common-that they seek to explain by some one simple formula the whole complicated development of human society. Most of them are, indeed, the work of professed "sociologists"; and they might be thought to exhibit, in this respect, the distinguishing trait of the branch of literature to which they belong. For, alike in Europe and America, quot homines, tot sociologiae. But in the pursuit of a Key to Society, there must, to judge from the number of publications, be something peculiarly congenial to the American temper. I suspect the explanation is to be found in the mechanical genius of the nation. The typical American is the man who, while Irish or Italian labourers are drudging away before him, sits on a fence and whittles a stick while he thinks out some easier way of getting the work done. And in like manner it might be said in a figure that America is full of "thinkers" who are sitting on fences, academic or other, and excogitating short cuts to the centre of the social labyrinth. It is a state of mind which has been productive of great things in the field of mechanics; it may still be fruitful in the study of society. For a too intimate knowledge of the difficulties of the problem does certainly put a drag on speculation; and there is much to be hoped for from the courage which is not afraid of large problems. Yet hitherto, it must be confessed, it is the dangers of such a mental attitude which have been most often apparent. The keenest powers of insight cannot safely invent the forms either of past or of present societies without sufficient information; and when, as in the present case, we have inadequate knowledge, a narrow ethics, a wearisome style, a barbarous terminology, and the constant confusion of assertion for proof, the result cannot but be disheartening, in spite of "the excellent earnestness," and even the occasional suggestiveness which the book undoubtedly displays.

Professor Patten's treatise sets before us what in one place he calls "a theory of history," in another "a social psychology," which unlocks, in his opinion, all the mysteries of social evolution-economic, political, religious, intellectual-from the beginnings of recorded time to the last days of man on the earth; though he modestly limits his more detailed exposition to the period from the "Early Germans" to our own time, and chooses England as his special field of illustration on the ground that "English thought" has been "more normal and. more uniform than that of her continental neighbours." This theory he regards as "scarcely open to question," and he does not hesitate to end with the assertion that it harmonises natural and revealed religion. But after all, it is nothing more than a little verbal legerdemain with our old friends "adjustment" and "environment," eked out with "sensory " and "motor." The "environment," first "determines" sensory ideas;" then the "struggle for existence" "determines" the "motor reactions" (or "motor ideas") which "constitute character." Progress is due to "a repeated transition from one environment to another; "but the "adjustment" to a new environment takes time

and is seldom complete; and "the striking features of every progressive nation" are due to the collision between the national character, due to the past environment, and the exigencies of the new situation.

Now, in the first place, this is a curiously simple and naïve psychology. The leap from sensory and motor nerves to sensory and motor ideas is one which no physiologist would venture to make, and which is made, so far as I know, by no recent psychologist of any reputation. And in the second place the effect of environment on character is habitually conceived of in the most mechanical and fatalistic manner. Yet even if the position were free from theoretic difficulties, the question would still remain: how comes the environment to change? The truth is that when you know the facts Professor Patten will supply you with a set of abstract-sounding formula to describe what you find to have happened; but he does not really help you to discover the facts; and the effect of the forces actually visible may be more concretely and intelligibly stated, and has again and again been so stated, in the ordinary literary language of the commonplace historian.

The "theory" is followed by a number of chapters which profess to show its applicability to the development of English character. They present an extraordinary jumble of shrewd observation and absolutely baseless and ridiculous assertions. Even when the remark is a sensible and true one, Professor Patten does his best to irritate the reader by insisting on translating it into his pseudo-scientific and barbarous jargon. Here is an example. "Methodism," he observes, "made people conscious of their emotional nature and educated them in its manifestations." This is true enough, though put somewhat too absolutely. He then proceeds: "Every one, it taught, must . . . recognise and manifest certain well-known motor feelings. A mere belief, or a sensory knowledge, that created no measurable motor reactions, was not enough," &c. "Motor reaction" has an impressive physiological ring, but a moment's thought will show that it does not penetrate one inch further into the mystery of the human heart than the old literary word "emotion."

Only too often, however, Professor Patten not only interprets his facts but invents them. Of general mediæval conditions he appears profoundly ignorant, while some vague impressions of certain facts of English history between the years 1300 and 1800 A.D., seem to float about in his mind unfettered by limitations of time, and to be grouped in one phantasmagoria after another as suits his argument. A few instances will suffice. "The great plagues were coincident with the rise of Puritanism. Labour had grown scarce and wages high;" and hereon a whole fabric of conclusion (p. 138). But the assertion is quite inconsistent with any merely human chronology. The bath-tub, according to Professor Patten, has been one of the greatest of moralising agencies. "It is hard to over-estimate its importance." "Much of the sensuality of earlier days was due . . . to an overplus of energy

and warmth of blood. A cold bath remedied this. . . . . The bath-tub is the parent of that English optimism of which the last two centuries have seen so many examples." One hardly knows which to be most surprised at; the child-like simplicity of the notion itself, or the fetterless imagination which gives the bath tub a couple of centuries for its beneficent work. "The new fashion of tea-drinking came in because hot drinks were grateful after the blood had been cooled off by a bath." For "tea" we should perhaps read " coffee," if it is breakfast that is in question; in any case it would be hard to justify the alleged cause chronologically; and Professor Patten does not see that if his remark is true, the warming-up by tea would counteract the moralising effects of the cooling-off by the bath! It is the drawback to all arbitrary constructions of history that it is equally easy to use the same fact, or supposed fact, in opposite ways. Our author might just as well have explained the vices of the age of Fox by the pernicious warming-up effects of hot coffee.

Perhaps the "fond thing " most "vainly invented" is the account given of the improvement in the condition of the agricultural labourer in the eighteenth century. He now began to enjoy, says Professor Patten, "new drinks, like tea and ale" (p. 195). The association of tea with ale, as if their introduction to the labourer's diet belonged to the same period, is enough to make Cobbett turn in his grave. To the same period, says our author, belongs the introduction of the oven into the cottage, and with it a new " oven diet." "So long as food had been cooked in a pot over an open fire, only watery dishes could be made, but ovens made dry foods possible. Bread and meat could now be cooked in more palatable forms than formerly," &c., &c. Has Professor Patten ever been inside an old-fashioned English cottage? When there is an oven, does he know where it is placed, and how it is made, and what it is used for? Has he ever heard of the frying-pan? Does he know what a turn-spit is?

I have said nothing of the elaborate estimate of the chief English philosophers and economists which occupies a large part of the volume. It is highly original; Professor Patten does not hesitate to correct the great writers themselves as well as their commentators in regard to the sequence of their ideas, and the order in which they occurred to them. He has given a good deal of attention to economic and philosophic literature, and hidden beneath a repulsive phraseology may possibly lurk many observations worthy of consideration. But to disentangle them would require far more time and space than I have at my command; they may be commended to the respectful consideration of the few scholars who really have any particular knowledge of the subject; to the general public the whole exposition will be impressive and unintelligible.

It is a hard duty to criticise frankly a book like this. If one could be sure that it would only be read by persons who had a fair acquaintance with English history and literature, one would be

disposed to pass over its wild statements and baseless arguments with a smile, and to dwell upon its stimulating and thought-promoting qualities. But one cannot but fear that the academic position and the, for many reasons, well-founded reputation of its author may give it a wide vogue in American colleges. I can hardly conceive anything more mischievous than to put this in the hands of students at a time when they ought to be learning the well-ascertained outlines of history, the methods of sound argument, the rules of evidence, and a lucid and unpedantic style. Lest I should be thought too severe, let me end with two examples of the sort of reckless assertion which proceeds from Puritan prejudice masquerading as psychological science. Professor Patten takes the Puritan side as against Laud with regard to the Book of Sports. There is a good deal to be said for that view. But what excuse can there be for such an assertion as this?

"The key to the situation lies in the once universal notion that the popular sports, by exciting the sexual passions, promoted the growth of population. The loss of virtue that the Puritans denounced in the May games was to the ruling class an argument in their favour. The noble who debauched the wives and daughters of his tenantry did not think that he did wrong; on the contrary he told himself that he was doing the nation a service" (p. 137). The explanation which the author provides for this remarkable condition of affairs only shows how dangerous a thing may be a little anthropology. And now read this:

"It must be remembered that the church was in reality a civil organisation whose main end was peace and security. Its clergy did not, therefore, express that condemnation of sensual indulgences which modern opinion demands, but were content if no breaches of the peace occurred, in the belief that the nation was safest when the attention of individuals was diverted from public affairs by opportunities to indulge their passions" (p. 133). This is not history: it is commonplace anticlericalism. W. J. ASHLEY

Local Variations in Wages. By F. W. LAWRENCE. (London: Longmans, Green & Co. 1899.)

THIS essay, which gained the Adam Smith prize at Cambridge in the year 1897, is a most excellent and useful piece of work. It not only contains most valuable information about wages arranged in a diagrammatic form, but also suggests many lines of new inquiry for the student who shall proceed with any further investigations into this field. The following criticism is deliberately aimed at picking out any faults, and does not in any way profess to make a fair estimate of the value of the work, in the belief that more good is served in the case of such a careful and valuable contribution to economic science by deliberate criticism in detail and other than general praise. A serious piece of work deserves serious criticism.

The author has no right to discard the ordinary methods of statistical inquiry. We have local variations in wages-what are the causes of these variations? In such a case the ordinary statistical procedure is to determine whether the variations in the variable we are considering (in this case generally mean wages per hour) are correlated with some other variable (for instance, the density of the population). It may be urged that the theory of correlation is only completely useful and intelligible when the number of instances is large and the distributions are normal. But even if the number of instances is small, the correlation coefficient remains the best test we have of the possible causal connection between two sets of variables, and though the distributions are skew the correlation coefficient still has a meaning. In a paper published in the Proceedings of the Royal Society, volume lx., Mr. Yule states in conclusion that " We can now see that the use of normal regression formulæ is quite legitimate in all cases so long as the necessary limitations of interpretation are recognised. Bravais' r always remains a coefficient of correlation." This conclusion is, at any rate up to the present, unchallenged. It is therefore very tiresome to find Mr. Lawrence saying (p. 8), “In order to test the correspondence between this population curve and the mean wage, there is constructed a curve showing the difference between the two so that the result may be expressed as

Mean wage per hour in pence, minus square root of population

500

And in order to consider how far the population in any way explains the fluctuation in wages, this new curve must be compared with the mean wage curve to see which is the least irregular." The construction of such a difference curve is, of course, a familiar statistical device and one of the greatest service in certain cases. But here the problem is-What is the correspondence between the population curve (that is, a curve showing the square root of the population divided by 500) and the mean wage? The answer is to be found by calculating the correlation coefficient. The answer to the inquiry is more simply and accurately stated as follows 1: In the case of 75 towns (London not being included) the mean wage is 7.86 pence per hour, with a standard deviation of 0.59 pence per hour. The mean of the function given by the square root of the population divided by 500 is 0.55, with a standard deviation of 0.265, and the correlation coefficient of these two is nearly 0-63-a high degree of correlation. The regression equation can now be written down, and in calculating the standard deviation we have done much "to form a numerical estimate of the total amount of the fluctuations for each of the two curves" (p. 9). The author adds together (irrespective of size) the amounts of the changes in passing from one town to the next. This procedure is open to the obvious objection (which the author very properly makes at 1 Mr. Lawrence kindly supplied me with the figures.