of money which preceded the crisis of 1889 in France, of 1890 in England, and 1893 in the United States, was not one which in itself made possible an increase in prices, but rather the reverse. Such increase as occurred would seem to have been due entirely to the expansion of deposits relative to checks, and perhaps also to increasing velocities of circulation. The experience of those years furnishes an interesting instance of the possibility of temporary advances in prices, culminating in a crisis, during a period when the money supply was not increasing fast enough to prevent, in the long run, a lowering of the general price level. Finally it may be observed that during the last fifteen years, notwithstanding the increase in the stock of money, both England and France have escaped crises. After all, general economic conditions, the temper of the business community, and banking operations are fundamental factors to be considered in the analysis of crises, Monetary changes pure and simple would seem to be, at the most, a contributing but not the controlling factor.

The most interesting and valuable part of the book is most certainly chapter XII, which with its appendix makes up nearly a fourth of the volume. Professor Fisher here

in most brilliant fashion subjects the quantity theory to the test of exact measurement, taking for the purpose the period from 1896 to 1909. He finds a remarkably close approximation between the actual course of prices and that which was to have been expected from estimated changes in the factors of the equation of exchange. A similiar attempt was made by Professor Kemmerer, in his Money and Credit Instruments (1905). Thanks to the investigations of the National Monetary Commission and to some special investigations made for him by United States Treasury officials, the data made use of by Professor Fisher are much more complete than those available to his predecessor. In particular, he has had the advantage of the two investigations conducted by Professor Kinley regarding check and cash payments, one for the beginning of the period taken for analysis and the other for the end. With these advantages, coupled with numerous improvements in the methods of calculation, Professor Fisher reaches estimates which doubtless make a much nearer approach to actual facts than those resulting from the pioneer work of Professor Kemmerer. Moreover, Professor Fisher makes estimates which had not been attempted by his predecessor, for certain magnitudes in the equation of exchange. Professor Kemmerer estimated the volume of check transactions (M'V',) but did not offer any separate estimate of the volume of deposits subject to check (M') and the velocity of circulation of deposit currency (V').

The most important contribution made by Professor Fisher to the statistical verification of the quantity theory is in connection with the velocity of circulation of money. Here something more was required than the use of more comprehensive data and the refinement of existing methods. An entirely new method is worked out for estimating what has generally been thought beyond direct calculation. The analysis, once made, is simplicity itself, but on that account the immense debt of all students of monetary problems to Professor Fisher is in no wise lessened. The amount of exchanges, effected by means of money (MV), once

determined, the velocity of its circulation is arrived at simply by dividing that magnitude by the amount of money in circulation. But no previous writer has been able to devise a satisfactory method for estimating the amount of exchanges effected by means of money. This magnitude is, according to Professor Fisher, equal to the total money deposited in banks plus the total money wages paid out plus a small miscellaneous item. In countries where deposit banking is highly developed "money, like checks, circulates in general only once outside the banks; but when it passes through the hands of non-depositors (which practically means wageearners) it circulates once more, thus adding the value of wage payments to the volume of ordinary money circulation, which is equal to the flow of money through banks" (p. 287). A few items composing the small miscellaneous total of money which circulates more than twice may be mentioned to illustrate their obvious insignificance: - the till-paid expenditures of "commercial depositors" in excess of money withdrawn by them from banks; money receipts of non-commercial depositors pocketed instead of being deposited; money payments between commercila depositors," between "other depositors" and those between "non-depositors." The amount of money withdrawn from banks in 1909 is estimated by Professor Fisher as $20.7 billions, and the amount of money expended in wages at $13.1 billions, while the total of various miscellaneous items is estimated at only $1.3 billions. The relative amount of the latter figure is so small that even if the estimate is far from exact, it would make little difference in the total of money payments, and accordingly in the velocity of the circulation of money. The data for the calculation of the amount of money withdrawn from banks are based upon Professor Kinley's investigations for 1896 and 1909. For other years the figures are interpolated. The amount paid out to wage-earners and other nondepositors is estimated largely from Census and Bureau of Labor returns.

66

Finally, the volume of trade and the level of prices are

calculated from a variety of data which are brought together in the appendix, but, as no new principles or methods are involved, extended comment is unnecessary. Future investigators, using more adequate data, and further refinements of method, will no doubt arrive at more accurate estimates of these as well as the other magnitudes of the equation of exchange. Professor Fisher has laid a broad and deep foundation. A long step has been taken toward the reasonably exact measurement of price-making factors.

In the final chapter, Professor Fisher attacks the problem of "making purchasing power more stable." Among the remedies which have been proposed at various times in the past, he discovers merit only in the tabular standard. But the complications that would follow its partial adoption make it unsatisfactory, and there seems no way of securing its simultaneous adoption universally by all individuals, as well as by all countries. The remedy which he proposes would require no change whatever in the circulating medium or in the methods of making payments between individuals. International agreement alone would be needed. The proposal is a combination of the tabular standard with the gold exchange standard. Such an arrangement, it is argued, would be not unlike in principle that already set up in India, the Philippines, Mexico, and Panama. "As the system is now operated the coinage is manipulated to keep it at par with gold, that is, to follow the fluctuations of the gold standard wherever they may lead." Manipulate the par of exchange to keep pace with the tabular standard and stable prices will be secured. The steps necessary to be taken in order to establish the system are simple. First, the mints of the world must be closed to free coinage of gold, thus giving coined gold a value higher than its bullion value; second, an international statistical office would be necessary to compute index numbers in the ordinary way. Dividing the market price of gold by the index number would give an official price for gold. At this official price the government of some one country, agreed upon for the purpose, would either buy or sell gold at the option of the