212 
The preceding table shows that the only important difference 
between these good and poor years consists in the fact that the latter 
have more rain and less sunshine. The influence of the tempera- 
ture of the air as such and of the number of days of freezing weather 
does not seem to be important, so that we must conclude that the 
cloudy weather which accompanies the rain and cuts off the sunshine, 
affects the plant unfavorably only by this loss of radiation. A defi- 
ciency of light is more unfavorable than excessive moisture in the 
soil. In general in France, and especially in dry countries such as the 
arid regions of America, it 1s the deficiency of water in the soil that 
affects the crops unfavorably. Where an abundance of sunshine 
exists the wheat plant can utilize more water than ordinary soils 
possess; hence the great advantage of irrigation, as long since prac- 
ticed in Mesopotamia, Egypt, China, India, Arizona, and South 
Africa. ‘The numerical data with regard to the quantity of water 
and the times of irrigation have been approximately determined at 
agricultural experiment stations, with results given in the next sec- 
tion of this present report. 
In the Annuaire for 1890 Marié-Davy gives climatic tables espe- 
cially adapted for phenological study. 
In order that meteorological data may be presented in form con- 
venient for the comparison of crop reports or for the prediction of 
the future development of the current crop or for other studies in 
the growth of plants it is necessary that the data should be compiled 
in a manner very different from that ordinarily given in climato- 
logical tables. The monthly means and other data given in the 
so-called international form recommended and urged by the recent 
international conferences of Europe have much more regard to 
dynamic meteorology and to questions in hygiene than to questions 
in agriculture. For our agricultural studies a continuous summa- 
tion must be made from the beginning to the end of the year, either 
by deeades, by weeks, by pentads, or even by days for each succes- 
sive year. From such tables we can calculate the total work that 
has been done upon the plant by the sunshine and the work that 
remains to be done before the harvest. Such tables can be compiled 
in an empirical approximate way from the data furnished by the 
international forms, as I have attempted to do in table—* But it 
is far better to prepare them from the original records, and they 
must be prepared for every agricultural experiment station in the 
United States before we can profitably study the influences of our 
a@This table is omitted in the present edition. 
