Theory of Selection.—Positive evidence of the regeneration of parts easily lost and therefore often found in nature as minia ture structures was obtained by simple experiment. Unfortunate ly, before extensive research in this direction had been carried out the interest of biologists passed to the theory of evolution. Starting with the idea that useful characters would be preserved by natural selection, it was held that all characters now observed in animals were due to this agency. It was cited, as proof of this thesis, that species much exposed to the attacks of enemies and easily injured readily regenerated lost parts. Some authors tried to establish that regeneration was restricted to these cases. They observed that birds with beaks liable to get injured in use, e.g., parrots and woodpeckers in pecking hard material, storks and fighting-cocks in their mating combats, were able to regenerate the injured beak. On the other hand, from a single instance of an old goose which was not able to restore the broken-off tip of its bill, it was rashly inferred that beaks like those of waterfowl, used only for handling soft substances, lacked the power of regen eration. This conclusion has not been borne out by subsequent experiments on young geese and ducks, which can regenerate the proper form of the bill when the tip is cut off. It has further been shown that homologous parts of the jaw of lizards regenerate, although in contrast to the tail they are strongly fixed to the body and scarcely ever injured. Other instances of correlation between liability to injury and capacity of regeneration have been sought in amphibians. We have already mentioned that common newts are capable of regenerating their legs. This capacity has been ascribed to the intervention of their foes. Now in some caves certain blind forms of tailed amphibians are to be found without any rapacious animals being present which would be able to prey upon them. Although former experiments had shown that the power of regeneration is possessed by this cave-dwelling Proteus, their inability to restore lost members has been upheld on the plea that they are not found in nature with regenerated legs or tails. But repetition of the experiments has proved their validity; and the reason for the absence of regenerates in nature is the low temperature of the caves and their proneness to infection, when the amputation has not been made under sterile conditions. If the theory of the acquisition of regenerative capacities by natural selection were correct, one would infer that rudimentary organs, whose regeneration would be of no value to their bearer, would have lost this potency. But here again the fact that her mit crabs regenerate their rudimentary abdominal legs contra dicts the theory. Nevertheless, the fascination of this view is so strong that it repeatedly crops up and is tested anew. Whereas the external gills of newts are exposed to injury, the internal gills of frog-tadpoles are well protected ; experiments have recently been carried out in expectation that they would prove devoid of regenerative capacity. The experimenter was disappointed by the prompt restoration of the excised gill-filaments. Many organs being internal in all animals were supposed not to regenerate : the lungs, for instance. The first experiments on their excision in newts being negative, this has been taken as corroborative evi dence. Later research on a larger scale has shown that partially extirpated lungs of salamanders and many other internal organs regenerate, provided that a small piece of the tissue is left in place. The continued discrepancy between facts and expectation made it necessary to look for some other explanation of the distribution of regenerative faculties than natural selection.
theless, there are certain restrictions as to the parts that are able to restore the animal as a whole. Passing this by for the moment, the next two groups in the scale of the animal world, the echinoderms and the worms, comprise species still capable of multiplication by fission, the several parts completing themselves by regeneration. As examples, the starfish Linkia, which regen erates from a piece of one arm, and the freshwater worms Lubri culus, Nais and Stenostomum may be cited. Many of the ascidians, too, are capable of complete regeneration. All other types show a great restriction in this regard, yet in the molluscs, arthropods and lower vertebrates the appendages and sense-organs can as a rule be restored, if not lost at a late stage in life. Thus the arms of squids, the eye-bearing tentacles of snails, the legs of spiders, all the appendages of crustacea and insects and the limbs of fish and amphibians grow again after injury. In the reptiles, further restrictions occur, the regeneration of the tail and jaw being the only ones of importance. Broken-off legs are never replaced by new ones, although examples have been found in nature with a tail-like appendage arising from the stump of a leg. In warm blooded vertebrates amputated tails or limbs are incapable of producing more than a short conical projection. The most achieved in the way of remodelling in mammals has been found in the tail of rodents, e.g., dormice, the broken vertebra growing to form a tip which by a tuft of hair simulates the normal tail end. From this brief summary three conclusions can be drawn : first that with ascent in the scale of animal types, regenerative capacities are more and more curtailed; secondly, that regenera tion is an original and general faculty of living matter, becoming lost by secondary complications in the course of evolution; and thirdly that natural selection has been impotent even to maintain the power of restoring lost parts shown by low types. Therefore, correlation of regenerative faculty with the probability of loss or the survival of the fittest must prove futile. Would not the pos sibility of replacing lost legs be of as much value to birds and mammals as to crustaceans and newts? Can we conjecture why regeneration has been thrown overboard in the highest classes of animals? For answer, we must turn to the conditions under which the replacement of losses is possible in genera more or less gifted in this respect. The mode of growth or physiological re generation varies in different creatures. Some types, on periodi cally shedding their skin, replace not only this outward covering but renew all the internal tissues too, with the exception of some cells which have lost the power to divide. Others, on the con trary, retain the great bulk of old tissue, as many of these have undergone a transformation inhibiting their further reproduction. Crustaceans, spiders and insects belong to the first, amphibians and reptiles to the second group. The former escape from the old shell or skin in a limp condition ; the latter keep the skeleton, tendons and other internal parts through their moult. The higher the grade of ossification the more stable the system remains and the less it is wont to undergo renewal in physiological regenera tion. Now the power of accidental regeneration wanes with in creasing ossification, indicating once more its correlation with the physiological processes of growth. There appears thus to be an advantage for the rigidity of the body to be maintained through the moults even at the cost of losing the regenerative power by the incorporation of stable elements. We know, indeed, the time of moulting to be a critical one. Even the mere change of skin in the cold-blooded vertebrates and the moult of feathers in the birds is of much inconvenience, these animals being then much more sensitive to external factors. It is therefore reasonable to suppose that in mammals the method of continuous repair has been evolved to avoid the disadvantage of moulting and that the nearly total loss of regenerative power has been the price. It may also have been of advantage to warm-blooded animals to economize growth material by not changing it completely at short intervals, as they expend energy not only in their movements and growth, but also in the maintenance of their body-heat.