Summary Some of the following propositions are to be read as suggestions or hypotheses, supported by circumstantial or direct evidence, but not yet rigorously demonstrated. An estimate of the significance to be attached to each should be gathered from the body of the paper rather than from the following brief statements. 1.The problem is posed: how do endophagous parasitoids counteract the haemocytic defence reactions of their usual hosts? 2.It has been demonstrated that the egg and young first‐instar larva of Nemeritis canescens have a coating on their surface which enables them to escape the attention of the haemocytes of their usual host, and to develop without exciting a defence reaction. The coating is applied to the egg before it is laid, and to the cuticle of the larva before it hatches. A little evidence suggests that some other ichneumon wasps of the subfamily Ophioninae may use this mechanism of resistance. 3.Older first‐instar larvae, and the second and later instars, of many parasitoids, both hymenopterous and dipterous, probably overcome the haemocytic reaction of their host by rapid feeding, which depletes its blood both of cells and of nutrients, and so drains its resources that haematopoiesis is prevented and encapsulation becomes impossible. 4.The common habit of parasitoids of lingering in the first instar, before ingesting much food, while the host goes on developing to another stage or undergoes diapause, may enable the larva to retain a protective coating that would have become ineffective if it had grown. When at length the larva does feed and grow, the preceding mechanism (3) comes into play. 5.The teratocytes and pseudogerms formed by many species in several families of Hymenoptera absorb nutrients on a large scale from the blood of the host. They act quickly, as soon as the larva hatches. I suggest that by their attrition of the host's reserves of food, and its consequent debility, they prevent an effective haemocytic reaction to the young parasitoid. 6.Some dipterous and hymenopterous parasitoids first inhabit the intestine of their host, and do not penetrate the body cavity until they are ready to overwhelm the defence reactions by rapid and gross feeding. 7.Parasitoids that live temporarily inside an organ of the host may there acquire a coating which protects them from reaction by the blood cells. 8.Species of parasitoids that occupy an organ of the host for a long period, and develop inside it, escape a defence reaction because they live within the connective tissue covering the organ, to which the blood cells do not react. 9.Eggs of hymenopterous parasitoids laid within the embryos of their hosts may be treated by the embryonic blood cells as a developing organ, and become covered with connective tissue as those organs are. Thereafter they would not be recognized as foreign bodies. 10.Parasitoid eggs laid in the eggs or the young larvae of their host may be coated with host substances, or covered by connective tissue (9), before the blood of the host be comes capable of vigorousdefe...

1. Defence reactions to metazoan parasites have been reported in fourteen orders of insects. The observations are brought together and reviewed in the first part of the paper.2. Examination of the various accounts that have been given shows that blood cells are always involved in insect defence reactions. They act by forming a cellular capsule, from which a connective-tissue envelope is usually developed, and in which melanin is often deposited.3. The reaction of the epidermal cells at perforations made by parasites is of the nature of wound-healing, and plays no part in defence against metazoan parasites after they have entered the body.4. Although several other tissues have been implicated, there is insufficient evidence to show that any of them make defence reactions, their response being limited to processes of regeneration.5. It is concluded that the blood cells of insects are their only known agents of defence to internal metazoan parasites.6. The principal groups of metazoan parasites infesting insects are considered in the third part of the paper, in order to see how the defence reactions made to them are related to their mode of attack and to the nature and consequences of their parasitism.7. Most parasites elicit a defence reaction when they are in unusual hosts.8. Some parasites, at certain stages of their life-history, are able to avoid eliciting a defence reaction in their usual hosts.9. Some parasites elicit a defence reaction in their usual hosts but are able either to endure it in a dormant state or to resist it.10. General problems of host specificity in relation to defence reactions are discussed. It is concluded that analysis of the stimuli that produce defence reactions has not yet gone far enough to explain the phenomena.11. The effects of insect defence reactions on metazoan parasites range from no perceptible effect to destruction of the parasite.12. The defence reactions of insects are influenced by the species, genetic strain, stage, instar, size, health and physiological state of the host; and by the species, genetic strain, physical and physiological activity, and health of the parasite. Environmental temperature and the presence of other parasites of the same or different species also have effects on the reactions.13. A brief survey of defence reactions made by invertebrates other than insects shows that encapsulation has been reported in Annelida, Mollusca, Crustacea, Acarina, and larval echinoderms. So far as it goes, the survey does not reveal in these other groups any reaction to metazoan parasites of a kind radically different from the reactions observed in insects.14. The historical development and present state of our knowledge of insect defence reactions is traced.15. The reactions made by insects to innocuous parasites are of theoretical interest but of little consequence to the species concerned. It is their effect on potentially dangerous parasites that determines the value of defence reactions. Consideration of the evidence suggests that the protection afforded to insects by their defence reactions is greater than has been generally supposed.16. The review makes apparent many gaps in our knowledge of the phenomena. A few of the outstanding problems are mentioned.I am indebted to Mr R. T. Hughes of the Balfour Library for helping me to obtain journals not available in Cambridge; to Mr M. J. Ashby for the photography necessary in the preparation of the figures; and to Miss G. M. Edwards for her careful typing from my manuscript. The paper would not have been completed without the goodwill and assistance of two persons: Professor D. Keilin, F.R.S., encouraged me to continue and finish it when my effort flagged; my wife not only gave me positive help in many ways but also exercised great forbearance in allowing me to devote vacations and spare time to it.