THE METABOLISM OF THE CARBOHYDRATES 713 



permissible to assume that a much more striking difference could be demonstrated 

 when the perfusion fluid is made to traverse all or practically all of the skeletal muscles, 

 as well as the heart. For this purpose an eviscerated animal may be employed that 

 is, one in which the abdominal viscera are removed after ligation of the celiac axis 

 and mesenteric arteries, and the liver is eliminated by mass ligation of its lobes. 

 Using such preparations, E. G. Pearce and Macleod29 found that the rate at which 

 glucose disappears from the blood, although very irregular, is in no way different in 

 completely diabetic as compared with normal dogs. They were thus unable to confirm 

 any of Knowlton and Starling's earlier conclusions. As has already been stated 

 Patterson and Starling subsequently pointed out that a serious error was involved in 

 the earlier perfusion experiments, partly on account of a remarkable but irregular 

 disappearance of glucose from the lungs, and partly because the diabetic heart may 

 contain a considerable excess of glycogen, from which its demands for sugar may be 

 met without calling on that of the perfusion fluid. 



More recent work by Starling and Evans, in which the respiratory ex- 

 change of the heart in heart lung preparations from diabetic (pancreatic) 

 dogs was determined, has revealed a low R.Q. (0.71) although the oxygen 

 consumption was normal. These workers consider that their results indi- 

 cate "a depression or abolition of the power of the diabetic tissue to utilize 

 carbohydrate." This conclusion, however, awaits confirmation by more 

 direct methods. 



Whatever may be the fundamental cause for the development of dia- 

 betes following pancreatectomy it is important to know whether the 

 condition supervenes because of the absence of an internal secretion, or 

 hormone, furnished by the pancreas or merely because some local action 

 of the gland has been removed. Attempts have been made (by Hedon, 

 Carlson and Drennan and others) to demonstrate the presence of such 

 a hormone in the blood of normal animals but with doubtful success. 



I That one may be present has however been shown by A. H. Clark, 65 who 

 perfused the isolated pancreas of a dog for some time with Locke's solu- 

 tion and then used this fluid to perfuse the isolated heart of the same 

 animal. He found under these conditions (i.e., of combined perfusion) 

 that sugar disappeared much more rapidly from the perfusion fluid than 

 when fresh Locke's solution was perfused through the heart from the 

 beginning. This indicates that the pancreas must contribute some sub- 

 stance which is carried by the blood to the heart, and presumably other 

 muscles, where it enables the muscle to consume sugar. 

 Until recently, however, no constant success has attended the injection 

 of pancreatic extracts in reducing the hyperglycemia or other symptoms of 

 diabetes. Thinking that this might be due to the destruction of the hor- 

 mone by the trypsin present in such extracts, Banting and Best, working 

 in the author's laboratory, made extracts either from the residue which 

 remains some weeks after ligation of the pancreatic ducts or from early 



fetal pancreas. In both these cases trypsin production in the gland is 

 absent, or nearly so, but it is believed that the Isles of Langerhans are 



