imperial city the census was taken by the censors in person. In the provinces the citizens made their declarations before the provincial magistrates, according to a form or schedule transmitted to the latter by the censors. All these lists being returned to Rome were reduced to a tabular form, so that the total number of Roman citizens, and the slaves and other property possessed by each, could be ascertained at a single glance. These records were preserved in the Temple of Venus Libitina; but none of them have been preserved to us in detail.

It will be discerned by the thoughtful reader that such exact details were prob. required for other purposes than a mere estimation of the number of the pop. It was so. The returns, obtained as described, were required and used for the fiscal purposes of the State; while they also had a distinct value in relation to military organization. Those who desire more details regarding the mode of taking and the purposes of the censuses of both Greece and Rome may, with advantage, consult Smith's Dict. of Greek and Roman Antiquities.

In the first instance the Roman census was taken every 5 years. After a time a good deal of indifference arose regarding it; hence it came to be taken only once in 10 years; and later, but once in 15 years. It is said, upon apparently good authority, that only 75 enumerations were made in the 630 years from Servius to Vespasian; after which it was entirely discontinued. Originally, after the census was taken in Rome, a sacrifice of purification or lustration took place. From this the term of 5 years came to be called a lustrum. The two great jurists, Paulus and Ulpian, each wrote works on the census in the imperial period.

It is stated that the Paternal Government of Peru has from a very early date kept a regis. of all births and deaths throughout the country, and has caused exact returns of the pop. to be made every year by officers appointed by the State.

One of the earliest methods employed with a view to numbering the people in Gt. Brit. was to enumerate the Houses, and then to multiply the houses by the number of people supposed to be occupying them. This method has been frequently resorted to, both here and in Ireland. Another aid was called in. In 1662 the Hearth or Chimney Tax was first imposed. The returns of this tax were supposed to elucidate the number of Families. After the Revolution of 1689, when the taxing of fireplaces was abolished, the computers fell back upon the returns of House and Window Duties. The Poll Tax had been levied on too small a portion of the pop. to be of much value for such estimates. It has been frequently shown that none of these methods could be at all depended upon. The books containing the account of hearth money were long since lost, and it is by no means certain whether Dr. Davenant, in stating the number of houses in England and Wales in 1690, as given in the Hearth Books, really meant the buildings in which families lived, or the families themselves. The returns of the House and Window duties were still less worthy of credit. The collectors were only required to make out and return to the Commissioners of Assessed Taxes, lists of houses within their respective collections chargeable with the duties in question. All cottages exempted from the usual taxes to church and poor were also exempted from the house and window duties; and there was no obligation on the officers to return an account of their numbers. The earlier returns of assessed taxes have never been regarded as very accurate. All these points are necessary to be considered in dealing with the various estimates made in the absence of the authority and the means for an exact enumeration.

Again, the regis. of births and burials have frequently been resorted to as means by which to estimate the magnitude of the pop. In applying them to this purpose, districts in various parts of the country were selected, forming as nearly as possible a fair average of the whole; and a census being taken of the pop. in them, it is learned, by dividing that pop. by the number of births and the number of deaths, the proportion which they respectively bear to the whole number of inhabitants in the districts that have been surveyed; and hence it followed, that to learn the pop. of the entire kingdom it was only necessary to multiply the total number of births, or the total number of burials, as given in the regis., by the proportion which either of them had been thus proved to bear to the whole pop. Thus, supposing that the average proportion of deaths to the pop. had been ascertained, by examinations being made in different parishes, situated in different parts of a country, to be as I to 45 or 50, the entire pop. would plainly be equal to the entire number of deaths in a year multiplied by 45 or 50; or if the proportion of births to the whole pop. had been ascertained, in the same way, to be I in 28 or 30, the pop. would be the product of the yearly births by 28 or 30. It is plain, therefore, that if the regis. of births could have been relied on as accurate, this would form a compendious and not unsatisfactory mode of forming an estimate of the pop. But the early regis. were in almost all cases very far indeed from being accurate. [PARISH REGIS.]

But though the regis. of births and deaths were kept with the most perfect accuracy, it would still be no easy matter to determine the exact amount of the pop. by their means. What may be considered the average and ordin. rate of mort. in a country-and the same thing is true of the average and ordin. proportion of births-is liable to be deeply affected by the occurrence of scarce and calamitous years, and conversely [FOOD, ITS INFLUENCE On Life and DEATH]; and unless all such exceptional circumstances were allowed for, error of greater or less amount must invariably be found in such estimates.

The earliest country in modern Europe which adopted a census of its pop. by actual enumeration was Sweden. A census of its entire pop. was taken in 1749, and again 1752 and 1755. [SWEDEN.]

One of the earliest English writers who appears to have discerned, or at least to have pointed out, the value of an exact enumeration of the people was Corbyn Morris, who in his Obs. on the Past Growth of the City of Lond., etc., first pub. 1751, suggested a B. of mort. arranged so as, after a series of years, to furnish such information. He says:

Under the B. of mort. proposed, one noble instance of information, which might clearly be drawn from it at any period, readily suggests itself. This is, that the total number of persons living of all ages, and also the respective numbers living of each age, might from hence accurately be ascertained: supposing this bill to have been kept for a time past, equal to the utmost extent of life, and also the accession of foreigners during that time to have been nearly equal to the egression of natives.

De Moivre, in the 3rd ed. of his Doctrine of Chances, 1756, speaks of the importance of Taking the numbers of the living, with their ages, through every parish in the kingdom: as was in part ordered some time ago by the Rt. Rev. the Bishops; but their order was not universally obeyedfor what reason we pretend not to guess. Certain it is that a census of this kind once estab., and repeated at proper intervals, would furnish to our governors, and to ourselves, much important instruction, of which we are now in a great measure destitute; especially if the whole was distributed into classes, to married and unmarried; industrious and chargeable poor; artificers of every kind; manufacturers, etc.; and if this were done in each county, city, and borough separately, that particularly useful conclusions might thence be readily deduced; as well as the general state of the nation discovered; and the rate according to which human life is wasting from year to year.

In 1783 the War of Independence in the American Provinces terminated. A Constitution had to be provided for the new U.S. In that Constitution pop. was made the basis of representation. This involved a periodical census of the people. It was resolved that this should be taken every 10 years. The first census of the U.S. was taken in 1790. The pop. was then returned at 3,929,827.

Various writers, many of whom will be noticed under Pop. and various other heads, continued to urge the importance of a correct enumeration of the people of Gt. Brit.; and at length urgency became so great that the Gov. yielded. The first systematic enumeration of the people of Gt. Brit. was therefore fixed to be taken in 1801. It must not be supposed that the proposal was carried through Parl. without opposition. On the contrary, it excited a good deal of alarm. Many considered it in the light of a preparatory measure for some more efficient plan of taxation, or some new scheme with respect to the levy of the militia. These fears operated rather outside than inside the House, and tended in many instances to false or defective returns.

The Act under the authority of which the first enumeration was to be taken is the 41 Geo. III. c. 15, An Act for taking an Account of the Population of Gt. Brit., and of the Increase or Diminution thereof. The day named for the enumeration was the 10th March. Many preparations were required to be made. The rector, vicar, overseer, or other authorized person in each parish in England-and in Scotland the parochial schoolmaster -was to fill up in a schedule sent for the purpose answers to the following questions: Ist.-How many inhabited houses are there in your parish, township, or place; by how many families are they occupied ; and how many houses therein are uninhabited? 2nd.--How many persons (including children of whatever age) are there actually found within the limits of your parish, etc., at the time of taking this account, distinguishing males and females, exclusive of men actually serving in His Majesty's regular forces or militia, and exclusive of seamen either in His Majesty's or belonging to registered vessels. 3rd.-What number of persons in your parish, etc., are chiefly employed in agriculture; how many in trade manufactures, or handicraft; and how many are not comprised in any of the preceding classes?

4th. What was the number of baptisms and burials in your parish, etc., in the several years 1700, 1710, 1720, 1730, 1740, 1750, 1760, 1770, 1780, and each subsequent year to the 31st day of December, 1800, distinguishing males from females ?

5th. What was the number of marriages in your parish, etc., in each year, from the year 1754 inclusive to the end of the year 1800?

6th.-Are there any matters which you think it necessary to remark in explanation of your answers to any of the preceding questions?

The information so obtained was valuable in the degree of its completeness.

It will

be spoken of under POPULATION, OCCUPATIONS, etc. The powers of the Act did not extend to Ireland.

In 1811 the second census of Gt. Brit. was taken; the form of returns being much the same as in the first. The prejudices attending the first enumeration had almost entirely passed away.

66

In 1813 the first census was taken in Ireland; but was regarded as a failure. (See 1821.) At the 3rd census of Gt. Brit., 1821, a return of the ages of the people" was first introduced. In 1821 also a very complete census was taken in Ireland. [IRELAND.] In 1829 there was pub. a pamp, Proposals for an Improved Census of the Pop. Upon this there was founded an able art. in the Edin. Review [vol. 49]. The writer pointed out in a forcible manner that an actual enumeration, or census, of the people "is the only means that can be safely depended upon for ascertaining their numbers."

He

reviewed many of the previous appliances, and showed how, and why, they had failed. He remarks:

Although the happiness of a country does not depend on the circumstance of the inhabitants being few or many, but on the proportion which they bear to the supply of necessaries, conveniences, and enjoyments at their disposal, still it is on many accounts extremely desirable to know their exact number. A nation, having only 10 millions of people, might be decidedly more powerful than a nation with 20 millions, if they were less instructed, less industrious, or less rich. But other things being the same, there can be no doubt that the political power and importance of a nation will be in a very great degree dependent on the amount of its pop. Although, however, the magnitude of a nation had no influence in determining the place which it must occupy in the scale of nations, still there are many most interesting subjects of inquiry which cannot be successfully prosecuted till this magnitude be known. It is impossible, for example, to determine the extent to which levies of individuals, either for the military service, or for any other object, may be safely carried, unless the pop. has been enumerated and classed. It is clearly too for the interest of a very large class of persons, or rather, we should say, of the public, that those questions which depend on the expectation or prob. duration of human life, such as those relating to L. ins., the constitution of friendly sos., and the value of L. annu., should be accurately solved. But this cannot be done without the aid of T. truly representing the laws of mort.; and these cannot be prepared without the aid of censuses, enumerating not only the total number of persons in a country or district, but the numbers at every different age from infancy upwards. The solution of such questions is not, however, the only, nor, perhaps, the greatest service, that may be derived from enumerations of the pop. By comparing together censuses made with the requisite care, and embracing a sufficiency of details, we obtain authentic information, not otherwise attain able, with respect to the proportion which the sexes bear to each other; the changes in the channels of industry; the increase and decrease of different diseases; the effect of epidemics; and an immense variety of other subjects, which are not merely matters of rational and liberal curiosity, but come home to our business and bosoms, and exercise a powerful influence over human happiness.

He said-speaking of course only of the first three-" There is good reason to think that no one of the censuses taken in this country is nearly so accurate as it might have been." He expounds his reasons for this statement, and adds:

A census in which the occupation of every individual and his age were specified would be a most invaluable document; not only would it show the number of individuals belonging to each separate profession or calling, but it would serve to exhibit the influence which different employments exercise on the rate of mort.; while by comparing different censuses of this sort future inquirers would obtain an accurate knowledge of the changes produced by the progress of society, both as respects the numerical relations of the different classes to each other, and as respects their longevity... Let not therefore the approaching opportunity of obtaining a correct census-a census worthy of the country and the age-be neglected. If the census to be taken in 1831 be executed with due care, both as respects the enumeration and the classification of the people, it will be one of the most important documents ever prepared under Parl. authority; and will not only afford a vast deal of information of immediate practical utility, but will be a point of comparison to all future times.

So far from these wise counsels prevailing, the census of 1831 retrograded a step; and a return of the "ages," which had been introduced in 1821, was on this occasion not asked for !

In 1830 a Parl. Committee sat and took evidence, in view of the census of 1831; but, as we have said, all efforts at improvement proved useless on that occasion.

In the census of 1841 the ": ages of the people "-one of the most important points of value for scientific purposes-were again taken. A special committee of the then newly formed Statistical So. of Lond. had been appointed in the preceding year (1840), with a view of making suggestions for improving the machinery to be employed in the census, and of enlarging its scope. This Committee made a very able report [Statis. Journ. vol. iii., p. 72], from which we take the following passage:

But a

The Committee conceive that a census to be made by Gov., extending, as in every country it has extended, beyond a mere counting of heads, contemplates various purposes besides the mere ascer tainment of numerical strength. Some of these concern immediate administration census is also generally extended to circumstances influencing the condition of the people, such as their industrial occupations; and to various facts illustrative of that condition, belonging to the field of vital statistics. It is in solicitude for the collection of the largest amount of these data, and for assuring to them a scientific correctness, that the Committee feel their appointment to the present labour to have originated.

This census was the first taken under the newly-formed organization of the Reg.-Gen. In 1842 Mr. G. R. Porter read before the Statistical So. a paper: An Examination of Some Facts obtained at the Recent Enumeration of the Inhabitants of Gt. Brit., so far as the same have been pub. by the Census Commissioners [Statis. Journ., vol. iv., p. 277, and vol. vi. p. 1].

In 1843 there was read before the Brit. Asso. at Cork, a paper, Obs. on the Census of the Pop. of Ireland in 1841, by Capt. Larcom, one of the Commissioners for taking the census [Statis. Journ., vol. vi. p. 323]. The paper is one of considerable interest.

In 1848 the Rev. E. Wyatt-Edgell read before the Statistical So. a paper, Remarks on the Plan adopted for taking the Census in 1841, with suggestions for its improvement. One of his suggestions was that the census should be taken at Christmas instead of in June, as it had previously been taken.

In 1849 a census of the pop. of Bombay and Colaba was taken. A statement showing the relative numbers of young, adult, and aged persons, classed according to religious persuasions, is given in vol. i. of Assu. Mag, p. 83. [BOMBAY.]

In 1850 another Committee of the Statistical So. was appointed, in view of the census of 1851. Among other recommendations of this Committee was one that the classification of houses be made similar to that of the Irish census of 1841, viz. by adding to the number of houses inhabited and uninhabited the number of rooms in each.

The 6th census of Gt. Britain, 1851, took a wider scope than any of its predecessors. It was made to consist of two parts: (1) Compulsory-comprising an enumeration of persons and houses; and an account of the age, sex, relationship, civil or matrimonial condition, occupation, and birthplace of all the inhabitants; (2) Voluntary-consisting of an elaborate inquiry as to the then existing provision for education and religious worship, and the extent to which that provision was made use of, viz. an enumeration of all the day schools and Sunday schools throughout the country, and of the scholars attending them.

The new features introduced into the compulsory portion of this census were the ascertainment of the various relationships, as husband, wife, son, daughter; the civil condition, as married, unmarried, widower, or widow; and the number of the blind and deaf and dumb-all very valuable details.

In 1853 Mr. Edward Cheshire pub., Results of the Census of Gt. Brit. in 1851, with a Description of the Machinery and Process employed to obtain the Return. He says:

The inquiries undertaken at the census of 1851 were of a far more extensive character than those pursued at any previous enumeration; for it was resolved to exhibit not only the statistics of parishes, and of Parl. and Municipal boroughs, but also of such other large towns in England and Scotland as appeared sufficiently important for separate mention; and the statistics of all the ecclesiastical districts and new ecclesiastical parishes, which during the last 40 years had been created in E. and W. The local machinery by which the objects thus contemplated were to be attained differed considerably in England and Scotland. In England and Wales the Regis. districts, which for the most part are conterminous with the unions were made available for enumerating the pop.; but in Scotland, which is, unfortunately, without any Regis. [a Regis. Act for Scotland was passed 1854], the census was taken through the agency of the sheriffs of counties, and the provosts, and other chief magistrates of Royal and Parl. burghs. The total number of enumeration districts thus apportioned in Gt. Brit. and its Isles was 38,740; to each of these a duly qualified enumerator was appointed.

It was necessary that these enumeration districts should be formed with a careful reference to the various divisions of the country, the pop. of which was to be separately distinguished in the returns. Accordingly, the instructions issued to Registrars in England, for the formation of these districts, directed that while the boundaries of parishes should be taken as the basis upon which to frame the various divisions, attention should be paid to other boundaries.

In the Companion to the [Brit.] Almanack, 1855, there is an interesting art. on the Census of the United States, 1850. [UNITED STATES.]

In 1855 Mr. Geo. Scott addressed to the Assu. Mag. [vol. vi., p. 47], a paper in the form of a letter, On Certain Means furnished by the Census of 1851 for Extending the Application of the Principle of Assu. to the Social Condition. The point of his paper was the providing annu. for unmarried females. His views took practical shape soon afterwards in the founding of the Female Provident L. office.

At the Social Science Congress held at Bradford, in 1859, a paper was read by Mr. Horace Mann, What Information, as to the Social Condition of England and Wales, would it be most desirable to Collect at the Census of 1861. The object of the paper appeared not so much to suggest as to invite suggestions. He said :

The

Perhaps no opportunity is ever presented so favourable to an extensive collection of facts upon many points of social importance as that which is afforded by the decennial census. necessary existence, for the purpose of enumerating the pop., of an elaborate machinery covering every portion of the land, and penetrating to every individual tenement, suggests of itself the desirableness of making as much as possible of so costly and complete an apparatus during the short period of its operation.

He then reviewed the range of the then more recent Census inquiries, especially of the last. On the same occasion Mr. J. T. Hammick read a paper, On the Direction in which the Census Inquiry may be Extended in 1861. He suggested several points for further development, especially as regarded sickness and infirmity. Mr. Nicholas Waterhouse also read a paper, Suggestions for the Next Census.

Mr. A. G. Finlaison, in his Report on Tontines and L. Annu., 1860, calls attention to the fact, that the purposes for which a census is generally taken have remained much the same from the days when Moses numbered the Children of Israel.

The Commissioners appointed to take the Census of 1861 were the Reg.-Gen. (Major Graham), Dr. Farr, and Mr. Hammick. In June of the same year they pub. a preliminary report, in which the following details were given :

The Act for taking the Census of England required that the 31,000 Enumerators employed should copy into as many books, the householders' schedules and other particulars collected by them in their several districts. These books were to be placed, with the schedules, in the hands of the 2197 Registrars, who were to subject them to a strict examination, and make all necessary corrections. This being accomplished, the books and other documents were to be trans. before the 30th April [the Census was taken 8th April] to the custody of 631 Superintendent-Registrars, who were required to test the accuracy of their contents by a further process of revision.

The Census in Ireland embraced inquiries regarding the religious denomination, and The Census in Scotland the answers obtained were regarded as most satisfactory. embraced some important details regarding "house accommodation."

At the meeting of the British Asso. held at Manchester in the autumn of 1861, Mr. Hammick presented a paper, On the General Results of the Census of the U.K., 1861, in which many most interesting details were furnished.

In the Assu. Mag., vol. x., p. 1, is given an abstract of the results of the Census of 1861. In 1862 there appeared from the pen of M. Maurice Block an interesting account of

the Census in France in 1861. The substance of these papers will be found in the Statistical Fourn., vol. xxv., p. 72.

In 1865 Mr. W. L. Sargant read before the Statistical So. a paper, Inconsistencies of the Census of 1861, with the Reg.-Gen. Reports; and the Deficiencies in the Local Registry of Births. The principal conclusions at which he arrived are the following:

1. That the census of 1861 is not to be implicity trusted, but requires further investigation. 2. That male infants below 1 year old are underrated by 36,546 or 12 p.c.; and the female infants by 30,831 or 10 p.c.; that in the 2nd year of life the deficiencies are 11 and 11 p.c.; in the 3rd year, 2 and 1 p.c.; and in the first 5 years taken together, 63 and 6 p.c.3. That this difference of error between male and female infants is prob. owing to the better regis, of male births, and not to a worse enumeration of males in the census. 4. That the males and females together, of all ages under 20, are apparently underrated by 510,440; but that some considerable deductions have to be made from this number. 5. That the males and females together, of all ages, are prob. underrated by more than half a million. 6. That the deficiency in the census is far greater in some districts than in others. 7. That the regis. of births is very imperfect in places; Liverpool and Hull appearing to be the worst, with Lond., Cheltenham, Plymouth, and Portsmouth following in order of demerit. 8. That we have but few materials for comparing the Census of Scotland with calculations made from the regis. of births; but that, as far as we can judge, the Scottish census is as inaccurate as the English one.

In giving these "conclusions," it must not be understood that we indorse them. Our purpose is to supply the substance of, or give reference to, all information upon the subject of which we are treating.

In the same year Dr. Farr read before the same So. a paper, On Infant Mort., and on Alleged Inaccuracies of the Census, in which he reviewed and answered the preceding paper. The two productions throw a flood of light over the question of the mort. of infants. [INFANT MORTALITY.]

In view of the 8th Census, then approaching, several of the learned sos. took action in 1870. The Council of the Statistical So. urged the repetition of the religious and educational census-this time to be made compulsory, especially as to answering the question whether every child or person beyond the age of 7 could read or write. It also urged inquiries as to house accommodation in accordance with the preceding Census of Scotland. The National Asso. for the Promotion of Social Science appointed a Special Committee, consisting of Mr. George Godwin, Dr. Stewart, Mr. Lewis, Mr. Walford, Mr. Safford, Mr. Powell, and Mr. Aldis. That Committee, after several sittings, passed the following series of recommendations;

1. That the Census of 1871 should be taken as nearly as possible at the same date as on the former occasions of 1851 and 1861.

2. That this Committee very strongly urges upon the Government the desirability of adopting a uniform system in taking the Census of 1871 for each of the three divisions of the U.K.; this uniformity not having been observed hitherto.

3. That alike for England and Wales, Scotland and Ireland, the Census "Householder's Schedule" should embrace the following particulars: 1. Name. 2. Sex. 3. Condition (civil). 4. Age (last birthday). 5. Degree of instruction, a, can read; b, can write. 6. Rank, profession, or occupation. 7. Relation to head of family. 8. Where born. 9. Language usually spoken. 10. Professed religion. 11. If deaf, dumb, blind, insane, sick, or infirm; if sick, the nature of the malady to be stated.-[Vide p. 129, Irish Census Report, 1861, Part III., vol. i.]

4. That it would be of the greatest utility if the Government would direct the Census Commissioners for the U.K. in 1871 to undertake, as a subsidiary inquiry, an industrial census, the principles of which were discussed in the last English Census Reports [vol. iii., p. 233]. Such inquiry would include: House accommodation, church and chapel accommodation, school accommodation, manufactures, trades, professions, wages of working classes.

5. That an annual enumeration of merely the number and ages of the pop. is greatly needed, at least for all the principal cities and towns of the kingdom.

The Brit. Asso. at its meeting in Edinburgh appointed a Special Committee consisting of Prof. Jevons, Mr. Dudley Baxter, Mr. Dawson, Mr. Heywood, Dr. Hodgson, and Prof. Waley. Their memorial to the Home Sec. embodied the following:

Your memorialists could specify a great many points in which there was divergence between the Tables of 1861; but they will mention only a few of the more important cases:

1. The detailed pop. T. of Eng., Scot., and Ireland, differ as regards the periods of age specified. The Scotch report gives 21 intervals of age, the Irish report generally 22, and the English only 13. Either one-third of the printed matter in the Scotch and Irish T. is superfluous, or that in the English T. deficient.

2. The classification of occupations is apparently identical in the 3 reports, but there is much real discrepancy between the Irish and English reports, rendering exact comparison difficult.

3. In the Irish report there is no comparison and classification of occupation according to age: classification according to religions being substituted, although such a classification could not be made in England or Scotland.

4. In the appendix to the English report appears a T. (No. 56), giving most important information as regards the number of the pop. at each year of age. Inconvenience has been felt from the want of similar information concerning the pop. of Scotland and Ireland.

5. In the appendix to the Irish report they find some interesting T. (II., III., and IV.), to which there is nothing exactly corresponding in the other reports, so far as they have been able to discover. 6. The T., even when containing the same information, are often stated in different forms and arrangements, seriously increasing the labour of research.

The General Assembly of the Free Church of Scotland also memorialized in favour of a religious and educational census, but it was all of no avail. The Ministry of the day did not understand the cry of "let there be light"! Three separate enactments were passed, as before, for the three divisions of the Kingdom, differing indeed in minor points, but all agreeing in excluding any provisions on the new and important points suggested. The Census was taken on Monday, 3rd April, 1871. The instructions given to enu