=190. Method of record.= As suggested elsewhere, there are four important kinds of records, which should be made for natural experiments, and likewise for habitat and control cultures. These are exsiccati, photographs, biometrical formulae and curves, and histological sections. These serve not merely as records of what has taken place, but they also make it possible to trace the course of evolution through a long period with an accuracy otherwise impossible, and even to foreshadow the changes which will occur in the future. The possibility of doing this depends primarily upon the completeness of the record, and for this reason the four methods indicated should be used conjointly. In the case of ecads and mutants, exsiccati, photographs, and sections are the most valuable, and in the majority of cases are sufficient, since both ecads and mutants bear a more distinctive impress than variants do. On the other hand, since variations are more minute, the determination of the mean and extreme of variation by biometrical methods is almost a prerequisite to the use of the other three methods, which must necessarily be applied to representative individuals. Exsiccati and photographs are made in the usual way for plants, but it is an advantage to photograph each ancestral form alongside of its proper ecads, mutants, or variants, in addition to making detail pictures of each form and of the organs which show modification. In the collection of material for histological sections, which deal primarily with the leaf or with stems in the case of plants with reduced leaves, a few simple precautions have been found necessary. Whenever possible, material should be killed where it is collected, since in this way the chloroplasts are fixed in their normal position. In case leaves that can not be replaced easily have become wilted, an immersion of 5–6 hours in water will make it possible to kill them without shrinkage. In selecting leaves, great pains must be taken to collect only mature leaves. When the plants have a basal rosette, or distinct radical leaves, mature leaves are taken from both stem and base. In all cases where the two surfaces of the leaf can not be readily distinguished, the upper one is clearly marked. METHOD OF HABITAT CULTURES