"The printed instructions issued to each observer contained the following rules as a guide in estimating the weights of vehicles: Special note will be made, in the column of remarks, of any unusu ally heavy loads, such as 6-horse trucks loaded with stone or iron, and an estimate be given of their weight.' " The traffic is divided into three classes, light weight (less than one ton), medium weight (between one and three tons), and heavy weight (more than three tons); and in order to reduce the personal equation of the different observers to a minimum, the directions specify what classes of vehicles are to be counted in each class of weight. Nothing is then left to the observer's judgment and estimation except the question of `heavy' and ' light ' loads in one-horse and two-horse wagons. The result of different estima tion in this respect between two observers would simply change a portion of the vehicles from one class to another, and the error in the final result could hardly exceed 5 per cent.
"The weight of the horses is discarded altogether, not because they do not constitute a factor in the wear of the pavement, but because they were discarded in the English reports, and it was desired as far as possible to make comparisons with them. The addition that would have to be made if the horses were included would vary with the traffic. On streets where light vehicles pre dominate (as on Fifth Avenue, New York). the addition to the ton nage by including the weight of the horses would be about 85 per cent; on streets with heavy vehicles (such as Wabash Avenue.
Chicago), it would be only about 40 per cent; and for other streets it would be between these two limits.
"The average daily traffic was obtained by dividing the total record for six days by six. To obtain the tonnage, the light weight vehicles were estimated to average one half ton each (in cluding their loads), the medium-weight two tons, and the heavy weight four tons. Multiplying the daily average of vehicles in each class by these figures and adding together the products, the total tonnage was obtained; and dividing this by the width be tween curbs, we get the daily average tonnage per foot of width.
" The average tonnage per vehicle is an almost infallible indicator of the character of the street, i. e., whether devoted to residential or business purposes. It ranges from 0.68 tons on Fifth Avenue in New York to 2.08 on a portion of Wabash Avenue in Chicago. The same character is indicated by the proportions of light and heavy vehicles on the street. On Fifth Avenue, for instance, 91 per cent of all the vehicles weigh less than one ton, while on Wabash Avenue only 25 per cent of them have so little weight. The general
average for all the cities is as follows: less than one ton, 67 per cent, between one and three tons, 26 per cent; more than three tons, 7 per cent." During the years 1888 and 1889 the Warren-Scharf As phalt Paving Co.* made observations on twenty-five streets in ten cities on seven kinds of pavements, the length of the observa tions for a particular street varying from one to seven days. In these observations the tonnage including both horses and vehicles, varied from 6 to 138 tons per day per foot of width of pavement.
In 1892 the Board of Public Works of Indianapolis, Ind., took a traffic census of thirteen of the principal streets of that city, the observations being made according to the system employed by Captain Greene (§ 879).t The traffic varied from 16 to 110 tons per day per foot of width of pavement.
In 1892 a census of traffic was taken on twenty-five streets of Montreal, Canada.$ The method was the same as that em ployed by Captain Greene (§ 879) except that the observations were made from 7 a.m. to 6 p.m., and except further that the weight of the horses was included, the weight of each being esti mated at half a ton. The tonnage varied from 17 to 146 tons per day per foot of width.
In Paris it is customary to state the traffic in number of "col lars " per unit of width, no tonnage being given.
It is desirable that engineers in charge of streets and roads should ascertain by direct observation the amount of ton nage passing over each particular pavement, in order that the service per unit of cost of different pavements may be accurately compared. The only measure of the durability of a pavement is the amount of traffic tonnage it will beltr before it becomes so worn that the cost of replacing it is less than the expense incurred by its use.