hollows. If all the surfaces are concave, as they are seen to be in Fig. 237, the scraping is rather a simple matter since the surfaces between the two wires are lower than the wires themselves. If a solid model were made for data such as that of Fig. 236, much care would be necessary, for in that case some of the surfaces are convex and it would be essential that enough material should be left between the different wires to permit giving the nicely rounded smooth surfaces which would correctly represent the data. When the surfaces have been scraped as nearly as possible to the proper shape they may be smoothed by rubbing with fine sandpaper. The external flat faces of the model may also be sandpapered advantageously to remove any marks which may remain from the surface of the box used as a mold. Lines such as are seen in Fig. 237 can be ruled on the surface and the different scales can be lettered by hand. A few coats of boiled linseed oil will harden the surface and give an attractive finish.
In Fig. 237 the lines giving oval figures represent certain operating conditions which, from experiment and from study of the model itself, have been found to be most desirable in practice. The oval lines show the limiting conditions best for actual practical operation, and give the real conclusions which the model itself greatly assists in portraying. Solid models and three-dimension charts like those shown are of supreme value in studies of complex data. It is to be regretted that these methods of presentation involve so much labor that they cannot be used as frequently as might be desired.
In political campaigns frequent use is made of statistical information. Campaign orators use figures which sound impressive when combined with a certain amount of eloquence. Too many of the statements, however, tell only a portion of the whole truth, and that portion is, of course, assumed to be the portion which the speaker most desires to have put forward. It is not ordinarily feasible in a speech to give all the facts over a series of years so that the hearer may draw any conclusions for himself. The whole system is weak in that the audience are forced to depend too largely on the statements made by the orator, rather than to draw conclusions of their own from data which are warranted to be authentic. When we have a larger number of people who know how to read curves, it will be a simple matter to present the arguments of a political campaign by means of a projecting lantern with properly prepared charts thrown on a screen. Even now the charts could probably be so made as to be understood
