inspection of starch-gel electrophoresis patterns. Histological examination of the tumour tissue was performed on paraffin sections of formalin-fixed specimens after staining with haematoxylin and eosin, eosin and methyl blue, and PAS-orange G. Extracts of pituitary tissue were prepared by homogenization in tris-EDTA-boric acid buffer at pH 8\m=.\9.The clear supernatant obtained after centrifugation was subjected to radioimmunoassay and to starch-gel electrophoresis. Radioimmunoassay was performed by the solid-phase method (Catt, Niall & Tregear, 1967) using Wilhelmi HGH 705 A as the assay standard. Extracts were diluted to within the range of the assay (0-1-5 ng. HGH) and estimations based on the mean values obtained from two dilutions parallel to the standard curve. Electrophoresis was performed in a vertical tray holding five 0-3 ml. samples of extract in buried-origin slots, each equivalent to 50 mg. of the original fresh tissue. After electrophoresis at 280 for 5 hr., the gels were stained with amidoblack and nigrosine. The HGH content of each extract was assessed visually on a scale of 0 to 5 +, 5 + being equivalent to the main band in the normal pituitary pattern. This method is based on the demonstra¬ tion that HGH forms a major component of the electrophoretic pattern of human pituitary extracts (Ferguson & Wallace, 1963).The results (Table 1) show a general agreement between visual estimation and radioimmunoassay and a marked variation between the mean HGH concentrations in the normal anterior pituitary glands (23/ig./mg.), the tumours causing acromegaly (4-5/tg./mg.) and the non-functioning chromophobe tumours (0-012/tg./mg.). The tumours causing acromegaly did not include a purely eosinophilic tumour, which may be very rich in HGH (Young et al. 1965 ;Lloyd & Meares, 1965). The non-functioning chromophobe tumours, containing minute concentrations of HGH, showed occasional cells with light eosinophilic granulation. The 37-yr.-old acromegalie patient had