“…In a recent letter to the Editor, Prangnell and Johnson (1976)1 pointed out that, using a Coulter Counter S with calibration of the haemoglobin levels against three cyanmethaemoglobin standards, the results averaged 0 5 g/dl higher than in laboratories using Coulter 4C to calibrate their machines. In the same way, Lewis (1976), in a comment on this letter, says that in the national quality control trials it has been found that participants using Coulter S produce a lower mean value for haemoglobin than when using other systems, although the difference is in the order of 0-2-0-3 g/dl; 0 5 g/dl is just within 2SD.…”