spinner exceed or equal those of himself. Those of the wife and children of an agricultural laborer, or of a carpenter or a coal-heaver, are generally unimportant-while the husband in each case receives 15 shillings a week, the weekly income of the one family may be 30 shillings, and that of the other only 17 or 18 shillings." The income of the family, it is evident, therefore, should be taken as the unit in estimating wages.

IV. No consideration is more needful to be observed in the reduction of Nominal to Real Wages than that of the greater or less regularity of employment; yet none is more neglected, not only in comparison of the remuneration of labor in different occupations and localities, but also in a still more important use of the statistics of wages, namely, the comparison of different periods to ascertain whether strikes and trades unions have been really successful in advancing the condition of the working classes. It is not unusual to see the fact of an increase of wages in certain occupations following a threatened or accomplished strike, put forward as proof positive of the efficiency of this instrumentality, without the question being raised whether the certainty and continuity of work may not have been affected injuriously in consequence. Yet it is clear that a nominal increase of wages may be offset by irregularity of employment so as not only

spinner or weaver must be discounted. Again, so far as the employment of the female head of the family in outside labor, or of very young children in any sort of labor, tends to reduce health and strength or to shorten life, this must be set off against the advantage of increased present earnings, in accordance with the principles to be noted in the paragraphs which immediately follow.

1 Lectures on Wages, pp. 8-9.

It is not only true that the opportunities for extra earnings vary greatly as between different occupations, as shown by Prof. Senior's illustration, but such opportunities vary greatly within the same occupation in different localities. Thus Mr. Purdy's tables of Irish agricultural wages show that the "harvest wages" for men range from 2 shillings 6d. a week above ordinary wages, all the way up to 11 shillings.-Statistical Journal, xxv. 448-50.

to render the advance nugatory, but, through the influence on the laborer's habits of industry, temperance, and frugality, to make the change highly pernicious. The neglect to make account of the regularity of employment is probably due not to want of candor in argument, but to the lack of a popular recognition of the vital importance of this consideration. Yet it ought to be evident to the earliest writer on comparative wages that the true time-unit is not less than the entire year. The hourly, daily, or weekly rate of payment is but one factor of wages; the number of hours, days, and weeks throughout the year for which that rate of wages can be obtained is the other.

Varying regularity of employment is due to (1) the nature of the individual occupation, (2) the force of the seasons, (3) social causes, (4) industrial causes of a general character.

In agriculture, for example, we find the first two causes operating to produce great variations in the monthly rate of wages. It is not alone the difference of seasons which makes agricultural wages so irregular; it is in part the

"In the

'This irregularity may be greater or less according to climate or the character of the crops. Some crops require far more days of labor in the year than others. Some countries are locked in frost half the year; in others the ground opens early and freezes late. countries on the Danube, these operations are spread over seven months; in the countries on the north of the Volga they must be concluded in four months."-Hearn's Plutology, pp. 74, 75. An English farmer is ploughing while a New-England farmer is hauling wood on the ice and snow. Mr. Purdy's valuable tables (Statistical Journal, xxiv. 352, 353) show that February is the worst month for employment in agriculture in England; August, the best.

Mr. Purdy gives a table which he deems fairly representative, exhibiting the divisions of agricultural wages between the seasons as follows:

nature of the operations involved. After the seed has been planted, time must be given it to grow, and this would be so even if there were no winter. So in the fisheries it is not stress of weather alone which obliges the laborer to lie idle portions of the year, but in part the reproductive necessities of the fish. In other instances it is the force of the seasons alone which makes employment irregular, as for example in the brickmaking,' quarrying, carpentering, house-painting, and sundry other out-door trades.

The loss of time from sickness, as shown by the statistics of friendly societies and by other evidence, varies greatly in different localities and occupations: an element that can not properly be excluded from the discussion of comparative wages, as such sickness involves not only loss of labor, but also, generally, a positive expense for attendance and medicine.

The following table from Mr. Alex. Glen Finlaison's report (1853) on sickness and mortality in friendly societies, shows the experience of certain large groups of occupations in this respect:

1In brickmaking, in England, it is estimated that men can be employed but 45 weeks in the year, in consequence of rain and frost. In the Northern States of America the failure of employment is for a much longer period.

What we call social causes in restriction of employment include the habits of a community respecting festivities and religious observances.' Vauban estimated the loss of labor in France from fête days and Sundays at 90 days in the year. In some Catholic countries the holidays. more or less scrupulously observed exceed, including Sunday, one hundred. Among the Hindoos they are said to consume nearly half the year. It is doubtless true that poverty sometimes joins with superstition' in imposing excessive fasts, and the want of work may account for the readiness with which a population surrenders itself to celebrating the virtues of a saint; yet there can be no doubt that a force not industrial operates in some countries in reduction of the number of days of labor. A very common multiplier taken in England and the United States in reckoning annual earnings is 300; yet there can be little doubt that this is an exaggeration.

But there are also industrial causes of a general nature

'Mr. Lecky remarks of holidays in Catholic countries: "The number that are compulsory has been grossly exaggerated."-History of Rationalism, ii. 323.

Diplomatic and consular reports to the British Government give perhaps the most recent and exact information on the subject of holidays in the Greek Church.

Consul Calvert reports from Montastir that, reckoning Sundays, there are more than one hundred days in the year when the Christians voluntarily cease work (1870, p. 244). Consul Stuart states the number of days besides Sundays which the Eastern Church attempts to withdraw from labor at 48. Formerly, he says, the number was greater; but the opposition of the working classes to the loss of so much time has caused a reduction in this respect, which will doubtless proceed further (1871, p. 780). Mr. Gould gives the number of working days in Greece as 265 (1870, p. 500). Consul Sandwith gives the number of fête days in Crete as 30 (1872, p. 382). Consul Egerton states that in Russia "besides Sundays there are about 24 holidays in the year, when no work is allowed. Some are saints' days; others, state holidays" (1873, p. 111).

2 Gibbon, chap. xlvii., of the Jacobites, whose five annual Lents the historian is disposed to regard as an instance of "making a virtue of necessity."

which of late years are operating more and more to interrupt the continuity of production and render employment precarious. These causes, though general in their origin, do yet affect localities and occupations very diversely, introducing thus a new element of great difficulty into the problem of wages. Thus there is no reason from the nature of the operations involved, why cotton-spinning should not proceed equably through all the months of the year, but in fact the demands of modern trade require that periods of heavy production shall alternate with periods of dulness and depression.' In the same way the aggregation of vast numbers of workmen into factories for the manufacture of boots and shoes has introduced an irregularity into that branch of manufacture which did not exist when it was confined to the small shop where the master worked with an apprentice and perhaps a journeyman, and made goods for a well-defined and permanent body of customers.

Among the industrial causes which introduce this disturbance into the employment of labor must of course be included strikes and lock-outs. Dr. John Watts has furnished some very instructive computations as to the first cost of strikes. Thus, assuming five per cent addition to existing wages to be the matter in dispute between the employer and the laborer, he shows that if the strike succeeds its results will be, roughly speaking, as follows:*

1Mr. Dudley Baxter, speaking of the operatives in this branch of industry, wrote: "We all know their periodical distresses. It may be said that these were accidents. They are not mere accidents, but incidents-natural incidents of our manufacturing economy. They are sure to recur under different forms, either from gluts, or strikes, or war, and they must be allowed for in computations of earnings.”—National Income, p. 45.

"In 1829 the weavers of Lancashire and Cheshire were earning, at best, from 4s. 44d. to 6s. per week when at work. The most favored had to wait a week or two between one piece of work and the next; and about a fourth of the whole number were out of employ altogether."-Martineau, History of England, iii. 167.

2 Statistical Journal, xxiv. 501. I have sought to show elsewhere (p. 391, n.) that all the time occupied by a strike is not necessarily lost.