In 1966 [Sir] Fred Hoyle published two papers on the possible use of Stonehenge as an eclipse predictor (1966a,b). He has now (Hoyle, 1977) returned to this theme with a very clearly written and well-illustrated book that is essentially an expansion of the material in the two 1966 papers. In brief, Hoyle proposes that the extant features of Stonehenge I provide the necessary means systematically to observe the sun and moon and to keep track of the nodes of the lunar orbit. This would provide sufficient information to allow eclipses of the moon to be predicted. He also proposes that by the time of Stonehenge III this method was superseded by the use of the Saros cycle of 18 years 11 days to predict eclipses. This last method almost certainly implies that written records of events at full moon were kept. Since there is no evidence from the British Late Neolithic or Early Bronze Age of writing or numeracy this proposal is purely speculative.