places—on the road, in casual wards, common lodging houses, public or charitable shelters, and prisons, besides which he has many other resorts, such as barns, brickworks, etc. Then, again, the number of homeless wayfarers varies greatly from time to time, and at different periods of the year, owing to conditions of trade, the state of the weather, or the attraction of seasonal employments."[7]

Although a simultaneous census of the entire vagrant population has never been taken, certain data exist which furnish the basis for at least an approximate estimate. Several of these will be mentioned.

(1) Up to 1868 yearly returns were collected by the Home Office from the different police forces of England and Wales showing the number of vagrants of all kinds known to them. The number on the latest date, April 1, 1868, was 38,179, against 32,528 on April 1, 1867. The number of persons relieved in the casual wards of the country on January 1, 1867, was 5,027, and on January 1, 1868, 6,129, showing that the "casual paupers" at that date represented only about one-sixth of the total vagrant class. If the same proportion to population still held good to-day the number of vagrants of all kinds, based on the mean of the known number of casual paupers on January 1 of the five years 1904-8, viz., 9,355, would be about 56,000.

(2) In the county of Gloucester a count has been made for many years on a night of April of the numbers sleeping in casual wards and in common lodging houses, and the results show that the lodging-houses contain five times as many vagrants as the casual wards. Allowing for vagrants who sleep out of doors, the ratio would not seriously differ from that shown by the police enumeration already mentioned. Applying to the whole country the number of vagrants per thousand of the population of Gloucestershire, the nomad army would be shown to be 30,000. It should be remembered, however, that Gloucestershire is a county of small towns, and lies away from the great streams of population; hence it should not feel the full effect of the vagrant movement.[8]

(3) An enumeration made on March 17, 1905, by the chief constable of Northumberland, by means of police officers placed at the most important points, of vagrants on the roads between the hours of 7.0 a.m. and 7 p.m. gave a total of 300 (exclusive of Newcastle and Tynemouth), equal to about 1 per 1,000 of the population of the area covered. On this basis he placed the number of vagrants in England and Wales at 36,000. Here the omission of two important towns largely invalidates computation; their inclusion would unquestionably give a much higher ratio.

(4) A careful census of vagrants, beggars, migratory poor, etc., is taken by the police for each county, city, and burgh police district in Scotland on two nights in the year, in June and December, showing the number of these persons in (1) prisons or police cells, (2) homes and refuges, hospitals and poorhouses, (3) common lodging-houses or other houses, (4) public parks, gardens or streets, outhouses, sheds, barns, or about pits, brick and other works. The two counts of 1908 gave the following result:—

This was equal to 2.1 and 1.8 per 1,000 of the population respectively, and if these ratios were applied to England and Wales they would represent aggregates of 76,000 and 63,000.

(5) An enumeration of homeless persons in the administrative County of London, made by the London County Council on the night of January 15, 1909, showed a total of 2,088. On that night there were also 1,188 persons in the casual wards of London, and 21,864 in the common lodging-houses and shelters, of whom 10 per cent. were supposed to belong to the vagrant class. This would give a total of 5,462 vagrants as follows:—homeless (sleeping out and walking the streets), 2,088; in casual wards, 1,188; in common lodging-houses and shelters, 2,186; total, 5,462. As the population of the administrative County of London at the date named was estimated at 4,795,757, this total is equal to a ratio of 1.14 per 1,000 of the population. The same ratio for England and Wales would give a vagrant population of about 41,000.

(6) Dr. J. R. Kaye, Medical Officer of Health for the West Riding of Yorkshire, in a report upon the influence of vagrancy in the dissemination of disease, published in 1904, estimated the roving population at 36,000. He has, at my request, explained the basis of his calculation as follows:—

"The estimate of 36,000 refers to England and Wales, and it includes the inmates of casual wards and nomads of the same class who inhabit alternately the casual wards and the common lodging houses according to the state of their pockets. The county police here (West Riding), make an annual census of tramps, and the figure comes out at about 1,000 persons, of whom about 200 are in the casual wards on any given night. Now the Local Government Board reports give the casual-ward population of England and Wales at about 10,000, so that if the same proportions hold good there should be about 50,000 wanderers. Or, on the other hand, if you take our ascertained 1,000 in the county area in relation to our population of 1,249,685, and apply the ratio to the population of England and Wales, we get a figure of 26,000. My figure of 36,000 comes about mid-way between the two estimates given above."

(7) A final estimate which may be quoted is that made at the request of the Departmental Committee on Vagrancy on the night of July 7, 1905, by the various police forces in England and Wales of persons without a settled home or visible means of subsistence: (a) in common lodging-houses; and (b) elsewhere than in common lodging-houses or casual wards. The result was as follows:—

These totals were made up of:—