20 
of atmospheric air. To accomplish this purpose 
without injury to the plant, we caused several cir^ 
cular tin dishes to be made, some of which were a-* 
bout six inches in diameter, and one and a half in 
depth, and others of little more than half these di- 
mensions. Each dish was divided through its mid^ 
die into two parts, and exactly in the centre of each 
part a semicircular cut was made, so that, when the 
two parts were brought together, a circular hole was 
formed, just large enough to admit the stem of the 
plant. One of the parts was made a little smaller than 
the other, so as to pass about an inch within it ; and 
to the bottom of this smaller part a flat piece of tin was 
partially fastened, in such a manner as to receive the 
bottom of the larger half in a sort of sheath, by 
which the two parts were rendered perfectly steady. 
As the plant grew in the pot, the two parts of the 
dish, previously separated, were brought together, 
one on each side the stem, and made to slide into 
each other without injury to the stem, which rose 
through the central opening,- while the circular dish 
itself now rested securely on the margin of the pot. 
By this arrangement, the plant was left growing in 
its natural situation, and all sources of fallacy from 
the mould were effectually removed. The junction 
of the two parts, and the aperture which received 
the stem of the plant, were now made water-tight by 
a proper luting, and the jar was then inverted over 
the plant into the tin dish, which was now filled with 
water. 
227. One of these dishes was made to inclose a 
young bean as it grew in a pot, to the height of three 
