would be susceptible to modification in the stage between the reduction divisions and fertilization.

This conclusion is borne out by my own results in which solutions were introduced into the ovary during this stage. The extent of the treatments, together with the diversity of results, makes possible an analysis of other features. Thus the induction of more than one form by the use of a single reagent suggests either that different chromatin elements were affected in the separate ovular reactions, or that unlike parts of the chains of catalytic action were interrupted or disturbed by the introduced substances. Some of the compounds used are inimical to enzymatic action, or may be capable of a negatively catalytic effect, or might indeed set up unusual splitting processes, a state of affairs distinctly favorable to the last named alternative.

Not only may irradiation and the introduction of unusual substances occur naturally to the modification of heredity in plants, but the climatic factors may, as in Tower's experiments, exercise an influence upon the reaction velocities of various parts of the metabolic series, or by variations in humidity, regulate the excretion or retention of active substances. All of these possibilities must be taken into account in attempts at explanation of bud-sports or bud-mutations in plants. It is to be seen that either egg or sperm may be affected by experimental agencies, and that the results do not differ in quality or degree. Gager's

atypic forms were obtained by the treatment of pollen; my own from ovarial injections which might have acted upon the egg, or sperm, and Tower's work was with both.

The new forms of beetles and plants which have thus been called into existence sustain the following general relations to their environment and to the strains from which they were derived:

1. Some of the species dealt with were growing in the open, and domesticated forms were not included in the experiments.

2. The newly arisen or modified characters maintained their distinctive appearance when crossed with the parental strains; in some no reversions have yet been shown.

3. Discontinuous departures induced by ovarial treatments in plants were full and constant within the limits of fluctuability with the first generation. Similar abruptness of divergence is exhibited by beetles in some cases, while in others more than one generation, after removal from experimental conditions, was necessary to secure full expression.

4. Many aberrations induced by irradiation in plants and by climatic effects in beetles were of the nature of closely continuous variations, the range of which was widened by the exciting agent. Some of the derivatives of Penstemon may prove to be of this character. The single derivative of Oenothera biennis obtained by ova

rial treatment with zinc sulphate is distinctly discontinuous with the parent.

5. Some of the modifications may be regarded as an increase of capacities already present; some imply the loss of characters or structures, and some are acquisitions; in more than one instance qualities new to the genus have been taken on. Changes such as the mottling of a solidly colored flower may be regarded as a loss of a portion of a design, the total effect of which was a shaded or a self color, or it may be taken as a differentiation in advance.

6. The behavior of the newly derived forms when subjected to natural conditions, competition, and possibility of hybridization with parental forms, has been extremely diverse. Some of the beetles have been swamped by hybridization with the parental form; others have displayed some power of endurance. The plant derivatives induced by ovarial treatments were weaker than the parent in some localities, and more enduring in others. The derivative of Oenothera biennis induced by a zinc sulphate ovarial treatment is less adapted to xerophytic conditions than the parent, does not readily hybridize with it when grown in contact, and its earlier characters appear to be dominant when crosses are made artificially.

7. The departures obtained by the experimental manipulation of external exciting agencies bear a general similarity to the initiatory

142 THE INFLUENCE OF ENVIRONMENT

and modificational phenomena exhibited by organisms in a state of nature, and it seems justifiable to conclude that the processes disturbed or set in motion are identical with some of those concerned in the main evolutionary development of organisms.

THE BEHAVIOR OF UNIT CHARAC

TERS IN HEREDITY

BY

W. E. CASTLE

NO ONE recognizes more frankly and joyously than would Darwin, were he here to-day, the great advance which has been made in our knowledge of heredity since his time. His work and writings have pointed the way to that advance, and it is largely owing to a return to the experimental method of testing hypotheses, which Darwin used so successfully, that the remarkable progress of the last decade has been made possible. We do, therefore, the greatest honor to Darwin if we pause to consider what superstructure of knowledge has been built on the foundations which he laid. This superstructure is, indeed, still in the building, and it is not easy in all cases to distinguish between the solid structure of proved fact and the scaffolding of hypothesis. Still, the attempt should be made, and it will give us encouragement to discover that, notwithstanding the considerable amount of rubbish lying about, there is, nevertheless, good constructive work going on here which gives promise of permanency.