To perform this service, there were, in 1866, 8125 locomotives at work in the United Kingdom, consuming about three million tons of coal and coke, and flashing into the air every minute some thirty tons of water in the form of steam in a high state of elasticity. There were also 19,228 passenger-carriages, 7276 vans and breaks attached to passenger-trains, and 242,947 trucks, wagons, and other vehicles appropriated to merchandise. Buckled together, buffer to buffer, the locomotives and tenders would extend for a length of about 54 miles, or more than the distance from London to Brighton; while the carrying vehicles, joined together, would form two trains occupying a double line of railway extending from London to beyond Inverness. A notable feature in the growth of railway traffic of late years has been the increase in the number of third-class passengers, compared with first and second class. Sixteen years since, the third-class passengers constituted only about one third; ten years later they were about one half; whereas now they form nearly two thirds of the whole number carried. Thus George Stephenson's prediction "that the time would come when it would be cheaper for a working man to make a journey by railway than to walk on foot" is already realized. The degree of safety with which this great traffic has been conducted is not the least remarkable of its features. Of course, so long as railways are worked by men, they will be liable to the imperfections belonging to all things human. Though their machinery may be perfect, and their organization as complete as skill and forethought can make it, workmen will at times be forgetful and listless, and a moment's carelessness may lead to the most disastrous results. Yet, taking all circumstances into account, the wonder is that traveling by railway at high speeds should have been rendered comparatively so safe. To be struck by lightning is one of the rarest of all causes of death, yet more persons were killed by lightning in Great Britain, in 1866, than were killed on railways from causes beyond their own control; the number in the former case having been nineteen, and in the latter fifteen, or one in every twenty millions of passengers carried. Most persons would consider the probability of their dying by hanging to be extremely remote; yet, according to the Registrar General's returns for 1867, it is thirty times greater than that of being killed by railway accident. Taking the number of persons who traveled in Great Britain in 1866 at 313,699,268, of whom fifteen were accidentally killed, it would appear that, even supposing a person to have a permanent existence, and to make a journey by railway daily, the probability of his being killed in an accident would occur on an average once in above 50,000 years.