PERSPECTIVES

rotation. One might not expect two halteres, or two antennae, to be capable of distinguish-ing three components of rotation, but they can, for a rather subtle reason. The spectrum of Coriolis forces has peaks at the frequency of the haltere beat and at twice that frequency, and the relative magni-tude of the peaks depends on the plane of rotation of the body. Thus, the sense organs in the stalk of the haltere experience a mix of two frequencies of mechanical stimulation, which changes as the plane of rotation changes, enabling each haltere to distinguish two components of rotation. The third possi-ble component (in the plane of the haltere’s beat) generates no out-of-plane force, but flies can still fly after one haltere has been amputated, suggesting that each haltere may be capable of distinguishing all three compo-nents. They could conceivably get the third one by sensing tension as well as bending in their stalks, but this would require the ability to measure tiny fluctuations in a much larger centripetal force (5). In any case, two halteres or antennae set at different angles, each capa-ble of distinguishing two components of rota-tion, can together provide full information about rotations in three-dimensional space. Now we would like to know whether other insects, as well as hawk moths, use their antennae to sense rotations. That capa-bility seems potentially useful to any flying animal