1911] TRANSEAU—APPARATUS FOR TRANSPIRATION  —_—57 
which dip into a small cup of mercury on the scale pan whenever a 
balance is established. The one gram weights are too heavy to 
obtain satisfactory records from many of the extreme xerophytes. 
For these plants I am using hollow brass balls standardized to 0.4 
gm. These are not as light as could be desired, but they are better 
than the gram weights. To be 
very satisfactory for comparative 
purposes, the interval between 
records should not exceed two 
hours. Where great differences 
exist between day and night 
rates, I have used the fractional 
weights at night and the gram 
weights during the day. 
THE IRRIGATORS.—Two points 
which became evident in the 
early experiments are that the 
water content of the soil of the 
plants to be compared must be 
essentially the same, and that the 
water content must be essentially 
the same throughout the experi- 
ment. The ordinary method of 
watering at 24-hour intervals did 
not give satisfactory results in 
some instances. In one experi- 
ment the ratio between two plants Hrs. ¢—-Eietails of the livigator, 
on successive days was reversed showing 
on account of differences in soi] nd connections. 
water content. To avoid errors 
of this kind the principle of irrigating plants by porous cups 
suggested by LivincstTon? was brought into use, and the apparatus 
shown in fig. 3 was constructed. It consists of a slender porous cup 
similar to those used in my vaporimeters.’ Thisis readily introduced 
into the soil of a 3-, 4-, or 5-inch pot by removing a core of soil with 
2 A method of controlling plant moisture. Plant World 11:39. 1908. 
3 A simple vaporimeter. Bot. GAZETTE 49:459. I910. 
porous cup, water reservoir, 
