In the _Essais optimistes_ he studied first of all the phenomena of old age in the different grades of the scale of living beings, of which he compared the life duration. He concluded that there was an indubitable connection between this and the intestinal flora. The shorter the intestine, the fewer microbes it contains and the longer the relative duration of life. As an example, he quoted the relatively great longevity of birds and bats. Those animals, adapted to aerial life, have to weigh as little as possible. To that end, they empty their intestine very frequently and this in consequence is not used as a reservoir for alimentary refuse; as it is but little developed, it contains a much smaller number of microbes. The longevity of flying animals is relatively much greater than that of mammals with a large intestine full of microbes, a constant source of slow poisoning. After treating the question of longevity, Metchnikoff dealt with that of death. Living beings die, in the great majority of cases, in consequence of diseases or accidents with an external cause; one involuntarily wonders whether there is such a thing as "natural death," _i.e._ arising exclusively from causes due to the organism itself. A review of known facts allowed Metchnikoff to draw the following conclusions: unicellular inferior beings have no _natural_ death; they merely die by accident. Their individual life is very short and comes to an end by multiplication or division of a unit into two; there is no trace of a _corpse_ in this loss of previous individuality. Among superior plants, certain trees attain considerable dimensions (dragon-tree, baobab, oak, cypress), live for centuries, and die from external causes. Their organism presents no _internal_ necessity for a natural death. On the other hand, a multitude of other plants have but a short life and their natural death coincides usually with the ripening of the seed. It has even been observed that it is possible to retard the death of a plant by preventing it from fructifying. For instance, lawns made up of grass mown before it runs to seed remain green and living whilst grass allowed to flower and bear seed becomes yellow and dries up. It is a well-known fact that fruits and seeds are frequently poisonous. Therefore Metchnikoff supposed that the death of the plant may be due to an auto-intoxication by poisons manufactured by it in order to defend its seeds and ensure the next generation; in Nature, the individual does not count, but the species. Once the survival of this is ensured the individual may disappear.