MANURING EXPERIMENTS WITH PADDY RICE. 5 
as soluble phosphates when applied as manures are not liable 
to be extracted in noticeable quantities by the usual irrigation. 
The water used far the irrigation of our experimental plots 
came from a spring situated about 500 metres from the field, 
and was guided through a ditch into a reservoir near the plots. 
It did not supply our plants with any remarkable amaunt of 
nutrients, as it is of exceptional purity. Analyses of this water 
before and after it had passed over a plot of about 400 square 
metres, made in 1882 in conjunction with J. Sawano, have 
shown 5 that it contains in 100 litres in grammes : 
Influent 
water. 
Effluent 
water. 
Difference. 
Total dry matter. 
4.500 
3.900 
—0.600 
Silica. 
I.291 
1-343 
+0.052 
Alumina, Ferrous and Ferric oxide .. 
0.666 
0-355 
—0.311 
Lime. 
0.782 
0.273 
+ 0 091 
Magnesia. 
0'493 
0.370 
—0.123 
Potash and Soda (as chlorides). 
1-103 
1.016 
—0.087 
Phosphoric acid. 
O 
0 
— 
Sulphuric acid. 
0.054 
0.059 
+0.005 
Chlorine. 
0.088 
0.113 
+0.025 
Nitric acid. 
O 
bo 
00 
0.430 
-0.457 
Ammonia. 
0.063 
0.076 
+0.013 
These figures do not, of course, exactly indicate how much 
of each ingredient is actually retained by the soil or assimilated 
by the plants, because the quantity of the water while it is 
slowly running over the field, is diminished by evaporation or 
increased by rain, whicn processes certainly do not quite com¬ 
pensate one another. The difference between the incoming and 
exit waters is, however, so trifling that in our case manuring ex¬ 
periments are not disturbed by the minute quantities of vegetable 
nutrients supplied or withdrawn by the water of irrigation. 
The experiments were carried out in square frames which we 
have successfully used on dry land since 1885. Each frame 
surrounding an area of g square shaku (=0.82645 square 
5 Landw. Versuchsstationen, 1883, vol. 30, p. 39. 
