low unicellular algae, and the more highly organized unicellular infusoria. They can be killed, but they do not die naturally. In this sense they are immortal. And, as Professor Weismann puts it, "Each individual of any such unicellular species living on the earth to-day is far older than mankind, and is almost as old as life itself."

Assuming, then, that the higher forms are evolved from the lower, when did natural death appear, and how could it have been-for this is Weismann's contention—a gain to the species, brought about by natural selection?

The answer is that when we pass from the unicellular to the multicellular organisms we are at first as far as ever from death. Multicellular organisms, like unicellular, are capable of being killed, but so long as the cells are homogeneous, there is no death. For the dissolution of the cellcolony is not death, since the separated cells still live. But when among the multicellular organisms we reach those in which a division of labour has taken place among the cells-that is to say, when we pass from homoplastides to heteroplastides—a new fact appears. The previously identical cells have become dissimilar, some being perishable, the somatic cells, the others, the reproductive cells, retaining the potential immortality, which belonged alike to unicellular organisms and to the cells of the homoplastid colony :

"The mortality of the somatic cells," says Professor Weismann, arose with the first differentiation of the originally homogeneous cells of the polyplastids into the dissimilar cells of the heteroplastids. And this is the first beginning of natural death."

At first, while the immortal reproductive cells far outnumbered the perishable somatic cells, the death of the latter would be hardly noticeable. The organism, as a whole, would seem to survive, though a part of it perished. But when, in the individual, the perishable somatic cells came to outnumber the reproductive cells, death would emerge into prominence. For that which lived would be relatively small and unimportant compared with that which died. And when, moreover, the potentially immortal germ cells, owing to the death of the somatic cells, lacked the conditions under which they might build up a new body, they would also die, and so the natural death of the somatic cells would become incidentally the cause of death in the germ cells. For instance, if a plant or insect dies before all the reproductive cells have matured, these remaining cells are killed by the death of the soma. And thus the mortal secures an accidental triumph over that which in its own nature was potentially immortal.

If we ask-Why did the mortal thus triumph over the immortal? the answer is that it must have been for the good of the species. But this at once suggests the further question,

If natural selection, operating on the material of minute quantitative variations in the cells, not only produced a quantitative distinction between mortal and immortal, but put a premium on death, why should not immortality some day reappear? If immortality is lost because it proved under certain conditions "a useless luxury," under other conditions it might reappear. Professor Weismann answers-The line between mortal and immortal is less sharp than we might suppose. For the immortal cells, which do not die, can be killed; and the mortal cells do recover the power of indefinite, if not infinite, reproduction, where long life is necessary for the good of the species. The duration of life is seen to be controlled by the good of the species, and long life, where it is so needed, is secured by an increase of the number of cellgenerations in the soma.

If we further ask-Why, then, may not this increase continue till the line between reproductive and somatic, immortal and mortal disappears? the answer is, that there is no reason, except that we cannot imagine conditions under which such an extension would be for the good of the species. A Tithonus, endowed with immortality but not exempted from the wear and tear of life, is as useless to the species as he is burdensome to himself. The soma being in its nature vulnerable, would, if immortal, become of less and less use to

the species, and the life potentially immortal would actually be limited by natural selection. Nature does not feed "useless mouths." The higher organisms, then, contain the germ of death, not because death is a primary necessity for living things, nor because in them a distinction between reproductive and somatic cells exists, but because in the somatic cells the power of indefinite multiplication ceased to be of use, and so was lost. In short, the death of the individual was for the good of the species.

Such is Weismann's theory of the origin of death. And it is interesting to find that, more than twenty years ago, Dr. A. R. Wallace had hit upon a similar explanation. In a note written some time between 1865 and 1870, but published for the first time as a footnote in the present volume, he says that while an organism, which increases by fission, would survive in spite of the destruction of its individual separated parts, those organisms which give off very small portions to form new organisms would be at a great disadvantage as compared with these smaller organisms in the struggle for existence, and would soon cease to exist::

"This state of things," he says, "would be in any case for the advantage of the race, and would therefore, by natural selection, soon become established as the regular course of things, and thus we have the origin of old age, decay, and death; for it is evident that when one or more individuals

have provided a sufficient number of successors, they themselves, as consumers of nourishment in a constantly increasing degree, are an injury to those successors. Natural selection, therefore, weeds them out, and in many cases favours such races as die almost immediately after they have left successors."

Here we have Weismann's theory of the origin of death, not indeed worked out as it is in his Essays on the "Duration of Life" and "Life and Death," but thrown out as a suggestion, which seems to have lain dormant and been forgotten even by the author till the essays of Professor Weismann were submitted to him in proof.

2. The other point in Professor Weismann's philosophy is his theory of heredity, which follows as a consequence from his theory of the origin of death and the separation of somatic and reproductive cells in the heteroplastids.

It is clear that in those organisms which increase by simple division the likeness which exists between the divided parts is simply the likeness of identity. The offspring is "a chip of the old block" in a literal sense, except that the question of age does not come in, each part being as old, or as young, as that of which it is a part. There is, as yet, no question of heredity. After the separation, any of the separated parts might by direct action of environment be modified, and so become different from the others, but when they in turn increased by fission their divided parts would, to start with,