impossible for the male to plant these "water eggs," as the author calls them, on its back. Therefore the early stages developed in the water as is the case with other amphibians. The habit of mating in the water became fixed, and persisted after the removal of the artificial conditions of temperature. The eggs, meantime, at each successive laying became smaller and smaller through the loss of yolk. The larvæ hatched at a much younger stage than the larvæ from normal eggs. The adults reared from the water eggs mated in the water at the first breeding season, even under normal conditions of temperature. Succeeding generations showed intensification of the new characters in the decrease of yolk, and also in the development of more gills, which changed in number from one pair to three pairs. There was, therefore, as in the preceding case, an apparently progressive adaptation to environment through the inheritance of acquired characters.
The effect of this change on the germ plasm was tested by a cross between the old type and the derived type. The new character, as judged by the instinct for mating in the water, behaved like a dominant Mendelian factor. Dominance, however, was of an unusual kind. The male, whether of the old or the new form, impressed its character on all the offspring of the first generation, but the second filial generation showed the usual kind of segregation of characters in the ratio of three individuals of the dominant form to one of the recessive. Clearly the unexpected feature in the behavior of the factors in this crossing lies in the peculiarity of the sex-limited potency, not in the isegregation of factors. The most interesting fact in the experiment is the attempt to prove a change in the germ plasm by the modern method of applying the test of cross breeding.
Another series of experiments was tried on the tree frog, Hyla. This frog lays its eggs in the water in bunches of 800-1,000, enveloped by the usual coats of gelatine. A number of frogs were kept away from the water, but were allowed to crawl about on a water plant which held small amounts of moisture in the bases of its leaves. During the mating season the frogs deposited their eggs in the moisture on the leaves, according to a habit which is common among some of the tropical representatives of the genus. The young remained in their envelopes until the gills had become enclosed, whereas the young under normal conditions begin a free swimming life before the gills appear. A new type of adult was produced marked by its small size. These dwarfs when reaching maturity laid their eggs in water after the usual manner. The new habit was not inherited. The offspring of the dwarf frogs, however, had external gills at the time of hatching, a stage half way between the old and the derived type, and, moreover, they grew into adults of a size half way between the two types. This experiment, therefore, showed results very similar to those shown by the experiment on Alytes.
