The dependent variable is made the vertical scale here, just as in most charts where there is only one independent variable instead of two. The use of the isometric ruling is not as common as it would be if it were more widelv realized how easily charts may be prepared to give the effect of solid models without the great labor which a solid model necessarily involves. Fig. 236 shows another method which may be used instead of the more laborious and costly solid model. Different curves are plotted by assuming at frequent intervals constant values for one of the vari-

336 GRAPHIC METHODS

ables of the data, and then plotting curves for the other two variables. These curves are made to the same scale on sheets of cardboard, and then the outline of the curve is cut out with shears so as to give a series of different cardboard curve sheets. The several sheets are carefully marked for their intersecting points, and are then cut halfway through in the upward and downward direction on the intersection lines so that the curve sheets may be fitted together to give an effect like that seen in Fig. 236. A cardboard exhibit on the scheme of Fig. 236 is, in many cases, just as satisfactory as a solid model and it has the advantage of being quite easy to prepare without any special apparatus or materials being required.

Fig. 237 shows a solid model of the type which may be considered the acme of graphic work when there are two independent variables. A model of this kind is ordinarily made of plaster of Paris, as that is a material easily handled and capable of being made into any shape desired. In making such a model the usual procedure would be to rule a flat board with lines at properly spaced intervals for each of the two horizontal scales.