When approaching a question of this sort, statisticians are apt to assume for their answers an appearance of mathematical accuracy. It is needless to say that this is a mere affectation. The best results which can be arrived at are, after all, mere approximations, and they also vary greatly year by year. The body of facts from which conclusions are to be deduced must cover not only a definite area of space, but also a considerable lapse of time. Even Great Britain, with its 17,000 miles of track and its hundreds of millions of annual passenger journeys, shows results which, one year with another, vary strangely. For instance, during the four years anterior to 1874, but one passenger was killed, upon an average, to each 11,000,000 carried; while in 1874 the proportion, under the influence of a succession of disasters, suddenly doubled, rising to one in every 5,500,000; and then again in 1877, a year of peculiar exemption, it fell off to one in every 50,000,000. The percentage of fatal casualties to the whole number carried was in 1847-9 five fold what it was in 1878. If such fluctuations reveal themselves in the statistics of Great Britain, those met with in the narrower field of a single state in this country might well seem at first glance to set all computation at defiance. During the ten years, for example, between 1861 and 1870, about 200,000,000 passengers were returned as carried on the Massachusetts roads, with 135 cases of injury to individuals. Then came the year of the Revere disaster, and out of 26,000,000 carried, no less than 115 were killed or injured. Seven years of comparative immunity then ensued, during which, out of 240,000,000 carried, but two were killed and forty-five injured. In other words, through a period of ten years the casualties were approximately as one to 1,500,000; then during a single year they rose to one in 250,000, or a seven-fold increase; and then through a period of seven years they diminished to one in 3,400,000, a decrease of about ninety per cent. Taking, however, the very worst of years,--the year of the Revere disaster, which stands unparalleled in the history of Massachusetts,--it will yet be found that the answer to the question as to the length of the average railroad journey resulting in death or in injury will be expressed, not in thousands nor in hundreds of thousands of miles, but in millions. During that year some 26,000,000 passenger journeys were made within the limits of the state, and each journey averaged a distance of about 13 miles. It would seem, therefore, that, even in that year, the average journey resulting in death was 11,000,000 miles, while that resulting either in death or personal injury was not less than 3,300,000.