encountered in permanent ground pools at Palmalmal. As many as two hundred adults of this active little “whirligig beetle” often occupy an area of only a square yard of the water surface.

Text-fig. 7.—Gyrinus sericeolimbatus.

Adult specimens of G. sericeolimbatus brought to the laboratory refused to remain in their containers. Even when placed in a large basin covered with netting they spent most of their time out of the water. Thus a simple field experiment was undertaken with a view to gaining some idea of the part played by this gyrinid as a predator of the Culicidae. This experiment was carried out in a temporary rain pool averaging 11 ft. 6 in. in length, 4 ft. in breadth, and 9 in. in depth. It was estimated that the pool held about five hundred larvae of Anopheles punctulatus Dönitz. There were no other aquatic animals present with the exception of a few chironomid larvae. One hundred gyrinid beetles were collected from a permanent pool supporting a heavy population of these insects, and were transferred to the temporary rain pool. A visit was made to the latter pool twentyfour hours later, when a search revealed the presence of only forty-six anopheline larvae. After a further twenty-four hours no anopheline larvae whatsoever could be found.

While working in Mauritius, d'Emmerez de Charmoy (1902) found that three adults of the gyrinid Dineutes indus consumed fifty anopheline larvae in three hours. However, Derivaux (1916) observes that even when a certain North American species of the genus Dineutes is abundant in a natural pool anopheline larvae are not particularly difficult to find. He states that although anopheline larvae are soon captured and eaten by these beetles in a basin filled with clean water, only a few larvae are captured, and these apparently by accident, when surface debris in the form of twigs is added to the water in the basin. Derivaux thus considers it improbable that gyrinids are of much significance as moquito control agents, as natural pools usually contain sufficient surface debris to shelter Anopheles larvae from these beetles.

Anopheline larvae were seldom recorded from natural pools supporting populations of G. sericeolimbatus at Palmalmal, although these pools seemed admirably suited for the breeding of Anopheles farauti, and surface debris in the form of fallen branches and floating