GRAPHIC REPRESENTATIONS OF DATA FOR FIVE YEARS.

It remains now to consider in the aggregate the graphic representations of the average data collected during the five years' study of this problem. First in this connection will be presented the charts showing graphically the aggregate data in the same order as that in which they were presented in the charts for each year.

In

Chart No. 4 includes the graphic illustration of the percentage of sugar in the beet, the percentage of sunshine, and the latitude of the station The stations are arranged in the order of the sweetness of the beets produced. Lexington, Ky., occupies the lowest position in this chart and Geneva, N. Y., the highest. The lines present irregularities and no attempt was made to select certain stations to secure a straight line. The curve representing the latitude in general lies in the same direction as that of the sugar content. fact, if each of these lines were reduced to a straight line by computation, they would lie almost parallel. The chart shows but little relation between the percentage of sunshine and the sugar content of the beet; in fact, the curve of sunshine is highest at the Lexington station, which is the lowest in sugar content. If only a part of this curve were studied, beginning with the Blacksburg Station, there would be less disagreement between the sunshine curve and the sugar content, but such a study is entirely inadmissible, as it would indicate the utilization of selected data in harmony with a certain theory and the exclusion of data not in harmony therewith. In order to present more clearly the relations of the two principal lines in this chart a special graphic illustration has been made (chart No. 7), including only the five stations which were compared throughout the whole five years of the study. These lines are superimposed as closely as possible in such a way as to transect each other proportionately. This method of illustration presents more clearly the existing relations.

In chart No. 5 a graphic illustration has been made of the average data for five years on the sugar content, purity, temperature, and length of day. Considering first the temperature, it is seen that, although there are extreme variations in some instances, there is nevertheless a distinct and marked relation between the temperature and the sugar content. As the temperature rises the content of sugar falls. The important exception to this rule is shown by the curve for Lafayette. The causes which produced so rich a beet at Lafayette have already been sufficiently discussed and account in a great measure for the variation. It is evident, moreover, that in so complicated a problem as the environment presents for a period of five years it was quite impossible to get a correct solution of all the equations between the different factors. In other words, there are

FIG. 4.-Sugar content of the beet compared with the latitude and sunshine record for five years.

more unknown quantities than there are equations with which to determine them. The curve for the length of day shows a very close correspondence, as shown by the graphic chart, to that for the con

FIG. 5. Sugar content of the feet compared with the purity of the juice, temperature, and average length of day for five years.

tent of sugar: since the length of day is only a function of the latitude there should be a very close agreement between the two curves. This agreement, however, is displaced by the injection of the factor

FIG. 6. Sugar content of the Feet compared with the rainfall, and altitude of station for five years.

of altitude into the problem. The altitude affects very materially the sugar content but does not affect the length of day. Any variation, therefore, between these two curves can be very properly accredited to the altitude. By this method of study it may be possible to trace more accurately than could be done otherwise the influence of the altitude upon the sugar content.

The line indicating the purity of the juice for the five years shows in a most convincing manner the effect of the sugar content upon purity; as the sugar content rises the purity increases. This relation is shown more vividly in chart No. 10, where only the sugar content and purity curves for the five stations for the five years are platted.

Chart No. 6 shows the average data for the period of five years on the sugar in the beet, the total rainfall, and the altitude of the station. There is an apparent relationship shown between the sugar content and the total rainfall, but it must be admitted that such a relation is largely fortuitous. The distribution of the rainfall, unfortunately, could not well be graphically illustrated for the five years, and inasmuch as the distribution is quite as important as the total amount of rainfall the two should be considered together for the purpose of reaching any valuable conclusion. The most instructive feature of chart No. 6 is illustrated by the tendency which the high altitude has manifested in compensating for the low latitude in the effect upon the sugar content. To a great degree this is shown in the results at the Blacksburg Station, where an altitude of over 2,100 feet was sufficient to counteract the natural depressing effect of the latitude of the locality, bringing the sugar content of the beet at Blacksburg almost up to a mean position in the whole series. To a minor degree this influence is also shown at Lexington, where the high temperature, higher than that of Washington, combined with a high altitude, higher than that of any other station with the exception of Blacksburg, has tended to raise the sugar content of the beet so that it is only one-tenth of 1 per cent below that at Washington. At the Lexington Station, however, it must not be forgotten that there was another factor, viz, the small size of the beets, which tended to increase the percentage of sugar.

In order to bring the gene al results for the five stations completing the five years work into a more striking comparison, graphic charts have been constructed in which each of the three most important elements of the environment affecting composition is represented separately, and a fourth graphic chart has been added showing only the relations existing between the purity and sugar content. These charts (Nos. 7 to 10) are constructed on the following principles.