Abstract
1. The egg distributions of five chalcid parasites of the knapweed gall-fly have been studied in a small area.2. A distinction is made between search for hosts by random movements, and random search in the sense of Nicholson; these terms are defined. Random movements may give rise to a distribution of parasitism different from that expected if search were purely random.3. The species discussed do not all distribute their eggs at random amongst the hosts. Some species superparasitize the hosts more, and one (Eurytoma curta) much less than would be expected if the egg distribution were purely random, owing to peculiarities in oviposition behaviour.4. Parasitism byE. curtais unevenly distributed in space, being higher in areas of high host density. This suggests that search is not exactly random, but is also concentrated in space. A tentative explanation is advanced based on the assumption that search is by random movements.5. Parasitism byE. robustais very patchy. This non-random distribution is attributed to random movements performed by a very few parasites.6. It is concluded that these spatial discrepancies do not seriously affect the numerical accuracy of the assumption that search is random in small areas, and that the theory of Nicholson & Bailey may be accurate to a first approximation.