substance is growing substance e. g., the water does not reproduce itself.

If then the increase of the body does not necessarily mean increase of the growing substance, there is no good reason for expressing the growth of any day in terms of the fraction of the initial weight which has been added during the day. If today all of the body substance is growing, an increase of 10% will not mean so high a rate of growth as an increment of 10 % in a body half of which is water. If we assume for the moment, in our ignorance of what part of the body is growable, that the dry substance is all capable of reproducing itself (and that is an overestimate), we get a somewhat different curve of percentage increments, as in Fig. 8b. The curve rises on the whole, but would have risen more rapidly if we had used only that fraction of dry substance which is growing. Here then we find no loss in the rate of growth of the growing substance, and in SO far a lack of confirmation of Minot's generalization.

One reason for the lack of agreement between the curve of percentage increments of the frog embryos and that of the guinea pig is due to the fact, that in our study of tadpoles we have been dealing with the beginnings of development, while in the guinea pigs, taken from birth on, we do not have the earliest stages. Doubtless a time will come in the life of the frog when it will not increase its weight a milligram a day; it will have ceased to grow. But are we justified in ascribing this cessation of growth to a gradual fading out of the growth force given by impregnation? We get some light upon this question when we go to plants and consider again the tip of the epicotyl. Here the growing tissue just behind the tip soon ceases to grow, while it undergoes histological differentiation. Is this cessation of growth due to the fading out of a growth force due to the remoteness of impregnation? It is clear that this explanation is unsatisfactory, for the protoplasm at the very tip of the epicotyl continues to grow during, it may be, years. There is no evidence that it ever loses its capacity for growth. It may go on developing for centuries, as we see in trees. It may go on growing far beyond the ordinary life of a tree. Thus, as we all know, there is in Cambridge an elm "under which Washington took command of the American army, July 4, 1775," and some cuttings have been made from this tree, which are growing quite as well as young trees that have presumably been derived from protoplasm more recently

impregnated. We see here, after the lapse of perhaps two centuries, no loss in the power of growth.

I would suggest that the reason why the animal ceases at last to grow is the same as the reason why the differentiated tissue below the tip of the epicotyl ceases to grow-not because there is a necessary limit to growth force at a certain distance from impregnation, but because it is advantageous to the species that the individual should cease to grow at this point. If it is not advantageous for it to cease to grow, it may go on for centuries like the tip of the plant.

To sum up, then, my studies lead me to recognize a general parallelism between the developmental processes occurring at the tip of a twig and in the animal embryo. In both there is first a period of rapid cell division with slow growth; next, a grand period of growth in which the general form of the embryo is acquired, the Anlagen of the organs are established, and the organism increases rapidly in size by imbibition of water; and, lastly, a period in which histological differentiation is carried on while the absolute growth increments cease to increase. Finally, the fact that increase in body substance is so largely due to a non-growing substance — water diminishes the value of the percentage increment curve of growth.

Printed, June, 1897.