The paper describes an investigation aimed at finding out whether solar radio bursts of spectral type III are due to disturbances which travel out through the corona with velocities exceeding O�lc, as predicted by the well-known hypothesis that the emissions are due to plasma oscillations. If the proposition is correct, emissions at different frequencies would be generated at different levels in the corona-the lower the frequency the higher the source. This property is tested by simultaneous directional observations at a number of frequencies between 40 and 70 Mc/s, using a swept-frequency interferometer.

From a study of simultaneous metre and centimetre solar radio bursts, it appears that the type III events which coincide with centimetre bursts are frequently followed by a particular form of broad· band emission. This last burst, termed type V, is observed mainly below 150 Mc/s on radio spectrum records where it resembles a bright glow lasting for about 1 min. Synchrotron radiation has been suggested as the mechanism for type V bursts.The accompanied centimetre burst lasts a length of time comparable with that of the metre bursts, is indistinguishable in intensity from unaccompanied centimetre bursts, and, on present evidence, may be due to either thermal or synchrotron radiation.A stream of highly energetic particles ejected from a :flare region on the Sun is suggested as the cause of both events, exciting centimetre bursts in or near the chromo· sphere and type V bursts at large heights in the corona.