difficult with a loose-leaf book to arrange a system for keeping hundreds of curves in such a way that quick comparisons between any of them can be made. When loose cards are used any card can be compared with any other card instantly, and, if desired, cards for any curve for a series of years may be laid out for comparison with cards for any other curve for any series of years. Anyone without experience in the analysis of curve records for large corporations may take it as a fact that no system of curve records should be installed which does not permit the instantaneous comparison of any curve with any other curve in the whole system.
Fig. 207 and Fig. 208 show the shipments from an automobile manufacturing plant as sales. Many of the automobiles recorded as sold were shipped to branch houses owned by the same company, to be stored there during the winter months when the branch-house sales are very small, for the reason that people do not wish to buy touring cars in winter. In Fig. 209 and 210 more curve records from an automobile business are given. In these cuts also, as in Fig. 207 and 208, cards are shown in groups of three, photographed against a black background. In Fig. 209 we have in the upper curve the actual sales of an automobile branch house selling direct to the automobile user. Notice that the sales in the spring months greatly exceed the sales at any other time of the year. In the first two fiscal years sales were at a maximum in May, while in the third fiscal year sales reached the maximum in April and were fairly large in both March and May.
In the lower curve we have the expenses of this same branch house. A very noticeable increase in the expenses occurred in the fall months at the beginning of the fiscal year, once in October and twice in September. This increase was due to local advertising announcing the new-model automobile for the next season. It was customary for all manufacturers of automobiles to announce their next season's models in the fall months, and the peaks in the expense curve shown here came as they did simply because of this custom of the trade.
It is quite easily seen from the upper curves that the sales for the second fiscal year were much greater than those for the first fiscal year. The total figures, however, show much more clearly the extent of the increase. Because of the excellent sales during the spring months, the curve for the third fiscal year at the right gives the impres-
