1. I have given a full historical survey of the problem of the rise in blood sugar associated with anaesthesia. • 2. I have examined the literature on the subject and the related problems in carbohydrate metabolism, in considerable detail. • 3. I have reviewed the various theories of the mechanism of ether hyperglycaemia. • 4. There is a well marked hyperglycaemia associated with ether anaesthesia. • 5. There is marked individual variation. • 6. The type of operation has no effect. • 7. The length of the operation has no effect - short of the production of shock which has a marked hyperglycaemic effect. • 8. The amount of anaesthetic used has no effect. • 9. Induction of the anaesthesia by ether and chloroform mixture, nitrous oxide or ethyl chloride makes no difference • 10. The pre -anaesthetic administration of omnipon and scopolamine has no influence on the hyperglycaemia. • 11. The pre-anaesthetic administration of glucose has no effect on the hyperglycaemic response to ether. • 12. The post -anaesthetic administration of 10% carbon dioxide in oxygen hastened the recovery from the anaesthetic. It also accelerated slightly the return to normal of the blood sugar. • 13. The rise in blood sugar is present at the end of the period of induction. It then continues as a plateau and has usually begun to fall again by the end of the operation. The impression gained was that it began to fall when the administration of the anaesthetic stopped. • 14. Ether anaesthesia has no effect on the diastatic index of the urine. • 15. The pre -anaesthetic administration of amytal almost entirely prevents the hyperglycaemia of ether anaesthesia. This is in agreement with the work of Campbell and Morgan (40) . • 16. The number of females in the series is small but the impression gained was that the hyperglycaemic response to ether, was slightly less in females. • 17. I believe that the mechanism of ether hyperglycaemia is mainly due to the depressant action of the ether on the piqure centre in the medulla. Alteration in the hydrogen ion concentration - increase - has a slight effect. This is probably caused by some respiratory mechanism. The grounds on which I base this view are :- (a) Amytal prevents the occurrence of hyperglycaemias of central origin. It prevents ether hyperglycaemia. (b) Amytal has no effect on the hyperglycaemia due to epinephrine. (c) The amount of anaesthetic - short of the production of shock - has no effect on the hyperglycaemic response. This rules out the direct effect of ether on the liver cells as a cause. (d) The blood sugar has begun to fall while the acidosis is still marked. This appears to exclude acidosis as the main factor. (e) The continuous administration of oxygen during the anaesthetic has no effect on the hyperglycaemic response. I gathered this from my own observations and the literature. (f) The experiments of Ross and Davis on anaesthesia in dogs without the pancreas and dogs with part of the pancreas remaining are inconclusive. The difference in the hyperglycaemic response in the two groups is too small to be of any importance. I consider that depression of the internal secretion of the pancreas is not part of the mechanism of hyperglycaemia of ether anaesthesia. • 18. The blood sugar has been estimated in 105 cases under general anaesthesia and I consider this a sufficient number to draw conclusions from. • 19. There is a rise in blood sugar under spinal anaesthesia. • 20. It is much less marked than ether hyperglycaemia. • 21. It appeared to have some relation to the fall in blood pressure. • 22. It was not caused by the ephedrine injected before the spinal anaesthetic was given, since the injection of ephedrine in a control group caused no rise in blood pressure. • 23. Trauma causes a rise in blood sugar even when there is no clinical evidence of shock. • 24. This was compared in a group of fractured tibiae and a group of head injuries. The average rise was slightly greater in the head injury group. • 25. A rise in blood sugar was obtained on injecting 30 ccs of 2.0 novocaine into the haematoma round a fracture. • 26. The response of the blood sugar was variable and sometimes did not occur at all. When it did occur it was prolonged - lasting about 48 hours and often associated with glycosuria. • 27. The explanation suggested is that the novocaine was absorbed rapidly and poisoned the liver cells, in those cases associated with marked hyperglycaemia. • 28. Small operations performed under 2% novocaine were not associated with hyperglycaemia. • 29. The explanation advanced is that the absorption of the novocain was slow. • 30. I can find no references in the literature concerning the blood sugar in spinal and local anaesthesia.