The following communication is an abridgement of a paper on the same subject presented to the Royal Society in 1858, with some additional matter that has been since disclosed by my experimental investigations. The original paper, being deposited in the Archives of the Society, is accessible for reference on points of detail that are here excluded. In 1848 it was announced by Bernard that the liver enjoyed a sugar-forming function. This statement appeared to rest upon irrefutable grounds, and the new function soon became almost universally acknowledged by physiologists. An animal which had been for some time previously restricted to an animal diet was suddenly killed. Sugar was found abundantly in the blood of the vena cava and hepatic veins, whilst none existed in that of the portal vein. The tissue of the liver was also found abundantly saccharine, whilst no sugar was to be detected in any other organ. I had seen this experiment several times performed in Bernard’s laboratory, and had often repeated it myself. From the correctness of the description of his results, I entertained no doubt as to the accuracy of Bernard’s deductions, and did not for a moment seek to question them. In the course of my experimental research, however, I was conducted step by step to a point which has placed me, involuntarily as it were, in antagonism with the glycogenic theory. By pushing investigation further than had hitherto been done, I have been compulsorily brought to arrive at conclusions of which I had not the most remote anticipation beforehand.