survival index. That the decreases need not be abrupt may be seen in 
figure 5. Here successive mortality rates (per cent per annum) were 
set up as 39.3, 38.91, 38.21, 37.11, 35.41, 32.89, 29.41, 25, 20, 15, 
10.59, ete. The resulting survival index is 60 per cent for 15 -con- 
Table 12.--Values of Survival Index for da x 100 
dy) 
dj and do = deaths in two successive years; 8) and-85 = survival rates 
same periods; q) and do = corresponding mortality rates. 


60.0 90.0 135 210 360 810 — 90 
10 10.0 22.5 38.6 
20 8.9 20.0 34.3 53.3 80.0 120 187 320 720 80 
30 7.8 17.5 30.0 6.7 70.0 105 163 280 630 70 
ho 6.7 15.0 25.7 0.0 60.0 90 10 24,0 540 60 
50 5.6 12.5 21.4 33.3 50.0 75 117 200 4,50 50 
60 hel) 10.0 17.1 26.7 0.0 60 93.3 160 360 4,0 
70 3.3 7.5 12.9 20.0 30.0 4s 70.0 120 270 30 
80 2.2 5.0 8.6 13.3 20.0 30 6.7 80 180 20 
90 11 2.5 3 6.7 10.0 15 23.3 40 90 10_ 
30. +80 70 +460 SO QO 36  #=+%F  i5 

secutive age intervals. While this result is identical to a time- 
specific survival rate (= survival index in the sense of this paper) 
reported for the lapwing by Kraak, Rinkel, and Hoogerheide (190), 
it seems to me improbable that such.a peculiar set of statistics will 
be frequently found to cover such a long span of time. The situation 
should be watched for in shorter periods of observation. 
Dynamic life-table analyses.-~In table 13, I have set up 
another hypothetical population in an effort to explore the differ-. 
ence between dynamic and time-specific analyses. Here three sets of 
mortality sources are identified as occurring in consecutive order 
each year. One source represents a fluctuating factor: 65 per cent 
of the first-year birds alive at the start of a hunting season are 
shot; at subsequent age intervals, only 50 per cent are shot. Under 
the conditions of this hypothesis, several additional facts about 
these methods emerge: 
(a) In populations with a fluctuating mortality factor 
confined to one part of the year, sampling time importantly affects 
3fhe successive mortality rates (q5).can be found by starting with a 
given mortality rate for the first year (a) and developing a series 
with the formula Qa) = 4% 
8; + 92 
38 
