The cells eluted at the end of the first generation of growth on the surface were progeny of the youngest cells in the culture at the time of labeling. Thus, the radioactivity in the newborn cells eluted during the first generation reflected the rate of thymidine incorporation during the division cycle of their parents, in reverse. Similar analysis applied to subsequent generations of elution (see Helmstetter, 1967 for a detailed illustration of this reasoning). What a feeling it was to suddenly realize how powerful this application of the technique would be to study a host of cycle-dependent phenomena with virtually no disturbance of the cells. As with the chicken epiphany, I was fully convinced it would work because all treatments, including labeling, would be done in untouched cultures. All we were asking was for the cells to divide on the membrane surface in the same sequence as they would have divided in the original exponentially growing culture. The first experiment was done the next day. I pulse-labeled a culture with radioactive thymidine, filtered it onto the membrane in the usual way, and I then had to head off to teach a class while a technician collected the samples and prepared them for counting. That experiment failed because I had added too much label, but it was repeated the next day with great success showing unequivocally that in glucose-grown cells, the rate of DNA replication appeared constant for the first half of the cycle, doubled at mid-cycle, and was constant at twice the rate for the second half of the cycle.