The soil in a flower pot is liable to be disturbed by irrigation, and the record thus vitiated by physical disturbance. This is obviated by wrapping a piece of cloth round the root imbedded in a small quantity of soil. The lower end of the plant is held securely by a clamp. In order to subject the plant to the action of gases and vapours, or to variation of temperature it is enclosed in a glass cylinder (V) with an inlet and an outlet pipe (Fig. 58). The chamber is maintained in a humid condition by means of a sponge soaked in water. Different gases, warm or cold water vapours, may thus be introduced into the plant chamber. Any quick growing organ of a plant will be found suitable for experiment. In order to avoid all possible disturbing action of circumnutation, it is preferable to employ either radial organs, such as flower peduncles and buds of certain flowers, or the limp leaves of various species of grasses, and the pistils of flowers. It is also advisable to select specimens in which the growth is uniform. I append a representative list of various specimens in which, under favourable conditions of season and temperature, the rates of growth may be as high as those given below:-- Peduncle of _Zephyranthes_ 0.7 mm. per hour. Leaf of grass 1.10 " " " Pistil of _Hibiscus_ flower 1.20 " " " Seedling of wheat 1.60 " " " Flower bud of _Crinum_ 2.20 " " " Seedling of _Scirpus Kysoor_ 3.00 " " " The specimen employed for experiment may be an intact plant, rooted in a flower pot. It is, however, more convenient to employ cut specimens, the exposed end being wrapped in moist cloth. The shock-effect of section passes off after several hours, and the isolated organ renews its growth in a normal manner. Among various specimens I find _S. Kysoor_ to be very suitable for experiments on growth. The leaves are much stronger than those of wheat and different grasses, and can bear a considerable amount of pull without harm. Its rate of growth under favourable condition of season is considerable. Some specimens were found to have grown more than 8 cm. in the course of twenty-four hours, or more than 3 mm. per hour. This was during the rainy season in the month of August. But a month later the rate of growth fell to about 1 mm. per hour. I shall now proceed to describe certain typical experiments which will show: (1) the extreme sensibility of the Crescograph; (2) its wide applicability in different investigations; and (3) its capability in determining with great precision the time-relations of responsive changes in the rate of growth. In describing these typical cases, I shall give detailed account of the experimental methods employed, and thus avoid repetition in accounts of subsequent experiments.