By the method in the lower part of the illustration the plotted line more closely approaches a true curve, there is much less variation in the direction of the lines from month to month, and the general trend of the curve line is easier for the eye to grasp. Compare the two curves for the summer months of the year 1909-10. In the upper curve a series of steps, and in the lower curve an almost straight line from April to September inclusive, indicate an increase by fairly equal increments during those months. Certainly equal increments are more easily represented by the straight

CURVES FOR THE EXECUTIVE

200 OOO

160000

160 000

140 000

1 20 OOO

100000

80000

60000

40 000

20000

200000

180000

160000

140 000

120000

100 000

80 000

6 0000

40 000

20000

1908-09

1909-10

19tO-ll

Fig. 202. Total Sales of the "Metropolis" Branch House of the "R.S.T." Automobile

Company for Three Years

These two curves are plotted from exactly the same figures. The upper curve is drawn by the method still used by some railroads, but generally going out of use. The lower curve is plotted by a simpler method which should be universally used

line as in the bottom curve than by the ragged series of steps shown in the upper curve.

Another disadvantage of the method of steps with flat tops as compared with the method using slanting lines and peak tops, is seen when two or more curves are so drawn that they intersect on the same sheet of co-ordinate paper. Curves plotted with peak tops can be drawn very close together and yet be fairly distinct from each other, as will be noticed in many of the illustrations in this book. If, however, two curves with flat tops hke those shown in the upper portion of Fig. 202 are plotted in such manner that they intersect each other, the re-