regenerate can be shown by cutting from the side of the gastrocnemius a square piece of muscle tissue. In the course of a few months I have found that the muscle regains its size, which seems to be due, in part at least, to the formation of new muscle, although hypertrophy of the remaining fibers may also assist in the enlarge

ment.

There is some indication that the delay in the formation of the new leg in the frog is due to conditions existing in the bones or muscles and, as I have pointed out,1 it is significant to find that in the vertebrates the loss of power to regenerate a limb appears where cartilage has been changed to bone. The result is not however due directly to the ossification, since the new material is derived from the periosteum and not from differentiated tissue.

Especially interesting is the evidence showing that the introduction of material, itself capable of regeneration (as when the tail-tissue of the lizard is introduced into the leg-pocket) does not incite the leg to regenerate. If the process of regeneration is due to some enzyme, or other substance of this nature, that arises in an injured. region, and whose presence incites the new growth, we might hope by introducing pieces of material capable of forming such substances to incite regeneration, but no such result followed. It would be unwise to lay too much weight on negative evidence of this kind, but the results as they stand indicate, perhaps, with some probability, that the primary cause of regeneration is not to be found in this direction.

Finally, to revert once more to the experiments that gave positive results. It has been shown that from the proximal end of a reversed femur new limb bones develop. This result calls for further analysis. It is clear that each level of the limb has the power to regenerate all of the parts lying more distal to it, and in all probaThe Harvey Lectures for 1906-7.

bility every level has potentially the power also to regenerate all the other parts of the limb proximal to that level. It is difficult to show that a distal part has the potentiality to produce more proximal parts, but the facts make this interpretation highly probable. How much of the distal end regenerates depends in part on its relation to what is left in the stump, and in part on the necessity of forming a distal structure. Between these limits the intermediate parts are laid down. The proximal cut end of a limb must have the same potentiality of forming distal structures as has a distal end and in those cases where the possibility exists of forming either an anterior or a posterior structure, as in pieces of lumbriculus, for example, some other relation must determine that from one end of a piece a head always develops, and from the other end a tail. I have suggested that the direction of the gradations of the old material (as expressed in their differentiation) is the factor that regulates this result. If we apply these same ideas to the special case under consideration we might expect the proximal end of the leg (or any part of it) to regenerate only proximal structures; in other words to complete the proximal end of the femur and produce a scapula at the exposed end of the leg, and, theoretically, one might imagine the further development of a salamander around the scapula as a center. The facts are the reverse. The conditions that determine in the case of the reversed femur what shall regenerate of the various possible ones are not so simple as just described. In the first place, the detachment of the femur from the rest of the limb may soon lead to changes in it that cause it to lose that gradation of materials on which the polarity of the new part depends. There is also the possibility that the polarity of the other tissues may have a counterbalancing influence. But far outweighing these possibilities there is another consideration of greater weight. The special group of tissues found in such an organ as a limb may be capable of forming only one

structure, if they form anything at all, namely, a leg, and not a salamander to take the extreme case. But why always the distal end of the leg and not the proximal, i. e., not femur and scapula? The determination of the distal end rather than the proximal must be due, I think, to the presence of the free rounded knob covered by the new skin which gives the stimulus for a distal structure, and the foot end of the leg is the only possible distal structure that exists for this organ.

The case is parallel to the formation of a heteromorphic tail in the earthworm, that develops, as I have shown, from the anterior end of a piece when cut behind the level of the twentieth segment, or thereabouts. Here also a distal structure develops, but the nature of the material is such that a tail rather than a head regenerates. While polarity, as an expression of the gradation of the materials, is one of the factors that determines a result, it is not the exclusive factor. In Lumbriculus a head forms at the anterior end of a piece at nearly all levels and a tail at the posterior end. Here we must assume that the kinds of materials are so equally balanced throughout the greater length of the worm that the polarity determines the result, while in the earthworm and in the leg of the salamander another condition determines a different result. In the latter cases the kind of material, or the organ-complex, makes it possible for only a tail or a leg to develop at either end of the piece, and the presence of a free end determines that its new structure must be the terminal part, hence a foot in the case of the salamander, and a tail in the case of the earthworm, is regenerated even from a reversed end.

THE PHENOGAMOUS PARASITES

BY DR. CHARLES. A. WHITE

SMITHSONIAN INSTITUTION

THE object of this essay is to describe in a popular manner the chief characteristics of the known kinds or groups of phenogamous parasites, to show their relation to one another and to normal phenogams, and to discuss their structure and habits with reference to the probable manner of their origination. In order to make a popular statement of the characteristics of each group of these abnormal plants and to discuss them clearly it is first necessary to summarize briefly the elemental structure and physiological characteristics of the normal phenogams. I have chosen to do this in verbal terms a part of which are somewhat unusual, but which are believed to be specially appropriate to discussions of this kind.

The elemental parts of a normal phenogamous plant are root, stem and leaves, the beginning of the differentiation of which structures is distinguishable even in the embryo; and to these are added, at the maturity of the plant, flowers and fruit. Every normal phenogam also consists of two incremental parts, an up-growing and a down-growing part, respectively, the latter entering the soil to form the roots. The normal phenogamous plant performs all its physiological functions within, and for, itself and lives independently of all other plants except in the matter of competition with them for the benefits of soil, moisture and sunlight, but the parasites escape the performance of those functions so far as nutrition is concerned. The normal plants derive the materials for their subsistence and growth from inorganic sources and elaborate them within their own tissues for their own use, producing thereby their new organic substance, but the parasites rob other plants of that substance in its elab

orated condition. The supply of inorganic material is obtained by normal plants partly in a soluble and partly in a gaseous condition, the former being contained in the food-sap which the roots derive from the soil, and the latter in the atmosphere which surrounds the plant. The function of the root requires a constant accession of moisture, and that function is vital with relation to the other functions of the plant presently to be referred to. The action of sunlight is indispensable in the condensation and elaboration of those inorganic materials into new organic substance. An essential step in that elaboration of new material is the production of chlorophyl, which takes place partly in the bark of the growing branches, but mainly in the parenchyma of the leaves. Fully developed green leaves are therefore among the chief organs of normal phenogams, and their absence from the greater part of phenogamous parasites is due to the inability of those plants to produce chlorophyl. It is for these reasons that chlorophyl is so frequently mentioned in the following paragraphs.

The reproduction of normal phenogams is by two methods, namely, parturital1 and blastemal.2 These methods have such relevancy to the subject in hand that it will be frequently necessary to refer to them. The first is the conjugative method and provides for the hereditary transmission of specific and other systematic characters, the geographical distribution of species and the multiplication of individual plants. It is periodically cyclic, the maturation of the seed ending one cycle and the germination of its embryo beginning another. The second is the autogenous method and pertains to the growth and preservation of the individual plant. Its operation is physically continuous during the whole life of the plant, and every bud of the plant is connected with all the other buds by living somatic cells. The horticultural processes of budding and grafting consist of transferring blastemal

2

1 Parturio, to bring forth young.

Blaσros, a bud. These terms are regarded as preferable to "sexual, "asexual,'' ''vegetative," etc., which are often used by writers.