4o8 



!. INAGAKI. 



side of the pot. (C) is an accumulator of the water that has overflown 

 from (A) after rain, wind, and diminution of atmospheric pressure. 

 The quantity of water collected in (C) is not to be neglected, as it 

 sometimes amounts to a factor of some importance. If we deduct 

 the quantity of rain taken from the ombrometer, from the quantity 

 of water collected in (C), the difference is exactly the quantity 

 of overflow due to wind or diminution of atmospheric pressure. 

 This quantity is also to be deducted from the measurement as made 

 by reading off the gauge. Then the remainder is the actual amount 

 of transpiration from the rice planted in the pot. 



In the course of the experiment made last year I became aware 

 of three points in which the apparatus could be improved, excluding 

 at the same time all sources of error and increasing its usefulness 

 for practical purposes. The following changes were therefore 

 made in the apparatus this year : — 



(i .) The side mouth (d) of the pot (A) was arranged, as shown in 

 Figure II, (e), so that the water level could be easily regulated by 

 means of a screw (f). By this new arrangemxnt, the inconveni- 

 ence arising from overflow was obviated, and the 3 screws for the 

 pot-stand could be dispensed with. 



(2.) A drain tube (g) w'as fitted to the pot (A), by means of which 

 about 250 c. cm. of water, the quantity to be deducted from the 

 measurement as made by reading off the gauge, is draw'n out from 

 the pot every day. This arrangement was made because in ordinary 

 rice fields there is a loss of about an equal quantity of water by trans- 

 mission through the soil, and this transmission seems to me to favour 

 the growth of the crop by avoiding the bad effect of stagnation. 



(3.) Several batches of rice were planted in the pot, so as to 

 secure as nearly as possible the conditions that obtain in the field. 



This arose from the consideration that in general the evaporation 

 from an isolated plant as (A) in Figure III is always larger than that 

 of a j^lant surrounded by other plants as (B), since in (A) there are 

 4 open sides {a, P, y, o) besides the upper surface (a), whereas in (B) all 

 the 4 sides are mostly covered in by the surrounding plants. Hence 

 we can not obtain the transpiration-water for 9 plants, such as are 

 shown in (B) by simply multiplying the quantity of transpiration from 

 one isolated plant by the number of plants in the batch. It has 

 long been a mystery to scientists that the quantity of evaporation 

 from a certain area of a cultivated field during the period of vegeta- 

 tion always exceeds very much the quantity of rain that falls down 



