Is natural death due to the gradual production in the body of harmful toxins or to the gradual destruction of substances required to keep up youthful vigor? If the latter, the natural duration of life would be the time necessary to complete a series of chemical reactions that would produce enough of the toxins to kill. Now, the period necessary to complete a chemical reaction diminishes rapidly when the temperature is raised, and increases when it is lowered. This time is doubled or trebled when the temperature is lowered by 10° C. The influence of temperature on the rate of these processes seems typical. If the duration of life, then, is the time required for the completion of certain chemical reactions in the body, we should expect it to be doubled or trebled when we lower the temperature. We can test this only where, as in our flies, infection is avoided. Northrop put their fresh eggs on sterilized yeast at a temperature of 0.2° C., and the higher temperatures selected were 5°, 10°, and 25°. All the flies died at nearly the same time when kept in the same temperature. The total average duration of life was 2½ days at 30° C., when nearly all of them died. At 10° C. it was 177 days. Thus heat accelerates all chemical action, and here we have the duration of life increased from 200 to 300 per cent. In man the body temperature is constant, for example, 35.5° C. whether in the tropics or the Arctic regions. If we could reduce our temperature, we might live as long as Methuselah. If we could keep the body temperature at 7.5° C. and follow the above ratio, we should live about 27 times 70 or about 1,900 years. Thus the duration of life seems to be the time required for the completion of a chemical reaction or a series of them. The latter may be the gradual accumulation of harmful products or the destruction of substances required for sustaining youth. Not only are unicellular organisms immortal and the life of all their successive generations a continuum, but a bit of cancer tumor can be transplanted to other individuals and there grow larger, and a bit from this second individual transferred to a third, and so on indefinitely; so that the same cancer cell continues to live on in successive transplantations throughout many individual lives. It has thus outlived many times the natural life of the mouse. Indeed, it seems to be able to live on indefinitely and Carrel has shown that this is true of other normal cells. Thus death may not be at all inherent in the individual cell but only be the fate of more complicated organisms in which the different types of structure depend on each other. Certain cells are able to produce substances that slowly become harmful to some vital organ or center and its collapse brings death to the whole.