change in shape, as indicated by the ratio of breadth to length; and, appropriate to the fourth proposition, that the American eggs are not indiscriminately distributed, but tend to gather about a mean type. This type is located on or near the ordinate of 73%, and is removed some little distance toward the side of sphericity, and away from the correlative ordinate (70%) of the British specimens.

The second curves, then, bring out in a more emphatic way the same general facts that were shown on the first diagram.

But it is quite evident that the mere ratio of breadth to length is not an adequate index of variation in shape. On this ratio alone, an egg that is conical, or pear-shaped, may not appear in any way different from one that is ellipsoidal or lemon-shaped. I have made several attempts to bring out these extreme variations in some practical arithmetical manner, but have felt each time that the eggs varied far more than the numerical results indicated.

For want of a better method, I finally adopted the following: Having placed upon each American egg a secret mark, the eggs of both countries were thoroughly mixed together in a single tray. A disinterested person was then requested to select, from the mixture of 1736 eggs, one hundred eggs which appeared to him to present extremes of shape-variation. If eggs from the two countries are equally variable, it is clear that approximately the same number from each would be selected; and, of course, if the American eggs are more variable, more American eggs would be selected. The result of this experiment was most striking, and in harmony with the evidence derived from the comparison of lengths and the ratios of breadth to length. Eighty-one of the selected eggs were American, while only nineteen were English; over four times as many of the former as of the latter.

As before mentioned, the colors of both European and American eggs are subject to variation, arising from modifications of the ground color and from the color and distribution. of the spots or blotches. Some are of a somber color, much like the eggs of our common song sparrow; others resemble the eggs of the kingbird; and still others have the delicate

ivory white of certain vireos. An attempt was made to arrange the colors in sequence, but after many fruitless efforts the plan of disinterested selection, above mentioned, was adopted.

The British and American eggs were thoroughly mixed together and the request was made that twenty-five eggs which presented the greatest variation toward the kingbird type should be selected first; then twenty-five of the somber type; third, twenty-five of extremely light color; and, fourth, twenty-five anomalous varieties. Some hours were spent in making the selection of one hundred eggs, and with the results indicated on Diagram III, where b represents the British eggs and A represents the American.

DIAGRAM III. - This diagram is designed to illustrate the preponderance of extreme color variation on the part of American eggs. A indicates American, b indicates British eggs.

Of the kingbird type and of the somber type there were over twice as many American as British eggs. There were among the light eggs nearly twelve times as many departures from the

mean of color on the part of the American as on the part of the British eggs, and among the anomalous eggs there were twenty-four times as many American extremes as British. (It may also be of interest to add that the single British egg was the last egg to be selected, that is, it presented the least departure from the mean of the twenty-five anomalous variations.)

Eighty-two of the examples of extreme color-variation were thus found to be American and eighteen British. That so large a proportion of extreme variation in color was found among the American eggs is interesting in itself, but a comparison with the relative amount of extreme variation in shape, enhances the significance of both results, for not only is the preponderance of variation among American eggs very obvious, but in both cases, in length and in shape, it is almost precisely the same (8119 in the first, 82:18 in the second).

Our data, then, whether it be gathered from comparisons of length, ratio of breadth to length, shape, or color, all point in one direction; and, granting that the sparrow since its introduction has been comparatively free from the action of natural selection, we may conclude that the predicted results of Panmixia have been realized.

The collection of a series of facts, for the mere support of some favorite theory, ought not to be the purpose of biological investigation. The relation that the facts may have to other facts and the bearing that they may have upon collateral theories should, at least, be indicated.

The following questions naturally arise:

Apart from the tendency to vary, is the new form, adopted by the American egg, the result of the selection of adaptive adventitious or fortuitous variations, or is it "determinate," the result of the direct action of a new environment? If due to the direct molding influence of a new environment, is the variation ontogenic, that is, does it occur anew and repeatedly in each successive generation, in obedience to reiterated environmental demands; or have the directive influences of the mechanism of heredity been so affected that the variation becomes

established as phylogenic? affected immediately, through the action of the new environment on the germ itself, or mediately, through the influence of ontogenic somatic change?

Is the mechanism of heredity

I think it improbable that the new form adopted by the American egg can be the result of the selection of adaptive fortuitous variations.

Fortuitous variation means chance-variation, and, although it is mathematically possible for the same particular variation to appear fortuitously in all or nearly all of the American eggs, it is absurd for us to suppose that this has actually happened. We cannot believe that the new form and shape, which are so universally presented by the American species, are variations which have arisen by mere chance. Again, even admitting for the sake of argument that a chance-variation has simultaneously appeared in nearly all the American individuals, what have we to show that this variation is adaptive, that it has selective value? Who will say that the shorter egg is a superior egg, or that the more spherical egg is, in the new environment, an improvement on the European type?

In the third place, even admitting the all-sufficiency of natural selection, there has not been sufficient time for the establishment of a new type of egg, that is, for the conclusion of the struggle between "Nature and Nurture." Neo-Darwinians deal with centuries and ages. Forty years can accomplish nothing.

If we again refer to the curves, we shall find other reasons for the belief that the American type of egg is not to be explained by the principle of adaptive fortuity.

Although the American eggs are unquestionably more variable, as is shown by the more elongated base lines, the curves rising to the culminating points of American variation are no less regular than those rising to the culminating points of British variation. This means that the new type is definitely established and that nearly all the eggs tend towards this type. Now, is it likely that mere chance-variation would yield an American curve so nearly parallel to the British curve? If the selection favors those eggs which are located on ordinates 21 and 73 (Diagrams I and II), that is, favors a certain type, why

do other eggs on distinct ordinates and of an entirely different type arrange themselves in an orderly manner?

This brings us to another point. The curves show that the British influence is still felt in America. There are distinct elevations in the American curves as they cross the ordinates of 22 and 70. These elevations, which may represent the

conservatism of certain individuals which still retain British. instincts, are perhaps of less interest than the elevations on the British curves which lie immediately under the American culminating points. One wonders why ruthless natural selection should have spared these particular individuals.

There has been a general reduction in the shape of practically all the eggs since the introduction of the birds into this country, and this reduction has taken place not only in the neighborhood of the new mean, but also at the extremes. Not only has the old culminating point been shifted, but the entire curve has been shifted. The larger eggs have become smaller, the medium eggs have become smaller, the smaller eggs have become smaller; and all the eggs, whether of the ellipsoidal or spheroidal type, have become more nearly spherical.

Concluding, then, that the evidence does not favor the view that the American egg is the result of the action of natural selection upon fortuitous variations, let us examine the alternative, that is, the variations are due to the molding influence of a new environment.

A new environment, offering new food, peculiar climatic conditions, etc., might affect a large number of individuals in certain peculiar and definite ways, and it is evident that the respective curves of variation given in Diagrams I and II are in harmony with such a conception of the march of transformation. It is, indeed, a phenomenon that is seemingly of the nature of a "mutation" (Scott, '94). This view, moreover, is not contrary to the later ideas of Darwin, who distinctly stated. that the greatest error which he had committed was in not allowing sufficient weight to the direct action of environment independent of natural selection.

Moreover, if the new environment is directly responsible for the new variations, the question of time is no longer a disturb