In plotting the index numbers I have in each case represented 100 by a heavy line. This is the norm shown by the ten-year monthly averages. As long as a curve remains above this line, the figures are above normal, that is, above the ten-year average for the corresponding months.

The general method is illustrated by Charts VI, VII, and VIII, which show the course of the index numbers for bank clearings, railroad gross earnings, number of business failures, commodity prices, stock market,

money rates, crops, and pig iron. The war has brought hardships untold to the statistician who wishes to study the indices of business conditions. It has led to the discontinuance of certain statistics, and a change in the form of others. The Copper Producers' Association, for example, ceased publishing figures for the production of

copper. The Boston Transcript reduced the number of stocks used in obtaining its barometer from twenty-five to twenty, and several other sets of statistics were upset. The charts here represented are only a part of those which have been worked out in my experiments, but they will suffice to explain this method of presentation. The bank clearing statistics used are those published monthly by Bradstreet's for the United States exclusive of New York City. The statistics for the number of business failures and also the index number for commodity prices are from Bradstreet's. Railroad gross earnings are for ten roads as given on Babson's desk sheet. The stock market barometer is that of the Boston Transcript. Money rates are represented by the average monthly rate on 60-90 day commercial paper in New York. For each of these subjects the index numbers from which the curves were plotted were obtained by dividing the actual monthly figures by the averages for the corresponding months during the ten preceding years.

For pig iron an index number for production was worked out upon the same general plan. Then in the same way an index number for price. In order to get a single index number for pig iron which should show the net result of changes both in production and in price I have averaged the production index with the price index. For example, the production index for January, 1913, was 157.5, the price index 99.5, and the average index, therefore, was 128.5. It may prove better to use these two indices separately, but this combined index seems worth trying and watching.

For crops the index number has been prepared first for winter wheat, spring wheat, corn, and cotton. Other crops might be added, but these serve to reprėsent the conditions in the great agricultural sections of the country. During the growing season the condition

reports of the United States Department of Agriculture are used. The index number for each of these crops for each month during this season is found by dividing the condition figure for the month by the ten-year average for the same month. When the final report of the Department of Agriculture is issued the yield per acre is taken as the best index and the index number for each crop is found by using as a base the average yield per acre for that crop during the preceding ten years. From December, when the final report of the Department of Agriculture is issued, till the new condition reports begin in the following spring, the index numbers for the crops remain constant. These constant index numbers during the winter and early spring give a proper representation of conditions, since the influence of the crops on the markets is practically without change during that time.

After the index for each of these crops was prepared, a weighted average was taken. Winter wheat was given a weight of one, spring wheat one, cotton two, and corn four. This weighting corresponds roughly to the relative total value of each of these crops. The final weighted average was taken as the crop index, which was to represent trade conditions in the leading agricultural districts.

Looking at the charts here given, it is apparent, I think, that they fairly represent some of the conditions prevailing during this period. The indices for bank clearings and railroad gross earnings (Chart VI) correlate closely and show the general trend of events. number of business failures (Chart VII) has been relatively high throughout, jumping sharply upward after the outbreak of the war. As regards the future, in

The

1 Further experiments are being carried on to ascertain whether an average index or a separate index for each crop is more satisfactory.

view of the length of time during which failures have been relatively heavy, this is a favorable indication; there has been an unusually severe liquidation and the weak spots must have been pretty thoroly cleaned out. The price index tended to fall until the war came. The stock market showed continued depression.

Chart VIII is, perhaps, most helpful in interpreting the general course of business during these months. The rise in money rates in the early part of 1913, was due to the Balkan war. This was probably one of the primary causes of the business depression which began in the United States early in 1913. The decline later in the year was accompanied by a brightening of business prospects in the fall of 1913. Most industries showed an appreciable improvement about September of that year, but this improvement did not hold. The sharp decline in crop prospects which began in June, 1913, at just the time when the index for money rates was at its highest point accentuated the depression which was setting in and helped to cause the slight crisis of that month. The relatively poor crops, as indicated by this curve, show why there was not more recovery in the fall and winter of 1913 and why business was depressed during the entire spring of 1914. Since other factors were favorable and the crop outlook brighter in the summer of 1914 conditions appeared ripe for at least a moderate business recovery. The breaking out of the European war, however, suddenly tightened the money market and upset the whole business world.

The pig iron index is added to this chart, not as an index of all industry, but as an illustration of this method of comparison. One of the merits of this form of presentation is that the various factors can be studied separately and evaluated. A composite index figure for numerous diverse subjects may cover up significant