For obtaining a general idea of the sensitiveness, the absolute of growth in the instance given above was 0·00018 inch per second, and the Balanced Crescograph was shown capable of discriminating a variation of 1 part in 27,000; hence it is possible to detect by this means a variation of 1/1,500 millionth of an inch per second. This method of unprecedented delicacy opens out a new field of investigation on the effect of changes of environment in modification of growth; instances of this will be found in subsequent chapters. I give below accounts of certain demonstrations which will no doubt appear as very striking. After obtaining the exact balance a match was struck in the neighbourhood of the plant. This produced a marked movement of the hitherto quiescent spot of light, thus indicating the perception of such an extremely feeble stimulus by the plant. Breathing on the plant causes an enhancement of growth due to the joint effects of warmth and carbonic acid gas. A more striking experiment is to fill a small jar with carbonic acid and empty it over the plant. A violent movement of the spot of light to the right demonstrates the stimulating effect of this gas on growth. The method described above is excessively sensitive; for general purposes and for the method of direct record, a less sensitive arrangement is sufficient. I give below accounts of several typical experiments in which the recording form of Crescograph was employed, the magnification being only 2,000 times. [Illustration: FIG. 96.--Record showing the effect of CO_{2}. Horizontal line at beginning indicates balanced growth. Application of CO_{2} at arrow induces enhancement of growth shown by the up-curve followed by depression, shown by the down-curve. Successive dots at intervals of 10 seconds. (Seedling of wheat.)]