Another great advantage, the most far-reaching of all, has been gained by the stabilization of the nervous ganglionic cells, which cease to divide in the mammalian brain comparatively soon after birth, thus guaranteeing a stable substratum for the preservation of memory. In several instances we have been able to prove this by experiment. Infusorians fed with carmine refuse to take the grains of this substance after a short time. But as soon as the animal has split into two new cells, this memory is lost, carmine again being taken. Certain predatory grasshoppers may be taught to take food at a selected place of their cage, but be tween two moults they forget what they have learnt before, probably because some ganglion-cells are renewed, as seen by his tological examination. The same behaviour has been observed in amphibian larvae, when they metamorphose. This raises the question as to what influence age and stage of development may have on regeneration. If our assumption be correct, that acci dental regeneration is directly connected with physiological regeneration and growth, we should expect most regeneration at those stages during which physiological processes and increase of size are at their height.
The rapidity of substantial change and growth decreasing with age, young animals ought to regenerate better than old ones. This conclusion is borne out by all the facts and experiments at our disposal, as long as the removed parts are strictly comparable. From this general statement it follows that higher organisms devoid of a certain faculty to restore parts of their body in a developed state may possess this faculty in their youth. The frog is unable to replace a limb, whereas it is able to regenerate the same member as a young tadpole. The most striking example of this kind is offered by the insects, which never regenerate limbs as imagos, but do so as long as they are in the larval state. Here, with the last moult to the definitive form, growth ceases. Examples have been brought forward in which seemingly the young stages show less regenerative capacity than older ones. If the tailed larva of the tunicates be cut in two, regeneration will not take place in both pieces, whereas the meta morphosed animal may be cut in two and regenerate from both pieces. But then the cut passing transversely through the middle does not separate homologous halves in larva and imago. Again, in the starfish the developed creature restores all missing parts when cut in two, whilst the pluteus stage is incapable of regrow ing its skeletal arms. Here, too, there is no homology whatsoever between the so-called pluteus arms and the rays of the developed starfish. These latter arise from a new rudiment at one side of the larva during metamorphosis. Since it is difficult to compare the regenerative capacity of larval and imaginal stages, it is neces sary to restrict our research to such periods of animal life, during which growth is yet proceeding, but no change in differentiation occurs, when seeking strictly comparable quantitative data of regeneration. Such material afforded by tadpoles, by newts in their larval or in their metamorphosed state, and by crustaceans.
has been treated statistically as to the increments of size in the regenerating tails, legs and claws respectively in equal periods. The increase in the size of the regenerating member is first only small, rising in time more and more, then again falling off and subsiding to the growth-rate of the normal member, as soon as the normal size has been attained by the regenerate. This rise and fall is mainly due to the increase of regenerating material, which multiplies from a small bud, but later on lessens its rate of multiplication. If we do not take the absolute mass that has been added to the miniature regenerating member, but calculate the coefficient by which we have to multiply the size at a given time to get its size at the next, this relative value is highest when the regenerating bud begins to differentiate, and falls off first quickly, then more slowly as time goes on. The curve of regener ation is inversely proportional to the time. We may picture growth as a stream losing energy as it flows on, but on the whole not showing a great fall from one point to the next. If, on the contrary, a part of the river-bed is dug out, so that its steepness is increased, the rapidity of flow will rise at the commencement of this section, then slow off and regain its natural rate, when it reaches a normal level. If this comparison is true, we must con clude that, everything else being equal, the rapidity of regenera tion ought to be greater if the cutting has been done nearer to the body than if it has been carried out further towards the tip. Now this has been observed over and over again. Unacquainted with these statements, some authors have postulated a direct cor relation between distance of cut surface from the regenerating body and the quickness of regeneration, on the plea that it would be easier for the animal to repair a small injury than a big one. But this latter thesis is not in accordance with experiment, for re generation after the amputation of several limbs is not only as rapid as that of a single limb but actually proceeds even more rapidly. There is, of course, a limit above which regeneration slackens or ceases, when the injury has been so severe that the re serves of bodily growth are not able to fill the gaps. The ex planation for the increasing velocity of regeneration with increas ing losses may be found in the decrease of the mass which has to be continuously overhauled by physiological repair. The less the old parts require of this process, the more material will be avail able for newly growing structures. It is perhaps in this connec tion possible to account for some observations of increasing veloc ity of reconstruction with repetition of the same operation. The posterior end of the worm Amphiglena is provided with branched gills. If these are cut off, new ones make their appearance. If these in turn are removed, they will be replaced in a shorter time, and so on, more quickly every time they have been lost. As the old tissue at the cut surface is also remoulded, less of this work will have to be done when the succeeding cuts pass through more and more rejuvenated tissue.