Lower Himálayas. Northwest. gradually to the Tibetan uplands, over which the Siberian temperate vegetation ranges. This is part of the great temperate flora which, with locally individualized species but often with identical genera, extends over the whole of the temperate zone of the northern hemisphere. In the Western Himalayas, this upland flora is marked by a strong admixture of European species, such as the columbine (Aquilegia) and hawthorn (Crataegus oxyacantha). These disappear rapidly eastward, and are scarcely found beyond Kumáun. The base of the Himalayas is occupied by a narrow belt forming an extreme north-western extension of the Malayan type described below. Above that there is a rich temperate flora, which in the eastern chain may be regarded as forming an extension of that of Northern China, gradually assuming westward more and more of a European type. Magnolia, Aucuba, Abelia, and Skimmia may be mentioned as examples of Chinese genera found in the Eastern Himálayas, and the tea-tree grows wild in Assam. The same coniferous trees are common to both parts of the range. Pinus longifolia extends to the Hindu Kush; P. excelsa is found universally except in Sikkim, and has its European analogue in P. Peuce, found in the mountains of Greece. Abies Smithiana extends into Afghánistán; Abies Webbiana forms dense forests at altitudes. of 8000 to 12,000 feet, and ranges from Bhután to Kashmír; several junipers and the common yew (Taxus baccata) also occur. The deodar (Cedrus deodara), which is indigenous to the mountains of Afghánistán and the north-west Himálayas, is nearly allied to the Atlantic cedar and to the cedar of Lebanon, a variety of which has recently been found in Cyprus. Another instance of the connection of the Western Himálayan flora with that of Europe is the holm oak (Quercus ilex), so characteristic of the Mediterranean region. The north-western area is best marked in Sind and the Punjab, where the climate is very dry (rainfall under 15 inches), and where the soil, though fertile, is wholly dependent on irrigation for its cultivation. The low-scattered jungle contains such characteristic species as Capparis aphylla, Acacia arabica (babúl), Populus euphratica (the 'willows' of Psalm cxxxvii. 2), Salvadora persica (erroneously identified by Royle with the mustard of Matthew xiii. 31), tamarisk, Zizyphus, Lotus, etc. The dry flora extends somewhat in a south-east direction, and then blends insensibly with that of the western peninsula; some species representing it are found in the Upper Gangetic plain, and a few are widely distributed in dry parts of the country. FLORA OF WESTERN INDIA. 767 This area is described by Sir Joseph Hooker as comprising Assam and 'the flora of the perennially humid regions of India, as of Malayan peninsula. the whole Malayan peninsula, the upper Assam valley, the Khási mountains, the forests of the base of the Himalayas from the Brahmaputra to Nepál, of the Malabar coast, and of Ceylon.' India. : The Western India type is difficult to characterize, and is Western intermediate between the two just preceding. It occupies a comparatively dry area, with a rainfall under 75 inches. In respect to positive affinities, Sir Joseph Hooker has pointed out some relations with the flora of tropical Africa as evidenced by the prevalence of such genera as Grewia and Impatiens, and the absence, common to both countries, of oaks and pines, which abound in the Malayan archipelago. The annual vegetation which springs up in the rainy season includes numerous genera, such as Sida and Indigofera, which are largely represented both in Africa and Hindustán. Palms also in both countries are scanty, the most notable in Southern India being the wild date (Phoenix sylvestris); Borassus and the cocoa-nut are cultivated. The forests, although occasionally very dense, as in the Western Ghats, are usually drier and more open than those of the Malayan type, and are often scrubby. The most important timber-trees are the tún (Cedrela toona), sál (Shorea robusta), the present area of which forms two belts separated by the Gangetic plain; satin-wood (Chloroxylon swietenia), common in the drier parts of the peninsula; sandal-wood, especially characteristic of Mysore; iron-wood (Mesua ferrea), and teak (Tectona grandis). CHAPTER XXV. Five sources of health returns. VITAL STATISTICS OF INDIA. THE vital statistics of India1 are derived from five chief sources. Of these, the first, or European army, consists of foreigners under special medical conditions, and subject to the disturbing influence of 'invaliding.' The second, or Native army; the third, or jail population; and the fourth, or police; are all composed of natives, but of natives under special conditions as regards food, discipline, or labour. It is dangerous to generalize from returns thus obtained, with regard to the health statistics of the ordinary population of India. For that 1 The literature of Indian health statistics and medical aid may be divided into nine chief classes :-(1) Separate treatises by a series of medical observers, dating from the latter part of the 18th century and continuing up to the present time. (2) Official special Reports of the Medical Boards of Bengal, Madras, and Bombay on the great outbreak of cholera in 1817; the Medico-Topographical Reports (1825-40) of the chief stations of the Madras Presidency, by the Medical Board of that Presidency. (3) The Transactions of the Medical Physical Society of Calcutta (1823-39), and of Bombay (1837-76); the Indian Annals of Medical Science (Calcutta) from 1853-80; other medical journals at different periods in the three Presidencies. (4) Reports on the Medical Education of the Natives of India, commencing with vernacular medical schools in Calcutta and Bombay (1820-30), developing (1835-57) into the Medical Colleges of Bengal, Madras, and Bombay, and extending into medical schools at Haidarábád (Deccan), Nágpur, Agra, Lahore, Balrampur (Oudh), Patná, Dacca, Poona, Ahmadábád. (5) Reports on Vital Statistics by the various Medical Boards, Medical Departments, and Inspectors-General of Hospitals; since 1827 these assume a prominent place. (6) The Annual Reports of the Sanitary Commissioner with the Government of India, since 1874, and of the Sanitary Commissioners to the local Governments; the Annual Reports of the Inspectors-General of Jails, of the Inspectors-General of Police, and of the Health Officers to municipal bodies in the various Presidencies and Provinces. (7) Reports by special Committees or Commissions, such as those on the Bardwán fever, on the cattle-plague in Bengal, the Orissa famine of 1866, the Madras famine of 1878, etc. (8) Annual Reports of the public hospitals, dispensaries, and other medical charities. (9) The Census Reports of India (Imperial and Provincial) for 1872, 1881, and 1891. The author has been unable to test all the dates in this footnote; but he reproduces some of them, unverified, from a memorandum supplied to him by Dr. Morehead, formerly of Bombay. REGISTRATION; DEATH-RATE. 769 population, however, a system of registration exists, and this system forms the fifth source of our data on the subject. tion of untrust In certain Provinces, registration is carried out with some Registradegree of efficiency. But the natives shrink from publicity general touching the details of their life. They could only be forced populato give uniform and absolutely trustworthy returns of births, tion. Why deaths, marriages, sex, and age, by a stringent legislation and worthy. a costly administrative mechanism, from which the Government wisely abstains. In municipalities, however, registration furnishes a fairly accurate account of the vital statistics of the urban population. For the rural Districts, special areas in some Provinces were selected for statistical supervision; and this has been now gradually extended, with the exclusion of certain exceptionally situated tracts, to practically the whole population. imperfect. But the results obtained are still necessarily of error. In treating of the public health of India, therefore, two Sources points must always be borne in mind. The data are obtained either, first, from limited classes under special medical conditions; or second, from a general system of registration spread over the whole country, but which has hitherto failed to yield altogether trustworthy results. General averages from such sources, struck for the entire population, can only be accepted as estimates based upon the best information at present available. Subject to the above remarks, it may be stated that till Death-rate within the last five years, the evidence went to show an annual in India; death-rate of about 33 per thousand in India. During the famine of 1876–78, the death-rate in Madras was ascertained to be equal to an annual rate of 53°2 per thousand. In 1877, in 1877; the death-rate among the European troops in India was 12'71 per thousand, being the lowest recorded up to that year; in the Native army, 13:38 per thousand; in the public jails, during the famine year of 1877, 61'95 per thousand, rising to 176 per thousand in the Madras prisons, which were flooded by the famine-stricken population. In 1883, the death-rate in 1883; returns of European troops in India showed a mortality of 10.88 per thousand; in the Native army, including men absent from their regiments, 14:31 per thousand; and in the jails (among the convict population during a normal year), 33'64 per thousand. In 1890, the death-rate was 11'4 per thousand in 1890. among the European troops; 15'91 in the Native army; and 30.8 among the convict population of the jails. It must be remembered that the system of freely invaliding sick soldiers General death-rate for all India. obscures the death-rate in the Indian army, and especially among the British troops. I have briefly summarized the death-rate returns for the three classes who, although disclosing special conditions of life, are the only large sections of the Indian population under complete and continuous medical observation. I now present the registered returns for the general population for nine years, from 1882 to 1890. But these returns must be taken subject to two warnings. First, that the registration of deaths is often evaded in India, and there are good reasons for believing that the number returned is considerably below the actual facts. Second, that during the whole nine years the ratio of deaths per thousand of the population is calculated on the same total population; namely, that yielded by the Indian Census of 1881. As a matter of fact, we know that the population has increased during the decade 1881-91 by 10 millions, or 9*76 per cent. calculated on the same area. While, therefore, the incompleteness of the registration renders the apparent deathrate lower than the actual death-rate, the increase of 10 millions in the population above the total on which the ratio is calculated, tends to give a higher death-rate than the actual. death-rate during the later years of the period. How far these circumstances counteract each other, it is impossible to say. But both the gradually increasing completeness of the registration and the known increase of the population, tend to bring out a higher apparent death-rate as the period progresses, quite apart from any actual rise in the death-rate. Subject to the foregoing remarks, the following tables exhibit. the death-rate of India, and the principal causes of death, for a period of nine years, as presented to Parliament in 1892 :— |