in July 1971, another (48) has not been seen since October 1967, and three (8, 13, and 24) had emigrated in good health in 1968, 1969, and 1971, respectively. Eleven patients had died. The table shows the actual number of deaths compared with the expected number in 5-year age groups. Cerebrovascular disease probably caused death in four cases, and cardiovascular disease in two. Two others died suddenly and unexpectedly while recovering from a severe painful crisis and the lack of causal pathology at necropsy was compatible with a cardiac arrhythmia. Forty two were traced during 1973, of whom 32 had regularly attended the clinic during the past five years, six had been specially contacted and examined for the present survey, and four (2, 22, 40, and 54) had not been seen but had corresponded or were reliably reported alive. Thirty three patients were in reasonable health, 22 of them in full time employment. In nine cases, symptoms of the disease interfered with normal activity, owing to leg ulceration in six, low haemoglobin levels in two, hepatic enlargement and abdominal swelling in one. When analysed in 5-year age groups the haematological characteristics of the elderly group did not differ significantly from those of younger ages.