It is necessary to correct the impression given by Philip Eden in the November 2012 issue (p. 306) that to obtain the CET Gordon Manley continued to advocate using the average of the Oxford and Lancashire sites. Eden's quotation comes from Manley's 1953 paper. In 1974 Manley wrote temperatures at the Radcliffe are gradually becoming less representative. He advocated using the mean of Rothamsted, Shinfield, Abingdon and Cirencester plus 0.4, combined with the mean of Lancaster, Kirkham, Preston and Stonyhurst plus 0.6. This contradicts Eden's claim that Manley's concept was to use just two stations.
It is, and I will use the word again, unfortunate that there are now two different versions of the CET series, the longest continuous series of monthly and annual temperatures from anywhere. It is surely necessary that a consensus is reached as soon as possible on the best way to extend Manley's great work.
Philip Eden responds: The point I was trying to make is that the Hadley Centre's CET uses a tri-polar scheme, weighted towards the south Midlands. Manley's CET uses a bi-polar scheme with equal weight given to the south Midlands and the Lancashire Plain.