apprehension lies in faulty comparisons, as well as in faulty original data. A few years ago a distinguished State senator, in a speech in the senate of his State, called attention to the decrease in marriages, the increase in illegitimate births, and the increase in divorces, and then he said that this alarming state of affairs had developed during the period in which the woman's rights question had been brought to the public attention; hence he argued that the woman's rights movement was responsible for the other conditions cited. This is what may be called concomitance in statistics with a vengeance. One might as well reason that because crime exists education is a failure, and, in fact, a writer a few years ago undertook to prove by statistics that crime, divorces, illegitimate births, everything that betokened bad conditions,- were more frequent and alarming in those portions of the country where the public school system prevailed than in those which had not established it on an expansive basis. His difficulty was in making comparisons between States where, on the one hand, statistics were fairly accurate, and where, on the other hand, they had few or none. And yet the statistical method was the very method by which his fallacious conclusions were completely upset, for this method proved that between the States which he used for his comparison there were no legitimate bases on which such comparison could be built. One State punished a crime by imprisonment in the penitentiary, while the other punished it by a fine of $20. One State had 150 or more distinct crimes in its penal code; the other a little over 100. Men were found in the prisons of one State sentenced for certain crimes, while no prisoners were found in the other sentenced for like crimes; hence the conclusion that one State was freer from crime than the other.

As a popular educational force, the statistical method is found faulty through the use of averages, for there are frequent fallacies in the practice of striking averages that add to the disturbing influences resulting from inaccurate enumerations, miscalculations, and unscientific comparisons. M. Quetelet, a distinguished writer on statistics, explained in his works the principles which ought to guide us in the matter of averages. To condense his illustration: He pointed out that an average may indicate two different things.

If

one measures Nelson's monument ten times, and each time with a slightly different result, and then adds the measurements together and divides the sum by ten, the quotient is an average or mean. So one may accurately measure the Duke of York's Pillar, the Parisian Obelisk, and the Column Vendôme, add the measurements together, divide the sum by three, and declare the quotient to be the average or mean height of the three monuments. It was contended by Quetelet, and very properly, that the results in the two instances are of such different significance as to require two separate names. He would limit the term "average" or "mean" to cases represented by the first illustration, the repeated measurement of one monument,- and the term "arithmetical mean" to cases represented by the second illustration - the measurement of several monuments. The repeated measurings of one monument result in a mean approximation to something actually existing, and this is an excellent definition of an average. Measurings and calculations having reference to a number of monuments result in no knowledge of anything existing; they simply and only indicate a relation among things actually existing. The principles of this statement are lost sight of oftentimes in popular calculations relative to wages. A concrete arbitrary illustration will suffice. There are men employed in a certain works at $1 per day, others at $2 per day, and still others at $3 per day. It is a common practice, in trying to find the average rate of wages paid in the works, to add the per diems $1, $2, and $3, making $6 and three rates of pay. Dividing the $6 by three gives an average of $2. Now this is not a true average. It is what Quetelet would have called an arithmetical mean. The true average is to be found by taking the number of men at each rate, say, for instance, 20 at $1 per day, 40 at $2 per day, and 60 at $3 per day. This means that 120 men earned $280 in one day. The true average is to be ascertained by dividing 280 by 120, which gives the average day's wages in the works assumed as $2.33 instead of $2, on the basis of the arithmetical mean. There have been many erroneous conclusions drawn through lack of recognition of this simple principle; even in so able a work as Thorold Rogers's "Six Centuries of Work and Wages," there is a very great lack of a true comprehension of the

principles of the average and of percentages also.

as

Perhaps the most deceptive statistics with which the public deals are those relating to the products of manufactures. Up to the present time there has been no method invented by which a fairly truthful statement of the exact value of the manufactures of a country at any given time can be determined. The United States is the only country so far that has taken an industrial census of her manufacturing industries. The results, stated, usually relate to the capital invested, the number of persons employed, the cost of materials used, the total wages paid, and the value of the product at the works. These results are given for the census periods. The true amount of capital invested has not been secured at any Federal census, owing to the omission of credit capital and the differences in opinion as to what constituted capital as applied to manufactures. In 1890 the attempt was made, through securing the amount of interest paid, to ascertain the credit capital employed. Money expended in purchasing raw materials has been stated with reasonable accuracy at each census, and so has the total amount of wages paid. The number of persons employed has been given on various bases, but usually the average number employed during the year. A truer method would be to ascertain the total number of positions paid for during the year, but this has not as yet been done. The total value of the products consists of the value of the raw material or stock entering into manufactures, the total wages paid, all the expense of production, and the margin of profit. The value is that given by the producer. The results for all the establishments of the country are then aggregated and the total production found to be worth so many millions of dollars.

Assuming that for each decade the classification has been practically similar, the total value of the products offers a reasonable basis for comparison for different periods. As a matter of science, the amount is absolutely worthless, and, as already intimated, no method has yet been devised by which the real, true value of the manufactured product can be ascertained. This results from the fact that the completed product of one manufac turer becomes the raw material of another, and this process may be extended through

two or more gradations. Again, the product of one manufacturer may have been secured in a prior year and used in the census year as the material or part of the material from which completed manufactures are made, all entering into the total value of products. To deduct the raw material or stock used from the total value and say that the difference represents the added value by labor is not a sound method. By these statistics the public is educated only in one sense, and that is the relation of total products one year to total products another year; but even this comparison becomes vicious when price is considered, for a hundred tons of pig iron may be worth $1,000 at the time of one census and $2,000 at another census. In cases where the unit alone can be brought into comparison, the quantity will reveal the truth as to whether manufactures are increasing or decreasing; but quantities of various kinds cannot be added together, and hence values must be taken when dealing with the whole volume of manufactures. Statisticians have for many years recognized these difficulties and called attention to them, but so far the complications have been too great to secure a scientific application of the statistical method.

The statistical method is not thoroughly satisfactory when the psychological side of a truth is sought. Data may be accurate, but the underlying motives which result in actions may not be revealed; hence the difficulty of applying the method for the purpose of ascertaining the increase or decrease of crime or the underlying causes of strikes. The apparent or superficial causes of strikes, and the apparent or superficial side of criminal conditions, can be ascertained by the statistical method. Sometimes, by a classification of causes and motives, or of the number of cases out of a certain total in which certain motives are ascertainable, this psychological side can be shown, but it is rather a dangerous matter to attempt to prove inherent and psychological conditions by statistical data.

The popular idea of statistics assumes too much. It should be remembered that numerous accounts or records of continuous events are essential to establish accurate knowledge. Herein statistical science differs from the exact sciences. One experiment may establish the fact that water freezes at a certain point, or

that the intermingling of two chemical elements will produce certain results, and the conclusion is that the same results will always be secured when the same elements are brought in contact; but the phenomena of life and of social and economic conditions are not so easily ascertained. These must be secured through a succession of accounts and should cover wide fields. The public expects much, often too much, from the statistician, and the public is not at fault in this expectation, because the free use of statistical tables has led the public to believe that all truth can be ascertained by the statistical method. Hence, the results of investigations, census-taking, and various other means of securing statistical data are often disappointing; yet up to a certain stage the results, if properly reached, are exceedingly useful in educating the people. The very usefulness of statistics in securing publicity of accounts, one of the surest means of correcting abuses, demonstrates the educational force of the statistical method. In no other way can the public ascertain the condition of railroads, of municipal works, the integrity of private corporations, the volume of foreign and domestic trade or of the products of our farms and workshops. All such statements stimulate the inquiry and help every man to solve some of the questions that present themselves.

The expansion of the method to such an extent that the real condition of corporate existence can be known, would help to correct many of the abuses of which the public complain. Such benefits are being reached gradually through railroad commissions, insurance offices connected with governments, the statistics of corporations generally, and other instrumentalities which serve to en

lighten the public as to what is going on in its midst. So everywhere the statistical method helps one to understand the true conditions of his own and other countries. Without it the world would grow larger instead of smaller; by it all parts are brought into a coördinated existence, and the commonest man is benefited. It must be the duty, therefore, of every man, whether producer or consumer, whether public or private citizen, whether employer or employee, to furnish, whenever the opportunity occurs, all facts which relate to conditions. It is only by a patriotic understanding of the

educational force of the statistical method that correct statistics can be secured. If people give false statements to official investigators, false conclusions must follow. If we wish to learn the true conditions of labor, both employer and employee must give exact facts when called upon. It is only by this patriotic comprehension of the true value of statistics that education in the broad sense receives a new element of force. Social science; political economy, its most important branch; political science,- almost every department of human knowledge,—can be aided and strengthened by the proper use of the statistical method.

On the whole, enormous as have been errors, false as have been many of the statistical statements of official reports, inaccurate as have been many of the calculations, and fallacious and almost monstrous as have been many of the inferences, political economy has, nevertheless, profited by what has been accomplished. Errors are gradually disappearing, and a very considerable remainder of truth is found.* We know far more than did our fathers of the progress of population, the resources of the nation, the earnings of the people, the cost of living, and the efficiency of labor, and far more of criminal conditions, of mortality in town and country, of vagrancy and pauperism, of crowding, of immigration, and, in fact, of all the conditions of life which make up sociology. Legislators and philanthropists can ill spare their statistical guides, lame and delusive though they may have been at times. CARROLL D. WRIGHT.

SPANISH literature - imaginative, poetic, and chivalrous has never had its due meed of tribute, partly from the lack of a good manual in English. Ticknor's "History" is too ponderous, and is now old. A good, popular textbook of Spanish Literature, by J. FitzmauriceKelly, has just been brought out by D. Appleton & Co., New York (price $1.50), which furnishes a compendious account of Spanish letters for the student, corrects the errors into which Ticknor fell, and brings the literature down to recent times. Though brief and succinct, the work is interestingly written, and with full knowledge of the subject. While sympathetic, and even enthusiastic, over such writers as Cervantes, Calderon, and others, the author rarely loses his impartiality of judgment, and hence is to be trusted.

Paraphrase from Sargant's "Lies of Statistics."

I

N HIS "Word About America" Matthew Arnold quoted the late Mr. James Russell Lowell's remark to the effect that "the English make their Anglicism the standard of all things," and categorically asserts "I do not share it." Without waiting to inquire either into the accuracy of Lowell's general statement or into its particular signification, we may say that Arnold unfortunately does share it. This critic who was so fond of holding up to us continental models was himself anything but cosmopolitan. He was perpetually detecting in his fellow-citizens notes of provinciality, and yet was himself provincial. His avowed aim was to "change English ideas," and he was himself affected with "Anglicism." This, I know, is a grave accusation to bring against the one of all England's critics who has generally been regarded as particularly free from such faults; who has been lauded for his catholicity, his flexibility, his openness of mind. Accordingly, I take pains to support it to the full.

We have now got all that Matthew Arnold wrote. His prose works and his letters occupy some dozen volumes of varying thicknesses; his poetry two thin ones. Yet there are those who venture to say that the verdict of posterity will be analogous to, and yet the exact contrary of, what he himself asserted would be the verdict of posterity upon Shelley: Shelley, he thought, would live by his prose rather than by his poetry; Arnold, we think, will live by his poetry rather than by his prose,- will live, that is, as he thought Shelley would live, by that upon which he set least store, by his play-work, not by his life-work. Years hence, probably, all that the average man will know of Matthew Arnold will be that he was a great, perhaps one of the greatest, of English critics; that he was an inspector of schools and an efficient one; that he did much and strove more to disseminate a liking for the things which are more excellent efforts that earned for him the title of apostle of culture; that he extolled sweetness and light; that he derided Philistines: and that he wrote some books about criticism and culture, God and the Bible. But that same average man,

**Civilization in the United States," pp. 74, 75 (Ed. Boston, 1888).

probably, will have on his shelves two little books, uniformly bound in claretcolored cloth and labelled "Poems," which he will often take down, and peruse lines upon lines from "Rugby Chapel," from "Sohrab and Rustum" and "Tristram and Iseult" and "Balder Dead," from "The Strayed Reveller" and "The Forsaken Merman" and "Empedocles on Etna» and "Geist's Grave" and the Sonnetsfrom, perhaps, one and all.

However, definitively to say what contemporary poetry will live and why, is to no living man given. Who has yet determined what are the attributes of a work of art tending to render it permanently popular? Which of its two components— truth of matter and perfection of form — contributes more to its ultimate success? Or is it only that work in which these elements are both supreme, and both in indissoluble and harmonious connection, that will live? What, after all, is truth of matter, and what beauty of form? Even if to-day's truth can never quite be to-morrow's lie, may not to-day's beauty quite easily be to-morrow's deformity? Truth remains, but the fashions in representing truth change. Who would now read an epic on Creation in twelve books of pentameter iambics, even if written on the most approved of evolutionary hypotheses, Lockyer's meteoritic hypothesis and Weismann's theory of the germ-plasm not omitted? Not, perhaps, till we have quite fathomed the depth of Keats's deep remark,

"Beauty is truth, truth beauty; that is all We know on earth, and all we need to know," shall we be able to apply the touchstone of immortality to contemporary poetry.

But that not all of Matthew Arnold will die is surely certain. He exercised a certain influence on his age- not, perhaps, on the masses, those masses which, nevertheless, he strove so hard to reach,- but on those elect few through whom ultimately the masses are reached; and anyone who exercises an influence on his age is pretty sure to command respect and excite curiosity after that age has passed. Not necessarily by means of those identical works that influenced that age. It was by his religious and political treatises that in his time Milton was chiefly known. To-day

to most he is but the author of the "Paradise Lost." So with Arnold: he was in his day most influential as a refiner of intellectual tastes - if we may so term the net result of his crusade on behalf of culture. But if his attainments had ceased there, it is doubtful if he would have been so much in the public eye as to-day he is. Most certainly his attainments did not cease there. Arnold was one of the few men from the perusal of whose total writings one rises with an exalted opinion of the writer. That, I think, is a good, and at the same time a severe, test. Well, Matthew Arnold stands it. And if we are to look for an explanation of the fact that we see in his life and writings a greater man than his lifework seems to warrant, we shall find it, I think, in his poems. However, I am concerned here not so much to try to point out that his poems will live as that his prose will die.

His prose will 'die first, because it is not really concerned with things of vital and universal importance. Compared with the Carlyle, his praise of whom he so largely qualified, or even with the Emerson, whom in Boston itself he so much belittled, he had little grasp of the mystery of life, and few are the passages in his works which throw light on that mystery. There is a gospel in Carlyle; you may find a gospel in Emerson; in Matthew Arnold it is not easy to find one, strive as he himself did in the search. And whether or not you can in a man find a gospel,find, that is, a rule of life, an explanation of Humanity and of the relation of Humanity to its Maker,- that, I take it, is one of the foremost tests of his greatness. For does it not prove that he has grappled with the problem of Life, and grappled with it with so much more earnestness and seriousness than other men, that he has been able to find for himself and them some, however small, solution? And the deeper and wider the solution, the greater the man.

Could any one have said to Matthew Arnold what Carlyle said to Goethe:-"If I have been delivered from darkness into any measure of light, if I know aught of myself and of my duties and destination, it is to the study of your writings more than to any other circumstance that I owe this.

MATTHEW ARNOLD

no one can deny. Least of all is Arnold's assertion that "he is not a great writer » a denial of it, because there he is concerned more with the word "writer" than with the word "great." One passage only in all Matthew Arnold's prose writings can I recall which possesses this luminous property. It is this:

"To walk staunchly by the best light one has, to be strict and sincere with oneself, not to be of the number of those who say and do not, to be in earnest - this is the discipline by which alone man is enabled to rescue his life from thraldom to the passing moment and to his bodily senses, to ennoble it, and to make it eternal. » +

*Correspondence between Goethe and Carlyle," p. 7 (London: Longmans, 1887).

+" Culture and Anarchy." Preface, p. xlviii.