211 
Ten to twelve kilos of soil, representing* approximately one 
fifty-thousandth acre foot, are used as the area unit, as this is 
approximately the area allotted to a clump of rice under field 
conditions. The soil is treated before it is put in the pot, the 
optimum moisture is supplied, and the rice seedlings are trans- 
planted at the proper time. If it is intended to weigh the pots 
during the growth of the plant, the weight of all the pots is first 
made uniform by means of the gravel in the bottom. All experi- 
ments are carried On in triplicate, and the uniformity of growth 
observed by the writer in hundreds of such cultures would indi- 
cate that the average is affected only infinitesimally by indi- 
vidual discrepancies. The cultures are grown in the open and 
under glass, and the results obtained seem to be very satisfactory. 
A galvanized iron pot of the same dimensions as the porcelain 
one is also used and some of these have been in service for ten 
years. They are, however, not considered sufficiently reliable 
where acids constitute part of the fertilizer. 
In addition to the above method, cultures are maintained in 
the field, in metal and earthen-ware cylinders which are sunk into 
the rice paddies. These cylinders are three feet in diameter and 
three feet in depth. Four inches are left projecting above the 
ground, and at a level with the ground surface there are several 
inch-wide holes to admit irrigation water when the paddy field 
is flooded. Five clumps of rice are transplanted to each of these 
cylinders. This method also gives very satisfactory results. An 
earlier, and somewhat cruder, method of growing cultures in the 
field was to sink a wooden frame three feet square into the paddy 
field, in which nine clumps of rice were planted. Better control 
and economy of space are the advantages claimed for this type 
of field cultures. 
The accompanying photographs give some idea of the ar- 
rangement and extent of these cultures. Field plot experiments 
with fertilizers are maintained only for demonstration purposes, 
when the plots cover several square rods. 
The adoption of pot and cylinder cultures in our work is 
strongly recommended for their accuracy and ease of use in 
carrying on a large number of experiments. 
Beginning in 1889, Kellner and his associates conducted a 
series of experiments to determine the exhaustion of the nutrient 
material in the soil by successive croppings, and the amount of 
artificial fertilization necessary to maintain fertility. These ex- 
periments extended over a period of six years. A detailed re- 
port of them will be found in Bulletins of the College of Agri- 
culture, Vol. I, 8-11, and Vol. Ill, pp. 371-406. All of these 
bulletins are now out of print, except Bulletin 11, which sum- 
marizes the first three years’ work. 
Taking the average of the results of a great many analyses, it 
was found that an ordinary crop, yielding 2,500 pounds of rice, 
removes from an acre of soil about 26 pounds of nitrogen, 16 
