Once it is granted that natural death actually exists, it is necessary to study its mechanism as closely as the existing state of knowledge permits. To exclude the possibility of the death having to be interpreted as violent, it would be necessary to know that some very rapid infectious disease does not attack these insects as soon as they emerge from the water. This possibility, although remote, must be examined. Instances are known of large numbers of insects dying very rapidly as the result of attack by a species of mould which causes an epidemic. Every one has seen, especially in autumn, dead flies anchored to the window pane by a little tuft of white fluff. As so many individuals die about the same time, we might be disposed to assign the fact to natural death. The actual cause, however, is an infectious and fatal disease caused by a parasitic mould. The occurrence of some terrible epidemic may be excluded from consideration in the case of ephemerids. I have made investigations which show that such an epidemic does not occur. The bodies of the dying ephemerids contain no microbe which could be the cause of death. Their death must be regarded as natural, as the result of their organisation, as essentially a part of the nature of the insects. Among the cells of their body there are many active phagocytes. Is it possible to attribute death to ravages that these cells may cause among the higher cells and tissues? Microscopic examination, so far from supporting such a possibility, shows that the organs are quite normal in their intimate structure. The brain and central nervous system, the muscles and other organs, show no signs of that invasion by phagocytes found in cases of senile degeneration. In this example of natural death there is certainly no possibility of phagocytic intervention. Some biologists have suggested that the rapid death of ephemerids and of some other insects is due to debility caused by the great effort of depositing the male and female sexual cells. On this supposition, the case would be analogous to the shock which is sometimes the consequence of a surgical operation. This hypothesis, however, may be excluded, for among the dead ephemerids there are many males that have not united with females. Among ephemerids males are much more numerous than females; many males have no opportunity of undergoing the sexual shock and of emptying the reproductive organs, and these, none the less, die as rapidly as the others. As yet we do not know if all the tissues of the ephemerids die simultaneously in natural death. Most probably the cells of the nervous centres perish first, and so bring death on the others. The investigation ought to be made.