RAILWAY ACCIDENTS. 



imprudence or want of due care and precaution on the part of the 

 traveller himself, a case of frequent occurrence, the railway 

 managers and the railway machinery must stand acquitted, and 

 the character of the casualty will indicate the nature of the 

 precautions which are necessary on the part of passengers, to 

 guarantee them from its recurrence. 



7. To estimate the chances of accident fatal to life or limb, it 

 is not enough to compare the number of passengers killed or 

 injured with the total number booked at the stations. This is 

 an error which is very apparent, yet it has been committed year 

 after year without correction by the government railway com- 

 missioners in their reports. Such an estimate is based upon the 

 implied assumption that all passengers run the same risk, what- 

 ever be the distances they travel. Thus, a passenger booked 

 from London to Aberdeen is assumed to incur no more risk than 

 one who travels from London to Greenwich. 



It is evident, on the contrary, that the risk incurred, other 

 circumstances being alike, is in the exact proportion of the 

 distance travelled. A passenger who travels an hundred miles 

 is exposed to ten times as much risk of accident as one who 

 travels only ten miles. The premium upon insurance against 

 railway accident should obviously be a mileage. 



8. To ascertain the extent of the danger incurred in travelling 

 on any proposed system of railways, it would, therefore, be 

 necessary to compare the number of accidents which take place 

 in any given time, a year, for example, with the total mileage of 

 the passengers in the same interval. This mileage may always 

 be determined with great precision. It is the sum total of the 

 distances travelled by all the passengers booked in that interval. 

 Now, since the fares paid by the passengers of each class bear a 

 fixed average proportion to the distances to which they travel, 

 their total mileage will be found with all the necessary exactitude 

 by dividing the gross receipts arising from each class by the 

 average fare* per mile. 



9. To render this method of investigation more clear, let us 

 suppose, that in a given time the quantity of passenger traffic 

 which has taken place on a given system of railways, is repre- 

 sented by an hundred millions of miles, that is to say, that all 

 the distances travelled by all the passengers booked, when 

 summed up, will make an hundred millions of miles. This 

 would then Joe the same as if a million of passengers were trans- 

 ported over a distance of an hundred miles. 



Now, let us suppose that in the same interval ten passengers 

 were killed, and an hundred wounded by accidents occurring in 

 their transport. It would then follow, that of a million of 

 164 



