112 REDUCTION OF VARIABILITY. 



males of last census, only ten had any off-spring whatever. So 

 these ten males averaged one hundred children each. And of 

 the five-hundred females, only fifty raised off-spring, they 

 averaged twenty young each. So that, after all, the number of 

 pigs out of the thousand of last year, which produced the thou- 

 sand of this year was sixty, or six per cent. Last year there 

 where six red pigs among the thousand. Two red young pigs 

 were born this spring, but they happened to be among the herd 

 of a farmer who sold out to the butcher. The chance of a re- 

 duction of the variability among the thousand pigs of last 

 generation, the chance that the potential variability of this 

 year's pig-population is smaller than last year's, looks greater 

 when we examine the facts, than when we accept the statis- 

 tician's calculation. 



While we are writing this, our cats are constanly coming and 

 going, bringing in field-mice, Microtis, for the kittens. They 

 live on these mice almost exclusively. The field-mice scamper 

 from under our feet in all the pastures, and hawks and weasels 

 and coyotes do as our cats do. But we know, that when the 

 drought has set in, the breeding will slow-up, and the cats and 

 the owls will begin to make appreciable inroads upon the num- 

 ber of mice. Very soon the cats will begin to take a renewed 

 interest in the kitchen. The statistician will tell you, that 

 the millions of Microtus in Strawberry canyon are descended 

 from as many millions that were living here last June, and 

 that on the average each one of these mice of last year has pro- 

 duced one of the mice of to-day. If we go out to hunt Microtus 

 in December, and find that they are decidely rare animals, we 

 get a different picture. How many dozen actually survive the 

 winter and are the parents and grand-parents of the June mil- 

 lions? How great a percentage is this number to the summer 

 number? Even if the numbers remain the same from year 

 to year, the group of survivors is so small, compared to the 

 group of mice from which chance has selected them, that the 

 potential variability of the group, if there was any, must have 

 been greatly reduced. (Fig. 16). 



