It was desired to confirm quantitatively pilots’ impressions that on turbine-jet aircraft in the region of 40,000 ft. the level speed stabilised very slowly.

The time required to make a conventional level speed measurement has been examined, for a representative turbine-jet aircraft, over a wide range of air speed and height. The results are compared with those for a typical piston-engined aircraft and their implications discussed.

It is concluded that the time required to make a conventional level speed measurement increases rapidly with height, in the absence of compressibility effects on drag, and is about five times as great at 40,000 ft. as at sea level. It is inconveniently large at high altitude and level speed measurements there will require very long runs, unless it is arranged that each run is started at an air speed near the steady level speed.

This increase will be present in most level speed tests, especially those in the region of the best speed for range. In maximum level speed measurements, however, compressibility effects may be present and the air speed may then be expected to settle down quickly.

The need for long runs at high altitude is not peculiar to jet aircraft, but has come into prominence concurrently with them.