The true function of such a chart is to portray comparative fluctuations. This desired result is secured clearly and satisfactorily when the factors or quantities compared are of nearly the same value or volume, but analysis will show that the result is not accomplished when the amounts compared differ greatly in value or volume. The extent or degree of the fluctuation as indicated on the ordinary chart depends in a measure on the proximity of the curves to the top or bottom of the chart. The chart registers the actual change in the value rather than the ratio or percentage of change. The wider the range of scale the greater the variation between the actual and the relative changes.

This same criticism applies to charts which employ two or more scales for various records or curves. If the different scales are in proper proportion the result is the same as with one scale, but when two or more scales are used which are not proportional, an indication which is absolutely false may be given with respect to comparative fluctuation. Charts made on a percentage basis are used to some extent in graphic work, and these correct the deficiency in the ordinary chart by showing the changes in the percentages of increase or decrease. In correcting one deficiency, however, another is introduced. The percentage scale gives no clue to the magnitude of the quantities represented by a curve. The true proportions of relative changes are shown, but the actual values are not indicated. The use of the percentage scale also requires considerable labor for computing the percentages of change.