288 GARY N. CALKINS 
other hand, the differences are so well marked that they may be 
seen at a glance. Chart 1 shows the full life-cycles of eighteen 
series and the history to date (January 24, 1920) of four series 
now living. Each of the eighteen curves represents the vitality, 
as measured by the division rate, of a single series from its start 
as an ex-conjugant until death from exhausted vitality. Each 
series 1s composed of five lines of individuals, each line comprising 
one of the first five individuals formed from the ex-conjugant 
and its progeny by division. Each curve represents the average 
division rate of all five lines of a series for periods of sixty days 
and each curve shows the time of origin of filial series, if there 
were any. The data on which the curves are based are obtained 
by: averaging the daily division rates of all five lines for ten-day 
periods. These ten-day averages are then averaged for periods 
of sixty days and the results are plotted. The first record of a 
curve represents the average division rate for the first six ten-day 
periods; the second record represents the average division rate 
for sixty days beginning with the second ten-day period; the 
third record represents the division rate for sixty days beginning 
with the third ten-day period, and so on for the complete life- 
eycle. Obviously there are as many sixty-day records as there 
are ten-day periods less five and the last five are filled out on 
the assumption that the division rate was zero. 
It is quite obvious from diagram 1 that extreme differences in 
vitality are manifested by some of the series. ‘Thus the Q series 
was a weakling as compared with any of the others, and the R 
series and a series were relatively weak as compared with most 
of the others, while the W series has shown a low grade of vitality 
_from the start. The strong series, on the other hand, are indi- 
cated. by the number of ten-day periods which lie above the line 
marking the rate of ten divisions in ten days. The curves, how- 
ever, afford very poor evidence of the difference in relative vitality 
of the different series, and for this the data must be more carefully 
analyzed. 
These differences in vitality may be due either to, 1) chance 
variations in different ex-conjugants; or, 2) to old-age weakness 
of parents at the time of conjugation; or, 3) to accumulating 
