cost of working the traffic as a whole, the difficulty is to find a sure basis upon which to apportion it between the three classes. The number of passengers carried is obviously not a safe guide, because it is necessary to take into consideration the wide difference in the conditions under which they are conveyed. To divide the amount in proportion to carriages employed is impracticable, because a great number of the carriages contain accommodation for all three classes in varying proportions. The number of seats provided in the carriages employed is a somewhat more reliable basis; but even this leaves something to be desired, because a first class seat occupies more space in a train, and adds more to its weight than a second class seat, while even a second class seat occupies more space than a third. In addition, the lavatories for first class, and the great size and weight of the saloons of various descriptions have to be reckoned with.
After very careful consideration of the subject, the writer has arrived at the conclusion that the nearest approximate results are to be obtained by apportioning the total passenger working expenses betwen the three classes in the ratio of the seat space occupied in the entire carriage stock of a company, and a calculation made upon this basis, with regard to the passenger traffic of the London and North Western Railway in 1888, and in 1871, the last year before the introduction of the changes which have been enumerated, gives the following results:
