I want to speak of the work we have been doing in Cardiff on the metabolism of the nervous system. The work was carried out there because of the importance of the narcosis treatment. It seemed to us there a pity that a treatment such as that should be given up because of the considerable toxicity possible in relation to it. The research was undertaken to see if we could diminish the toxicity, at the same time seeking an idea as to how narcotics work. I ask that you will realize that the main substance burned by the brain is glucose. The dominant form of metabolism in the nervous system is connected with the breakdown of glucose and lactic acid, and this can be proved by experiment in the living animal and with brain-tissue in vitro. In doing experiments we are not able to carry out work with human brain, because we cannot get human tissue fresh enough, so we have to carry out experiments with animals. They are carried out in this way. We cut slices of the cortex of the brain as soon as the animal is dead, that is to say, within ten minutes of death the brain is out and slices have been cut. They are placed in a physiological medium in the presence of glucose, and we follow the metabolism of that tissue, which allows us to estimate the amount of oxygen being taken up by the brain. If luminal, chloretone, hyoscine or somnifaine be placed with the brain-tissue, then the respiration, instead of being at the usual level, starts lower down, and maintains a straight line. We wanted to see whether this action is reversible or irreversible. If the latter, then on removing the brain-slices from the narcotic it should no longer behave like a normal piece of tissue. Actually, when the brain-slice is removed and placed in Ringer solution, with no narcotic, the respiration goes up and becomes equal to that shown by the slice which had no narcotic. That is to say, the process is reversible.