=192. Methods.= All field experiments in evolution are based upon a change of habitat. The latter is accomplished by the modification of the habitat itself, or by the transfer of the species to one or more different habitats, or to different areas of the same habitat. In both cases the choice of habitats is made upon the basis of efficient differences of water-content or light. Saline situations do not constitute an exception, since the chresard is really the effective stimulus. Cultures at different altitudes, which afford striking results, appear to concern several factors, but in the final analysis, water-content and humidity are alone found to be really formative. Cultures may furthermore be distinguished as simple or reciprocal. Simple cultures are those in which a species is transferred to one or more habitats, or in which a habitat is modified in one or more ways. Reciprocal cultures are possible only with polydemic species, or with endemics after ecads have been produced by experiment. Modification or transfer is made in the usual way, but reciprocally, i. e., the original form is transferred to the habitat of the ecad, and the latter to the habitat of the former; or the shade in which some individuals of the ecad are growing may be destroyed, and at the same time individuals of the type may be shaded. Both transfer and modification may be applied to the same species, but since the same measured change of factor can be obtained in either way, the use of both is undesirable, with the exception of the rare cases where they serve as checks upon each other. The transfer of a seed or plant is so much simpler and more convenient that this method is the one regularly used. It sometimes happens, however, that a change of water-content or light intensity is readily and conveniently made, and is desirable for other reasons. It is evident that both transfer and modification require that the factor records of the various habitats or areas be as full as possible, at least so far as water-content, humidity, and light are concerned. In the case of the areas that are to be modified, these factors are determined before the change is made. Afterward they are read from time to time during the growing season, and are also checked by readings made near at hand in the unmodified formation. The readings made in the beginning should correspond closely to the check readings, but in case of disagreement the latter are to be taken as conclusive.