Soon after Banting and Best's discovery, insulin was found capable of correcting hyperglycemia, the apparently basic abnormality in all diabetes mellitus. Lack of insulin is thus one potential cause of diabetes mellitus, and this has been well demonstrated by Wright and his colleagues who produced acute diabetes mellitus in animals with single injections of an anti-insulin serum. 1 Subsequent discoveries, however, have made defective insulin secretion an insufficient explanation for the syndrome seen in many diabetics. As methods of assaying insulin-like activity in blood have developed, only some types of diabetes mellitus have been found to show low levels of circulating insulin—typically the thin, ketotic, juvenile type of diabetic, while the obese type usually shows normal levels. 2" 5 Recently, even abnormally high levels have been reported in some untreated diabetics6" 8 and also in prediabetes. While overproduction of hormonal antagonists such as growth hormone and cortisol can explain some cases of diabetes with normal insulin levels, particularly those in whom the diabetes is obviously part of the syndrome of acromegaly or Cushing's, nevertheless in most cases of diabetes there is little evidence of overproduction of either of these hormones. As possible explanations of these cases, there are now emerging various other factors which may affect the action of insulin on the tissues but whose interrelation with these hormonal antagonists and with each other is not yet clear. Insulin antagonists have been demonstrated in blood, ie, substances probably not hormones which interfere with the biological effect of insulin on tissue; 9" 11 they have usually been shown by their inhibitory effect on in vitro assays of insulin activity using the rat diaphragm method; for they are not active on the fat-pad assay. 12 Their demonstration in prediabetes suggests that they might be primary, 9 though their mode of action is unknown. Further and independently, plasma insulin-like activity has been shown not to be