Abstract
A discussion is given of problems which arise in using rocket-borne mass spectrometers to determine the composition of the atmosphere at altitudes above100km. A description is given of a magnetic deflection instrument which has been used. Some of the results obtained are discussed.