In a recent note, Collier and Lockwood (1974) compared measurements of total solar radiation on cloudless days at an inland site in England with radiation calculated from empirical expressions derived from (i) data from weather ships in the eastern Atlantic (Lumb 1964) and (ii) measurements of radiation and turbidity in central England (Unsworth and Monteith 1972). Collier and Lockwood concluded that radiation received inland under cloudless skies was only about 60% of that received over the sea. As this figure is inconsistent with the whole literature of radiation climatology in Britain, we believe there must be large errors both in the measurements they recorded and in estimates they derived from independent empirical relationships. Collier and Lockwood tabulated hourly mean values of total solar radiation S, for three cloudless days (20 August 1971, 14 October 1971 and 21 January 1972) and they plotted the dependence of S, on solar elevation O for these and other days. Hourly means of S, are recorded by the Meteorological Office and other establishments for a number of sites, and comparison shows clearly that for given values of 0, under cloudless skies, Collier and Lockwood's values are 30-40 % lower than those recorded elsewhere in Britain.