“Finally, individuals, usually females, show complete degeneration. A female taken in a fine-meshed net showed no trace of organs in the head; the eyes, the brain and the intestinal tract had disappeared almost completely. The antennæ were reduced to stumps consisting of the lowest joint and a portion of the second. These were clear indications of the senility that precedes death” (p. 194). Such evidence not only supports the hypothesis that the natural death of Monstrilla is due to inanition, but is opposed to a similar interpretation being applied to the case of male rotifers, in which death is not preceded by wasting of the organs. The death of some insects, which comes rapidly after the adult stage has been reached, cannot readily be attributed to starvation. In the strange butterflies known as psychids (_Solenobia_) some of the females lay eggs without having been fertilised,[89] and their life in the adult condition lasts only a day. On the other hand, other females of the same butterfly are fertilised before laying their eggs and in this case survive for more than a week although they take no food. The rapid death of the first-mentioned set cannot be attributed to inanition. In some Ephemeridæ, which supply good cases of natural death, the end comes after a few hours of adult life without any sign of degeneration of the organs. As in others (_Chloë_), life lasts for several days without food having been taken, it is clear that inanition is not the cause of the swift arrival of death in the first set. It is much more probable that the natural death is due to an auto-intoxication which takes effect at different intervals of time in different circumstances.[90] In the higher animals such as vertebrates the conditions are less favourable than in the case of insects for the investigation of the causes of natural death. Vertebrates have always well-developed organs of digestion and so live a relatively longer time and encounter a greater number of chances of accident, with the result that in most cases death comes from external accidental causes. Vertebrates usually perish from hunger or cold, or are devoured by their enemies or killed by the attacks of parasites or diseases. There remains only the human race amongst the more highly developed animals, in which to study the onset of natural death. And in the human race cases which may be designated as natural are extremely rare. III NATURAL DEATH AMONGST HUMAN BEINGS Natural death in the aged—Analogy of natural death and sleep—Theories of sleep—_Ponogenes_—The instinct of sleep—The instinct of natural death—Replies to critics—Agreeable sensation at the approach of death