No precautions that sanitary science can suggest have been omitted, yet the weekly reports now show an average of twenty thousand deaths from the bubonic plague alone. The officials explain that that isn't so high a rate as inexperienced people infer, considering that the population is nearly three hundred millions, and they declare it miraculous that it is not larger, because the Hindu portion of the population is packed so densely into insanitary dwellings, because only a small portion of the natives have sufficient nourishment to meet the demands of nature and are constantly exposed to influences that produce and spread disease. The death rate is always very high in India for these reasons. But it seems very small when compared with the awful mortality caused by the frequent famines. The mind almost refuses to accept the figures that are presented; it does not seem possible in the present age, with all our methods for alleviating suffering, that millions of people can actually die of hunger in a land of railroads and steamships and other facilities for the transportation of food. It seems beyond comprehension, yet the official returns justify the acceptance of the maximum figures reported. The loss of human life from starvation in British India alone during the famine of 1900-1901 is estimated at 1,236,855, and this is declared to be the minimum. In a country of the area of India, inhabited by a superstitious, secretive and ignorant population, it is impossible to compel the natives to report accidents and deaths, particularly among the Brahmins, who burn instead of bury their dead. Those who know best assert that at least 15 per cent of the deaths are not reported in times of famines and epidemics. And the enormous estimate I have given does not include any of the native states, which have one-third of the area and one-fourth of the population of the empire. In some of them sanitary regulations are observed, and statistics are accurately reported. In others no attempt is made to keep a registry of deaths, and there are no means of ascertaining the mortality, particularly in times of excitement. In these little principalities the peasants have, comparatively speaking, no medical attendance; they are dependent upon ignorant fakirs and sorcerers, and they die off like flies, without even leaving a record of their disappearance. Therefore the only way of ascertaining the mortality of those sections is to make deductions from the returns of the census, which is taken with more or less accuracy every ten years. [Illustration: AN EKKA OR ROAD CART]