Appendix table 1. — Data for the regression of percentage of the kill from a year class at ages 3 

 and <+ taken at age 3 on date of termination and median date of the kill at age 3, year classes 

 1947-61, St. Paul Island 



For the 1962 year class the median date of the 

 kill of 3-year-old males is 27 July and the 

 terminal date is 9 August. Thus, for the 1962 

 year class p^ is estimated as 65.1 (in the actual 

 equation the median date was estimated by 

 interpolation to tenths of a day; for the 1962 

 year class, m = 6.7). The kill at ages 3 and 4 

 from the 1962 year class is estimated as 

 29,200, i.e., 19+0.651; hence the estimate of 

 the kill of 4-year-old males is 10,200. The 

 standard error of the estimated p^ for the 

 1 962 year class is 9.6. This value is equivalent 

 to an error in the forecast total of about 4,500. 



Regression of returns on air temperature . -- 

 Appendix table 2 shows the data for this 

 regression, i.e., the mean tennperature at 

 St. Paul Island for the 12 months ending 30 

 June of the year of birth, and the total kill from 

 the year class at ages 3 and 4. Use of the total 

 kill at ages 3 and 4 represents a modification 

 of the dependent variable previously used. 

 Previously, the dependent variable was the 

 kill of a year class prior to 3 1 July at ages 

 3 and 4 plus 80 percent of the kill of 3-year- 

 olds in August. This "adjusted" kill was used 

 to obtain comparable data from early years 

 when the kill ended prior to 31 July and from 

 recent years when the kill usually has been 

 extended into August. 



The adjustment, however, made it difficult 

 to evaluate the error of this forecast. More- 

 over, the analysis given in the previous section 

 shows that the terminal date plays a small role 

 in the percentage taken at age 3 and presumably 



Appendix table 2. --The kill of 3- and 4-year- 

 old males and mean air temperature, year 

 classes 1950-61, St. Paul Island 



a still smaller role in the total kill from a year 

 class. 



The regression equation is: 



Y = l6.6+1.05T r = 0.81 



For the 1962 year class T = 21 andhence: 



Y = 38.7 



27 



