Mr. E. W. James, of the Bureau of Public Roads, U. S. Dept. of Agriculture, makes an analysis of the distribution of traffic over the roads of a township located along the section lines of the United States land survey. The market place is taken at the center of the township.[181] [Illustration: Graphic representation of distribution of traffic on roads located along section lines.] His analysis assumes the lay of the country makes all roads equally traversable and that the traffic seeks the nearest highway thence to the main traveled road east and west or north and south through the market center. This analysis shows that 4.8 per cent of the total mileage carry 39.3 per cent of the traffic; that 9.5 per cent of the roads carry 71 per cent of the traffic. In his opinion this analysis corroborates the observation of engineers to the effect that 20 per cent of the roads carry 80 per cent of the traffic. Of course the most important roads, measured in traffic, are the ones nearest the market, 15-22, 15-16, 16-21, 21-22. Following these naming only one of the four symmetrical roads, in the order of importance are 14-23, 14-13, 13-24, 13-x, 14-15, 11-12, 12-x, 12-13, 1-x, 11-14, and 1-12. ------------+---------- Road between| Relative Sections |Importance ------------+---------- 15-22 | 100 14-23 | 60 14-13 | 25 13-24 | 20 13-x | 15 14-15 | 13 11-12 | 7 12-x | 7 12-13 | 2 1-x | 2 11-14 | 1 1-12 | 1 ------------+---------- The same objections to this method hold as to the preceding. Local conditions always affect the travel on roads; hills, valleys, soil, drainage, nearness to other cities, railways, streams, and location of farmhouses, schoolhouses, churches, and factories, all enter into the estimate. A reconnaissance and the good judgment of the observer must supplement any method of formal procedure.