When these eggs are deprived of oxygen at the time they reach the eight- or sixteen-cell stage, it can be noticed that the membranes of the blastomeres are transformed into small droplets within half an hour or more, according to the temperature. These droplets begin to flow together, forming larger drops. [Figures 48 to 51 show the successive stages of this process.] When the eggs are exposed to the air in time, segmentation can begin again; but if a slightly longer time is allowed to elapse, the process becomes irreversible and life becomes extinct. Such clear structural changes cannot often be observed in the eggs of other animals under the same conditions. Are these changes of structure (apparently liquefaction of solid elements) responsible for death under such conditions? In order to obtain an answer to this question, the writer investigated the effect of the lack of oxygen upon the heart-beat of the embryo of the same fish _Ctenolabrus_. This egg is perfectly transparent and the heart-beat can easily be watched. When these eggs are put into an Engelmann gas chamber and a current of pure hydrogen is sent through, the heart may cease to beat in fifteen or twenty minutes; it stops beating suddenly, before the number of heart-beats has diminished noticeably, and ceases beating before all the free oxygen can have had time to diffuse from the egg. In one case the heart beat ninety times per minute before the hydrogen was sent through; four minutes after the current of hydrogen had passed through the gas chamber, the rate of the heart-beat was eighty-seven per minute, three minutes later it was seventy-seven, and then the beats stopped suddenly. It is hard to believe that this cessation could have been caused by lack of energy. Hydrolytic processes alone could furnish sufficient energy to maintain the heart-beat for some time, even if all the oxygen had been used up. The suddenness of the standstill at a time when the rate had hardly diminished seems to be more easily explained by a sudden collapse of the machine; it might be that liquefaction or some other change of structure occurs in the heart or its ganglion cells, comparable to that which we mentioned before. In another fish _Fundulus_, where the cleavage cells undergo no visible changes in the case of lack of oxygen, the heart of the embryo can continue to beat for about twelve hours in a current of hydrogen. In this case the rate of the heart-beat sinks during the first hour in the hydrogen current from about one hundred to twenty or ten per minute; then it continues to beat at this rate for ten hours or more. In this case one might believe that during the period of steady diminution of the tension of oxygen in the heart (during the first hour), the heart-beat sinks steadily while it keeps up at a