Data, by pentads, for Montsouris, 1889, as published by Descroix or specially computed—Continued.

THE RESULTS OF RECENT STUDIES BY ANGOT.

In 1880 the Central Meteorological Bureau of France, under the minister of public instruction, organized a system of phenological observations; the resulting data have been studied by Angot in a series of memoirs.

In his first paper (1882) Angot grouped the dates of the wheat harvest as observed during 1880 and 1881 at several hundred stations in France in groups of four or five stations and plotted these upon maps showing the elevations of the stations. By a careful comparison of neighboring stations he shows that the date of the harvest is everywhere quite uniformly retarded with increase of elevation, and at the rate of four days in time for every hundred meters of ascent. Apparently this retardation is the general result of a complex system of influences in which rainfall, drainage, soil, sunshine, temperature, and other local peculiarities combine. It is evident that the special influence of any local climate on the crop can not be successfully studied until the observations have been corrected for the general influence of elevation. He therefore reduces all the dates of harvest to sea level by applying the preceding correction.

A similar calculation showed him that the phenomena of flowering are also retarded at precisely the same rate of four days per 100 meters of elevation and these dates also are thus reducible to sea level.

Angot's charts, showing the dates of flowering and harvesting thus reduced to sea level, show great regularity and the isanthesic lines show the perfect regularity with which the reduced epoch of flowering begins in southern France on the 11th of May and advances northward until it reaches the northern boundary of France on the 25th of June; in a similar way the harvesting of winter wheat begins in southern France on the 10th of June (reduced epoch) and in northern France on the 9th of August. The variations of these isanthesic lines from year to year may be compared with the ordinary charts of temperature reduced to sea level or with other meteorological data in a very simple manner.

Angot has modified and apparently improved the methods of determining the influence of temperature on the date of flowering and harvesting. He says that since 1837 Boussingault's idea that the ripening demands a certain sum total of heat, which is constant for each species of plant, has been generally adopted. At first this sum total was calculated by adding together all mean daily temperatures from the germination of the seed or the beginning of vegetation after rejecting such means as were below freezing point. Then, as C. H. Martins, De Gasparin, and A. de Candolle had shown

that the temperatures useful to the plant vary with the species and are decidedly above freezing, therefore students have taken other limits. Thus Gasparin and Hervé Mangon adopt 6° C. for the initial temperature in the growth of wheat. In order to ascertain the proper method of counting temperatures Angot has accomplished the labor of prosecuting three parallel computations by three different methods, as follows:

(A) First method. By observations of daily maximum and minimum temperatures. In this method Angot has examined separately the observed maxima and minima of the thermometer in the shade. After rejecting all observations below 6° C., he subtracts 6° C. from all the others and takes the separate sums of the remaining maxima and minima for each month and then the average of these two sums, which consequently represents a sum total of heat received during the month in excess of 6° C.

(B) Second method.-By the daily means. In this method the mean of each day is first computed by taking the average of the maximum and minimum; 6° C. is then subtracted from each of these daily means and all negative remainders are rejected. The sum of the positive remainders represents the sum total of heat received in excess of 6° C.

(C) Third method. By maximum temperatures alone. In this method, which is a modification of that proposed by Hoffmann, a maximum thermometer is exposed to the direct rays of the sun and the sum total of the maximum temperatures is used by Hoffmann. But Angot prefers to use the maximum thermometer in the shade, as in the first method, and, as before, takes the sum total of all the positive remainders after subtracting 6° C.

In all these methods the principal difficulty is to fix the epoch from which the summation should begin. Sometimes the date of sowing has been adopted as this epoch, but from the date of sowing up to the date of sprouting the seed and the young plant are subjected only to the temperature of the soil, and not to that of the air, which often differ considerably. It would perhaps be better to start with the date at which the plant appears above the earth, but the date of sprouting is not generally given by observers. He therefore provisionally adopts the 1st of December as the point of departure and calculates the sum total of temperatures for the nine stations in France for which the dates of flowering and harvesting of winter wheat have been best determined for the years 1880 and 1881. The agreement among themselves of the numbers calculated by these three methods for nine stations and two different years is such that no decision can be arrived at as to which method is the best, and such decision is reserved for a future study of other harvests.