the beginning of the recent changes in their relative value. We may be unable to devise, or even certainly to define, a stable monetary standard in any absolute sense, butwe can determine whether the one or the other metal shows a nearer approach to our indefinite ideal, and even show more or less satisfactorily why. STABILITY AS SHOWN BY STATISTICS. The three diagrams which present this comparison to the eye are based on data from the Statistical Abstract of the Treasury Department, for the one at the top, and from the Report on Wholesale Prices, Wages and Transportation prepared for the Senate Finance Committee in 1892, by Hon. C. D. Wright and other officers of government, and collated by Prof. Roland P. Falkner, for the two below it. The two upper diagrams both relate to prices in the United States, the lowest one to England. In all of them TABLE OF RELATIVE PRICES AND WAGES; GOLD AND SILVER VARYING 1853 1854 1855 1856 1857 1859 1859 1860 1861 91.5 100.6 99.4 75.6 77 6 84.5 82.8 87.2 92.9 100.0 1862 101.8 100.2 99.8 96.7 105.7109.7 1863 1864 98.0 93.4 1867 1868 1863 1870 100 6 99.4 101 8 105.6 10.9 100.6 99.4 108.3 102.2 100.4 99.6 92.1 91.5 82.2 100 3 99.7 92.1 109.6136.4 100.3 99.7 106.9 111.0] 86 5 100.3 99 7 91.7 112 6 145.2 109.0 95.5 1865 100.5 99.5 100.6 107.9 69.0 100.599.5 95.7106 3136.2| 96.4 1866 98.8101.2121.2 124.5 100.5 99.5 124.7 131.7 111.7 100.4 99.6 99.2106.9133.7 116.1 107.1 99.1 100.9 15.3 105.2 101.0 99.0 123.0 121.4 123,0 100.9 99.1 104.5 104 8 116 5 108.1 99.4 100.6111.6 102.2 101.4 98.6 120.3 127.4 120.8 101.0 99.0 103.7 104.1 100.7 99.4 100.5 117.3 108.9 101.1 58.9 121.4 113.5 124.9 101.0 99.0 98.7 102.9105.2 125.4 111.8 99.4 100.6119.7 112.9 101.3 93.7 128.4 120.5 138.7 100 9 99.1 97.5 101.2106.9 128.4 115.5 1871 99.4 100.6 114.4 106.9 101.3 98 7 154.9 124.5 152 3 100.9 99.1 102.7 103 5102.11 98.0 1872 99.3 100.7 97.3 99.9 101.0 99 0 123.4 122 6 154.7 101.1 98.9 107.1 113.7109 9 145 0123.7 1873 99.9 100.2 98.3 99.9 101.9 98.2 117.4 113.7 150.2 102.0 98.0 114.9 118 9 116.8 153 8129.9 1874 100.8 99.1 104 4 100.2 102.9 97.2 121.4 120.0 150.1 102.9 97.2111.9110.1118.4 159.0134.8 1875 101.9 98.1 113.3 104 4 103 9 96 2 120.5 119.1 145.9 104.2 96.0 109.5 104.5 108.0 163 5 139.8 1876 104.3 95.9 114.7 102.6 105.8 94.5 115.4 115.0 142.0 108 2 92.5 113.6105.5 108 3 164.8 12.2 1877 105.4 94.9 110.9 101.5 104.0 96 2 117.8 111.3 140.8 106.1 94.3 112 2 103.3 105 6 154.2 14.5 1873 105 095 2 100 9 92.5 107.2 93.3 113.1 110.6 149.0 108.3 92.3 109.5 98.8 100.9 157.4 152.6 1879 108.5 92.2 83.2 83.2 111.4 89.8 108.7 105.8 155 3 109.7 91.1 103.0 95.1 89 9 161 0 52.0 1880 107 0 93.5 95.3 90.7 108.8 91.9 117 1 114.1 155.6 108.7 92.0 106.8 99.8 100.3 168.4 150 9 1831 107.2 93.3104.0 97 8 103.9 91.0 121.9 119.1 165 6 109.0 91.7 103.7 96.5 94.6 171 81:9.2 1852 107 4 93.1 118.1 104 8 109 3 91.5 129.8 119 2 167.1 109 1 91 7103.4 96.0 96 6 173.7 150.1 1883 108.2 92 4 122.2 106 4 111.1 90.0 132.0 118.4 176.9 110.4 90 6104.0 94.9 91.0 181.3 152.5 1884 108.4 92.2 108.6 97.0 110.4 90.6 120.2 113.3 171.3 110.2 90.7 92.9 87.4 91.1 1885 109.5 91.3 99.2 90.1 111.9 89.4 110.4 104.4 174.5 112.7 88 7 88.9 85 0 86.5 188 112.7 88.7 90.7 81.3 115.2 86.8 114.6 107.6 179.5 116.6 85 8 89.4 84.4 84.4 1887 115 4 86 7 92.9 86.9 115.4 86.7 120.2 109.1 180.7 117.6 85 0 84.7 83.1 86.9 1888 117.0 855 99.9 90.9 118.3 84.5 129.4 113.8 186.8 119.9 83.4 91.8 88.6 94.6 188) 118.5 81 4 101.8 91 7 120.8 82.8 135.2 119.0 196 8 120.2 83.2 95.5 92.2 96.2 1890 116 2 86.0 91 8 85.6 118 3 84.5 123.7 110.8 199.0 113 7: 88.0 88.0 86.4 97.4 1891 112.1 89 2 90.7 83.4 114 3 87.5 118.8 107.9 192.7 117.0 85 5 94.6 88.2 92.6) 1892 118 2 81.6 97.9 86 8 120 5 83 0*120.9*109.6 +196.1 124.5 80.3 1893 124.8 80.1 111.7 95.7 127.2 78.6 §121.6 §116.0 §210.0 131.6 76.0 January prices: factor in 6th col. + July, 1891; gold 116.6. October, 1891; gold 118.1. § October, 1892; gold 126.5. 94.0 97.6 155.4 162.9 the par value is made a mean proportional between silver and gold, so that the two metals vary reciprocally with reference to it; they are denoted by crescents and circles respectively. The full line denotes, in the middle diagram, Professor Falkner's means of all the two hundred and twenty-three articles of the Senate report, weighted according to consumption, while the dotted line is the indiscriminate mean of the fifty-three articles included as food" in the same list. Corresponding to these, the upper full line is a mean, here unweighted, of nineteen articles in general use-sixteen of the prices are export rates, and three others (Ohio weol, soft coal and pig iron) as reported by dealers in the Treasury Abstract. The dotted line is a mean of eleven export prices of foods. The full line in the lowest diagram is the general mean from Sauerbeck's list of articles: this is preferred because fuller and in slightly better correspondence with the United States figures than the Economist list, which is also added in the broken-anddotted line, for the most part closely agreeing with it. The lowest line of dots only is a mean of the three columns headed Vegetable foods," Animal foods" and "Tea, Coffee and Sugar" (the last having half weight) in the summary of Sauerbeck's relative prices. Broken lines in the middle diagram show the movement of average wages, Professor Falkner having weighted the means for seventeen out of twenty-one different industries, proportionately to the number engaged in them according to the Census. All prices in the Senate report are compared with those for 1860, and there seems to be no good reason for changing this. As the Treasury Abstract prices go no farther back than 1866, however, a different date had to be chosen1 with them, and that fixed upon was the mean of the three 1872-4 In the Senate report, prices are given in gold as well as currency. The Treasury prices for 1866 to '78 are here reduced to gold by dividing by the mean gold premium for each fiscal year. The ratios having been thus put on a gold basis, they were reduced to the chosen mean-proportional basis by first comparing the ratio of gold to silver for each fiscal year (upper diagram), each January (middle diagram), and each calendar year (lower diagram). with the values, 15.78 (for mean of three years, July, 1871, to June, 1874), 15.19 (for January, 1860), and 15.29 (for whole year 1860), respectively; and then multiplying each gold price by the square root of the corresponding ratio of ratios so derived It may, perhaps, be worth while to add that the reason why neither gold nor silver was chosen as a standard of value, but a geometric mean of the two, so that the rise of value in one of them is exactly offset by a proportional fall in the other, is not because of any notion that such a mean has any practical significance whatever, or ever could have, or that it is desirable for the purposes of any other statistical inquiry than this; but simply because in deciding between two alternatives, those alternatives ought to be treated as impartially as possible, without discrimination of Trojan and Tyrian. Whether the gold or the silver standard is more nearly followed, is in these diagrams not a question between curve and straight line, but between two curves of opposite course, The use of the upper and lower diagrams is mainly to confirm the middle one. which is best adapted for the proposed test, because founded on more widely derived and better digested data. It will at once be seen that the two upper curves of price have nearly the same relation to each other and the same general direction as the corresponding middle curves, though taken from an independent source, made up of far from identical (though similar) elements, and very differently treated. Both show a lower height, indicating a lower relative price, for the foods than for the other items in 1866, while after 1875 the foods through all vicissitudes are decidedly and continually higher. Notwithstanding this divergence, the direction of both, after 1866, is more nearly that of the silver price than of the gold; following the former even in its details until 1880, and after that becoming somewhat more nearly parallel to the latter. But it makes a significant difference whether 1872-4 or 1860 is taken as our date of reference. By choosing the former, as in the upper diagram, prices are made to follow silver very closely indeed; whereas the 1860 date brings the prices of the last fifteen years to substantial uniformity compared with gold. The price urves for England, at the bottom, agree in showing a general diminution after 1873, and no such diminution, but on the whole an increase, before that date. Allowing for the violent changes of our civil war. in market prices as in many other things, almost the same words might be used to characterize the United States curve in the middle diagram: only putting the change of direction at about 1870 or 1871, or perhaps even as far back as 1866, instead of 1873. The relation in price between food-stuffs and commodities in general, is substantially the same in the Sauerbeck list as in the Treasury Abstract list and the fuller list of the Senate report; the foods generally lower for the former half, and uniformly higher for the latter half of the period considered. The divergence of the two from 1862 to 1866 is abnormally large, and is fully explained 1 A disenssion of these prices by Mr. E. B. Howell, in the Review of Reviews for October, 1893, adopts 1873 as the standard. by the effect of the single item, cotton, whose price was more than four times greater in 1864 than in 1860. The Economist list is generally in satisfactory agreement with Sauerbeck's, but its abnormal elevation from January, 1862, to January, 1867, is even more excessive, cotton being allowed a much greater proportional weight. The divergence after 1886 is more troublesome to explain, but it does not seriously affect the question before us. The effort to get hold of some indication as to the movement of wages in England, within the time here covered, has met with no success; the only statistics to be had are scattered, fragmentary, desultory, showing no approach to a continuous series for any occupation. Hon. Carroll D. Wright, in calling attention to a recent collection of statistical notes on the subject by Mr. Giffen, said that it would take some months' labor of a considerable clerical force to obtain from them any results that could be used as indications of progress. The wage statistics here depended on to show that the movement in this country is not exceptional, are some scanty and limited returns of coal and iron-ore production in France, and 'rom the cotton spinning and weaving industries in the German Rhine Provinces. The average annual wages per laborer, as tabulated in the first report of our commissioner of labor (on Industrial Depressions), appear to have been found by dividing the total wage-payment by the mean number of laborers; and in combining them, care was taken to allow weights proportional to the number employed; dividend and divisor were added separately, for a new quotient. No combination of French and German figures, in mining and cotton working, was attempted. The results were compared with 1860 for gold-price ratios, then reduced to a mean proportional between gold and silver, as already explained. They are thus exhibited on the middle diagram, being distinguished from United States ratios by the letters F and G. It will be seen that the course of European wages, so far as we can infer it from these very scanty data, corresponds fairly well with that in this country, shown by the heavy broken line. The general upward course meets a few interruptions in each line; these, in the French and German cases, may very possibly be incidental to the special industries examined. The first exception to the general rule of increasing compensation in this country occurs during the Civil war. Wages nominally increased in those days, every year; but the currency in which they were reckoned lost in value much faster than they increased, so that real compensation fell off as shown. Wages, as is well known, change more slowly than prices; so that, as already indicated, a depreciation in the unit of values makes those who depend on them the worst sufferers. The price-lines, it will be seen, do not show the effect of depreciated currency during the period nearly so much as the wagelines. The second exception marks the great industrial depression of 1873-78; the third the minor depression of 1884; and the fourth, the one which began in 1890 and still continues. In the chart, there appears to be a sudden depression for 1891; followed by a rapid increase, but that appearance is due to coincident changes in the price of silver. Following the gold standard, and taking account of relative importance, wages rose suddenly in 1889, but became substantially stationary in 1890 and 1891, and showed a very slight falling off in 1892. The figures for 1894, if they could be had, would doubtless show a grievous downfall. INTERPRETATION. Now, in answer to the question, which of the two metals furnishes the better mone tary standard, the diagrams have a great deal of important information to give us. First, as to the general run of prices of commodities since 1873, when the change in the gold-to-silver ratio first became considerable. The upper diagram shows that prices have followed silver and not gold. The middle diagram likewise, for it is by direction that we must judge, not position. The lowest diagram shows the same thing for England, with even greater emphasis, general prices having there fallen relatively even to silver; though this is less marked with the Economist than with the Sauerbeck figures. There is a difference between foods and products in general, the former having had less share in the general downward tendency. This is a consequence of the fact that recent mechanical improvements, bringing greater economies of production and transportation, have affected foods less than other commodities If the use of a measure of values is to preserve un formity in general prices, under such conditions as have obtained since 1873, then unquestionably silver is a better measure than gold. Next, carrying the limit back to 1860, we find a noteworthy difference in the answer. As a representative of prices about 1860, as the middle di gram shows us, gold has for the last fifteen years answered very well indeed in the United States. It has the same purchasing power that it had then, far more nearly than silver has. But, strangely enough, the same rule does not hold for England. There, prices still follow silver, and keep strikingly close to it if we consider how careful that country has so long been to accept only gold as its measure of values. What is the reason of this difference? A closer inspection of the price curves appears to confine it within a few years. There is no very important and lasting divergence between our curve and the British, down to 1862, nor after 1866; for such fluctuations as appear in either disappear again after a few years. But between those two dates there is shown a heavy increase in prices in this country, British prices for 1866 remaining almost exactly what they were for 1862; which increase has not yet been offset by a relative decrease. If the U. S. prices for 1866 had been held down to equality with 1862, and had followed afterward the same course they actually followed, they would have been very like the British, and have kept closely to the silver line. The explanation of this increased scale of prices in those special years is found in our revenue legislation, which then underwent its most significant modifications. If we regard the movements of prices in England, where there have been no overhaulings of the tariff, as the normal movement, our conclusion that silver has been better adapted than gold to exhibit uniformity in prices of commodities, remains undisturbed by the further extension of the period considered. But a constancy over long periods of time, it has been already shown, is less important in a monetary standard than a steadiness in opposition to shorter and sharper fluctuations. There are several cases, within the time here covered, of sudden changes in the relative price of gold and silver; these are represented on the chart by steeper slopes of the gold and silver curves, which mutually approach or recede. Consider first the years 1875-80, where gold shows two rises and two falls in value, and silver of course the reciprocal changes. Prices, as shown in the middle diagram, follow silver with a fidelity, which in the case of foods is quite remarkable. The upper diagram shows the same thing, only the changes in the commodities have a greater range. English prices are less decisive for this period, since they seem to follow gold about 1876 and silver about 1879. Consider next the years after 1888, where the changes are of still greater range and abruptness. Here the price curves follow gold in all three diagrams, thus appearing to show that these important changes affected only silver and not commodities in general; or, in other words, that it was not gold that altered in value. It may be claimed that this following of gold is merely an evidence of the conservatism of prices, taken in connection with the fact that our business is done on a gold basis; so that, if we had regulated our prices by silver, they would have been comparatively uniform in that metal, while gold would have appeared to be the wanderer. Professor Laughlin insists that this very question has been recently tested in Santo Domingo, where the currency was silver, the unit a Mexican silver dollar, and yet prices fluctuated with regard to it, whenever its value fluctuated with regard to gold. But this may be taken as proof, not that gold was inherently the steadier metal, but that markets in Santo Domingo, a region of only 400,000 population, are too strongly influenced by prices in this country to show what they would be independently. India, a country great enough to have a market of its own, does not show the effects noted by Professor Laughlin. It may reasonably be admitted, therefore, that the claim has something in it, and that the conservatism of prices, already mentioned, after all forms a far more important factor in resisting these most dangerous changes than the choice of the better adapted metal. Another very significant point should be borne in mind, in the same connection. The fluctuations of prices which run their course in short periods of time, the very fluctuations against which it most concerns us to guard if possible, were even greater before the two metals so widely parted company in 1873-not only in this disturbed country, but in the comparative quiet of Great Britain-than they have since been by either metal. The food-curves then showed very little more irregularity than those of general prices, and the great upward movement which followed the California and Australia gold discoveries, the reaction attending the commercial depression of 1857, the rise in cotton due to our civil war, and the effects of our increased tariffs, all occurred while the gold and silver ratio continued nearly uniform. Were it not for the existence and importance of this fact, it would hardiy have been worth while to continue these diagrams back of 1873. But the proof that we here have, that the greatest oscillations in price come from causes with which divergence in metallic standards has nothing to do, ought not to be overlooked when we are estimating the importance of inquiries into metallic standards. To the principal irregularities of the heavy broken line, showing the mean wages of labor, regard being had to the number engaged in the different occupations, attention has already been called. Passing over these, which are chiefly downward turns, symptomatic of our greater and lesser industrial depressions, and comparing the curve with those for the two metals since 1873, we see not a single instance in which it shows a change in direction agreeing with that of the silver curve for the same date; while its correspondence with the gold curve, even in its minute changes, is as striking as is its general agreement in direction, whether we start from 1873 or from a remoter date. We find the upward turn in 1877 and the curious alternation of higher odd years atd lower even years between 1878 and 1885, in the gold curve, faithfully mirrored in the wage-curve; as are the downward turns of 1886, 1889, 1890 and the upward turns of 1887 and 1891. If this correspondence be ascribed to the conservatism already spoken |