In enumerating the houses, some definition of the term was required. In the great majority of instances no difficulty is presented, yet, in certain exceptional cases, the difficulty of defining "what constitutes a distinct house" was considered insuperable by Mr. Rickman; and in the earlier Censuses it was left to "those who made the Return," to decide "whether a college, or inn of court, or a town-house in Scotland, containing as many separate habitations as stories or 'flats,' was to be deemed one house or many." With the exception, however, of a few towns in the northern counties of England, the confusion likely to arise from the term "house" being variously understood, was confined to Scotland, and in 1851 the difficulty was surmounted even there. It was decided that in every part of Great Britain, under the term "house," should be included all dwellings isolated and "separated by party-walls." The enumerator in Scotland, as well as in England, was instructed to the effect that flats and sets of chambers must not be reckoned as "houses." The houses thus defined were returned throughout the country on a tolerably uniform principle.

The character of the houses, the nature of their tenure, and the extent of house accommodation in Great Britain, did not form parts of the Census inquiry; but it was necessary, in order to secure uniformity in the returns, to take into account the great difference in these structures. And this difference is of importance in many points of view; for on the question whether the owner of a flat is the owner of a house, or whether the occupier of a part of a house is a householder, the possession of the electoral franchise, or exemption from the house-tax or from other imposts, may depend. The definition of "house" in the Census Abstracts was laid down, it is perhaps scarcely necessary to say, only for the guidance of the enumerators, and had no reference to objects in which other elements should be probably taken into account.

In respect to the occupation of dwellings in different parts of England and Wales, it is found that certain districts, in which 10 houses are occupied on an average by more than 11 families, contain 843,468 houses, inhabited by 1,209,935 families of 5,612,573 persons; while in the other districts, containing 12,315,036 people, nearly all the families live in separate dwellings, as 2,434,571 houses were occupied by only 2,502,355 families. In the former districts, an average of 143 families of 665 persons lived in 100 houses; in the latter, comprising more than two-thirds of the population, 103 families, of 506 persons, lived in every 100 houses.

The first class of districts consists of towns, the second includes many large towns also, as well as the country districts; from which it may be inferred, as the family is on an average composed of five persons, that some of the families in towns are imperfectly con

stituted, and have less than the normal number of members.

The variations in the dwellings are considerable in the 11 statistical Divisions under which London and the counties of England and Wales have been grouped; and it is difficult to account for all the anomalies which they present. In London (Division I.), 533,580 families, of 2,362,236 persons, occupy 305,933 houses; the average number of families to a house, in several districts, is between two and three, comprising 10, 11, and even 12 persons. Other districts present a different aspect: the 8276 houses in Wandsworth contain 10,117 occupiers, 50,764 persons; the 5927 houses in Lewisham are occupied by 7144 families, comprising only 34,835 persons, or less than six persons to a house. It will be recollected that the houses differ greatly in size, and that the house-room to each person is probably greater in such districts as Marylebone, where large houses have, on an average, 10 persons to a house, than in the small houses of Bethnal Green, where there are not, on an average, seven persons to a house.

In the county-towns, and many considerable towns of the South Eastern Division (II.), containing 532,463 people, the families exceed the houses in number by 22 per cent. (houses 90,705, families 110,576); while, in the other districts of this division, 207,349 houses are occupied by only 218,737 families of 1,095,923 persons. Guildford, Brighton, Hastings, Portsea Island (comprising Portsmouth), Southampton, and Windsor, are the districts in which the excess of occupiers to houses is greatest.

In the South Midland Counties (Division III.), certain districts in Middlesex, with Hertford, Northampton, Cambridge, and a few other districts, have, on an average, more than 11 families to 10 houses; but the families generally live separately.

Few of the houses in Essex, Suffolk, Norfolk (Eastern Division IV.), contain more than one family.

In the South Western Division (V.), Dorchester contains, in 1109 houses, 1573 families, 7383 persons; Exeter, in 5111 houses, 7327 families, 32,823 persons: Falmouth, in 1446 houses, 2048 families, 10,220 persons; Bath, in 10,877 houses, 16,173 families, 69,847 persons; Plymouth, and the adjacent districts of East Stonehouse and Stoke Damerel (Devonport), had 102,380 inhabitants, constituting 22,868 families in 10,132 houses; and consequently, on an average, more than two families of 10 persons to a house. No other town in England, and only a part of London, exhibits such proportions. In 1801, Plymouth contained rather less, in 1811 rather more, persons to a house than are found in 1851.

A larger proportion of the families in the West Midland Division (VI.), around Worcestershire, live in separate houses; the Bristol district, however, had 65,716 persons, 14,613 families, in 9122 houses; Clifton, Gloucester, Hereford, and Birmingham, had also an excess of families in the houses.

In the North Midland Division (VII.), Nottingham and Wilford (a subdistrict of Basford) contained little more than five persons in a house, although they have nearly 11 occupiers to 10 houses. It is a remarkable fact, that in the counties of Leicester, Rutland, Lincoln, Nottingham, and Derby, nearly all the families dwell in separate houses; and that in all Yorkshire (Division IX.), the City of York, and Hull, are scarcely exceptional cases to the same rule; for York has, in 2412 houses, 2801 families, 14,405 persons; Hull, in 9733 houses, 11,325 families, 50,670 persons in Leeds and Sheffield the average number of persons in a house does not exceed five.

More than 300,000 òut of 472,907 families in Lancashire and Cheshire (Division VIII.) | live in separate houses; the population of the districts in which the families are to the houses in the ratio of 11 and upwards to 10, is 785,255, consisting of 150,563 families, living in 121,707 houses; while in other districts 1,705,452 people, consisting of 322,344 occupiers, live in 314,281 houses. Liverpool, West Derby (adjacent to Liverpool), Bolton, Manchester, Salford, and Wirrall (including Birkenhead), are the chief districts in which two or more families, in many cases, occupy the same house. The Liverpool district contains 258,236 people, 47,271 families, in 35,293 houses; the Manchester district, 228,433 people, 44,621 families, in 36,701 houses.

In the Northern Division (X.), across the Tees, a very different arrangement of the dwellings exist; in proceeding from Stocktonon-the-Tees to Durham and Sunderland, to South Shields, Gateshead, and Newcastleupon-Tyne, the proportional number of families and persons to a house increases. In Sunderland 8990 houses contained 15,984 families, of 70,576 persons; in Newcastleupon-Tyne, 10,685 houses held 18,632 families, 89,156 persons. The Scotch style of building, or a style approaching it, prevails in the towns, while in the country the isolated system is general; so that, in some districts of the Division, 604,269 persons, constituting 128,437 families, dwell in 94,410 houses, while in the rest 364,857 persons, 72,957 families, dwell in 70,284 houses.

In WALES (Division XI.), the system of isolated dwellings prevails. Chepstow, Pontypool, Newport in Monmouthshire, Cardiff, and Carmarthen, are the only districts in which the proportional number of persons and families, in the houses, exceeds the average to any extent.

It is not improbable that the houses were made larger and stronger in the seaport towns, and the border counties of England and of Scotland, than they are in the secure inland towns of England, to meet the exigencies of the troubled times in which the style of building originated; and that, from various reasons, the fashion extended and was perpetuated. Scotland probably adopted her style of building houses in the large towns from the Continent; and the insule of an

cient Rome are reproduced in the blocks of building on the Clyde.

Another analysis has been attempted, in 14 Subdistricts, of the Families in connection with the Houses which they occupy; reckoning here, for the occasion, single lodgers, and widowers or widows without children, as constituents of other families, or as not in themselves constituting separate Families. The result is, that in 35,876 inhabited houses, there were in this sense 48,985 Families; 3901 Widowers, 1955 with children, 1946 without children; 10,473 widows, 6356 with children, ¦ 4117 without children; 14,315 single lodgers, 9434 being single men, 4881 single women. 48,985 is nearly the number of families returned by the enumerators, who can, therefore, rarely in these subdistricts have counted single lodgers as separate occupiers. The total population was 242,164, and consequently there were, on an average, nearly 7 persons to a house, 5 to a family.

Of the houses, 1862 were uninhabited; 1608 were in charge of persons who did not occupy them; 26,309 were occupied by single families, each of 4789 by two families, of 1523 by three families, and of 38 by ten families and upwards, as shown in the following Table :

mates, 2503 eight inmates, and so on; and one house held 74 persons. Several of the houses belonged to families with large establishments; but the houses which contained the higher number of inmates were lodginghouses or hotels. Four persons to a house was the proportion of most frequent occurrence. 17,046 houses contained 6 persons or more; which was less than the number that contained from 1 to 5 persons. Only 5765 houses contained ten or more inmates: 988 twenty or more inmates. More than the half of the total population dwelt in 9956 houses, containing eight or more inmates, and twelve on an average. About one-fourth part of the population dwelt in 3102 houses; each of which held 19 persons on an average.

The number of inmates in the respective public institutions of these subdistricts was, in seven workhouses-68, 77, 100, 189, 196, 308, 1656; four barracks, 107, 263, 354, 562; three hospitals, 107, 196, 263; two prisons, 145, 954; one asylum, 67; a training school, 244.

It will be observed that the transition in the numbers is less abrupt than from a comparison of the average number of inmates in houses, with the average number of inmates in public institutions, would suggest the lodging-houses and the hotels form a connecting link between the two series.

The family, as before observed, consists of a head and of dependent members, living together in the same dwelling. The type of the family is the community in a house, consisting of the husband, wife, children, and servants; but the most common of all particular cases is that of a husband, wife, and children. Five classes may be distinguished on the basis of the natural family, and these may be subdivided, according to the numbers and character of the members. Thus (1), the husband-and-wife, as head, may be alone, or have children in the house, with servants, without servants, with or without relatives, visitors, or workpeople, in a great variety of combinations; (2), a widower is the head of the family; (3), a widow; (4), a bachelor; (5), a spinster is the head of the family-all subject to the same combinations as Class I.

In order to throw some light, by classification, on the constituent parts of families, the returns of the 14 subdistricts have been analyzed; and the following are the chief results-41,916 heads of families were husbands-and-wives; 10,854 widowers or widows; 14,399 bachelors or spinsters; in 440 and a few more cases the head of the family was absent from home. 36,719 (more than half) of the heads of families had children living with them-they were parents; 7375 (nearly 1-10th) had servants-they were masters and mistresses; 4070 (1-17th) had with them visitors they were hosts; 8543 had relatives with them; 1020 (1-67th) had apprentices or assistants in their respective trades-they were masters. Of the 67,609 families, only 3503 (5.2 per cent.) consisted of husband, wife, children, servants; whilst 4874 consisted of man, wife, and servants, which

Aristotle characterizes as the constituents of a family.* The heads in 24,180 instances had neither children, relatives, visitors, servants, nor trade assistants; like some corporations they may be characterized as "sole." Thus, of 41,916 married couples, 8610 were "sole:" 29,969 had children residing with them, either alone or in other combinations; namely, 21,413 had children alone—that is, without servants, trade assistants, visitors, or relatives; 3132 had children, and other relatives (alone); 2269 had children and servants; 1421 had children and visitors; 149 had children and trade assistants; 550 had children, relatives, and servants; 245 children, relatives, and visitors; 360 had children, visitors, and servants; 33 had children, relatives, and trade assistants; 65 had children, visitors, trade assistants; 166 had children, servants, trade assistants; 69 had children, relatives, visitors, servants; 50 had children, relatives, servants, trade assistants; 34 had children, visitors, servants, trade assistants; 5 had children, relatives, visitors, trade assistants. Only 8 families consisted of husband and wife, children, relatives, visitors, servants, trade assistants. Of 41,916 natural families (comprising husband and wife) nearly 21 per cent. (1 in 5) consisted of the husband and wife sole, 71 per cent. of the husband and wife, with their children in various combinations: 8 per cent. of the husband and wife, with servants and others. Of 10,854 families, at the head of each of which was a widower or widow, 3264 were heads sole, 6405 had children in various combinations, 1185 servants and other connections: the proportions of the three classes were 30, 59, and 11 per cent. respectively. Of 14,399 designated families having a bachelor or a spinster at their head, 12,306 were sole (lodgers generally); 238 had children residing with them, born out of wedlock, and 1855 had servants, relatives, or visitors with them. The percentages of the whole of the classes stand thus:

Of the 41,916 natural families, each having a man and wife at the head, 4854 had servants; or 12 per cent. of the number (1 in 8 or 1 in 9) of these families kept servants.

The average number of members in a family depends to a considerable extent on the fact whether (1) single lodgers are or are not taken as families; (2) on the number of children at home; and (3) on the number of servants. The number of children at home in families varies considerably; the greatest number of children at home in one family was 12, in

* Aristotle, Politics, Book 1.