Page:Graphic methods for presenting facts (1914).djvu/129

 curves can best be compared if a pencil line is drawn in such manner that the peaks above the pencil line are approximately equal to the valleys below the pencil line for each curve.

From the 1910 Annual Report of the Health Department, City of Boston

Fig. 98. Deaths in Boston of Children under Five Years of Age, under One Year, and from Five of the Principal Infectious Diseases, Expressed as a Percentage of the Total Mortality

Curve A. Deaths of children under five years of age as a percentage of the total mortality

Curve B. Deaths of children under one year as a percentage of the total mortality

Curve C. Deaths from Diphtheria, Scarlet Fever, Measles, Typhoid Fever and Smallpox as a percentage of the total mortality

Note that Curve A shows a much steeper slope than Curve C, yet Curve C drops in 1910 to less than half the figure for 1871. Curves plotted by rectangular co-ordinates should not be compared by the slope of the different curve lines

Fig. 99 contains some interesting information. Though the chart proves fairly well the close dependence of the price of cast-iron pipe upon the price of the pig iron from which it is made, the chart is nevertheless misleading in that the first glance would indicate a much greater fluctuation in the price of pig iron than actually occurred. The reader is apt to overlook the fact that the vertical scale of the chart does not extend below $11 per ton. He is quite likely to think that the price of pig iron had all the rapid fluctuations which would be indicated by the changing vertical distances between the pig-iron curve and the bottom line of the chart itself. The amount of fluctuation would look much less if the chart extended to the zero line of the vertical scale.