XV. OTHER EUROPEAN COUNTRIES.

The more important countries of Europe have been treated at length, but, in order to complete the study, statistical tables setting forth the development of urban populations in the remaining European countries are appended. Trustworthy data are usually lacking for these countries except in recent years, but the best authorities have been relied upon.

NOTE.-The five great cities are Madrid, Barcelona, Valencia, Sevilla, Malaga. The fifteen cities include these five and ten others exceeding 50,000 in population in 1887. The authority for 1800-10 is Hassel 1809; for 1820, Hassel 1823; for 1857, Censo de la Poblacion de España, on 21 de Mayo de 1857, and Kolb, 1868; for 1887, the Census. The statistics relate to the commune or township, which in Spain is exceedingly large and contains a rural population. The estimates for towns 10,000-20,000 are arithmetical (multiplying 15,000 by 72 and 140); but they probably approximate actual conditions. According to the Nomenclater de Espagna, 1888, the urban population was 4,851,903, and the rural 12,713,369, or 27.89 and 72.11 per cent. respectively,

a = decrease. NOTE.-The eight cities are Oporto, Braga, Funchal, Coimbra, Setubal, Evora, Angra, Elvas, being the large cities of 1800. As regards the sources, the data for 1890 are from the Statesman's Year Book, 1897; for 1878, from Censo No. 1 de Janeiro, 1878, Populacao, Lisbon, 1881; for 1801 and 1857 the population of Portugal is given in Block, Bevölkerung Spaniens und Portugals (1861), p. 53, while the population of the cities in 1800 is from Hassel, 1809, and other early hand-books; in 1857, from Kolb, 1860, etc. Additional references are Balbi, Essai Statistique sur le Royaume de Portugal, 2 vols., Paris, 1822; Minutoli, Portugal und seine Colonien im Jahre 1854, 2 vols., Stuttgart, 1855. The earlier figures are regarded as very inaccurate.

SOURCES.-For 1879, Almanach de Gotha, 1885; the figures include parts of Thessaly annexed in 1881. For 1889, Supan. For 1852, Kolb, 1860.

SOURCES.-Census of 1885, in Statesman's Year Book for 1897, p. 1018. The territory of Turkey has been so frequently altered that not even the old estimates can be given for comparison. Constantinople is said to have had a population of 597,600 (Hassel, 1823) at the beginning of the century, and 700,000 at its middle (Kolb, 1860). It has apparently gained in population in recent years, for so late as 1885 the Almanach de Gotha credited it with only 6-700,000.

SOURCE.-Von Asböth, Bosnia und Herzogowina, Wien, 1888.

SOURCE.-Statistique du Royaume de Serbie, Belgrade, 1892-3.

Servian communes are unusually large (38.2 sq. km. on the average), but in this table only the agglomerated population is counted. If the unit taken were the commune instead of the dwellingcentre, the result would be considerably different (1890):

The statistics of 1800-10, from Hassel, 1809, are little more than guesses, but, even when discounted, show a large city population.

1 Including East Roumelia. The authority is Résultats du Récensement de la Population, 1888 (Sofia 1888) and Supan (Ortsstatistik) p. 73–5; for 1850, Kolb 1860. According to estimates at the first of the century, Sofia, the capital, had a population of 46,000 (Hassel 1809). But recently Sofia has grown rapidly and at the census of 1893 had a population of 47,000 (St. Yr. Bk. 1897, p. 1034).

AUTHORITIES.-For 1800-10, Hassel 1809; for 1859-60 Almanach de Gotha; for 1889-90, Supan, except Bucharest, which was returned at 194,633. This inaccuracy shows that the attempt in 1890 to repair the defects of 1889 were unsuccessful. The census of 1859-60 is also of doubtful value. The earlier figures are, of course, only estimates. It should be added that the unit of city populations is the territorial subdivision.

XVI. ASIATIC COUNTRIES.

Statistics of Asiatic Russia have already been presented. Additional statistics for this ancient grand division are not worth much space on account of the untrustworthiness of mere estimates. The modern periodical census so familiar in the Western world has scarcely been introduced in Asia. Even progressive Japan has not thoroughly learned the art of numbering the people. Hence the only really valuable data for present purposes are in the English census of India. § 1. Asiatic Turkey covers a vast extent of territory and contains numerous cities. But only the larger cities are known to the statisticians, and these imperfectly. The estimated population in 1885.

Asiatic Turkey includes Asia Minor, Armenia and Khundistan, Mesopotamia, Syria and Arabia. The four great cities are Smyrna (200,000), Damascus (200,000), Bagdad

'Statesman's Year Book, 1894.

(180,000), Aleppo (120,000). No statistics of growth can be given, beyond noting that in 1823 Hassel published these estimates: Damascus, 130,000; Bagdad, 96,000. The Oriental cities are, it appears, not stationary in population. § 2. Persia.-The latest estimates of the population of Persia are as follows:

I

§3. British India. The official statistics at the present time can be compared in accuracy and trustworthiness with those of the Western nations. And they are especially instructive as presenting the distribution of population in a country which is as densely populated as Europe, and has therefore emerged from the barbaric state and even attained a considerable degree of civilization. But the industrial organization of India is totally different from the European. Industry in India is mainly carried on in local, autonomous and selfsufficing communities. Between these communities there is little commerce, for each devotes nearly its entire population to supplying its own wants. This is essentially true of the

1 St. Yr. Bk., 1897, 811.

The Census of 1891 (General Report, p. 94), shows that nine-tenths of the population are engaged in local industries:

Percentage: