530 PROF. D. T. MACDOUGAL ON THE GROWTH OF 
(see figure, p.545). This method of sealing exerted no injurious 
pressure on the plant and allowed it to expand in a normal 
manner—a very important consideration in such experiments, 
where soft-stemmed herbaceous plants were used. The plant was 
covered with a bell-jar of a capacity of 4 to 8 litres, sealed to the 
glass plate, and provided with two tubulures. To one tubulure was 
fitted a series of potassium tubes or vessels containing potassium 
hydrate in solid form, and saturating a mass of asbestos fibre. 
The second tubulure was connected with an aspirator, by which 
the air was occasionally renewed. Several vessels containing 2 to 
5 grams of solid potassium hydrate were placed inside of the bell- 
jar. The potassium absorbed water rapidly and soon dissolved. 
To provide against the dryness of the enclosed air thus induced, 
a large sponge saturated with water was placed near the plant. 
These precautions furnish normal conditions, except in the com- 
position of the air, from which almost all the CO, is abstracted. It 
is, of course, understood, that the plant is constantly giving off this 
substance as a result of its oxidation processes, and it may be 
imagined as forming a diffuse stream from the plant to the vessels 
containing the potassium solutions. The amount actually present 
in the bell-jar at any time, however, must have been quite small. 
The potassium solutions were renewed once each week. 
Plants of Arisema grown in the apparatus described above 
exhibited a normal development during the opening of the bud 
and the preliminary stages of the unfolding of the leaves, which 
are crumpled in the bud, during a period of from 2 to 4 days. 
The unfolding process was arrested, however, at a very early 
stage, and the lamin: were unfolded only so far as to expose the 
dorsal surface, and the crumpled appearance was not lost. Jn 
10 to 14 days after the beginning of the experiment, the lamine 
assumed a yellowish colour, as a result of the decomposition of 
the chlorophyll, and other signs of deterioration appeared, ending 
in the death of the organ a few days later (see Plate XIX. fig. 2). 
The structure and arrangement of the tissues had undergone 
but little differentiation from the forms present in the folded 
condition, and differed from the normal forms by the size of the 
single layer of palisade-cells and the globular form of the 
spongy parenchyma, and seemed, moreover, to be in a state of 
hunger. It is to be noted that the sheathing spathe also under- 
goes similar abnormalities; but since it is never folded, and since 
its development consists principally of a longitudinal expansion 
m 
