MANIFESTED IN HUMAN SOCIETY JOT 



and New Zealand have very low rates, as we shall 

 presently see. 



Now, the connection between birthrate and deathrate 

 is only indirect. Let us suppose that a certain municipality 

 provides an excellent supply of pure milk and thus cuts 

 down the rate of infant mortality. As this will not 

 affect the adults, there is no reason why it should affect 

 the birthrate. And now let us suppose that another 

 municipality, through some neglect, causes a considerable 

 increase of infant mortality which does not affect the 

 birthrate either. This will cause a big discrepancy 

 between the birthrate and deathrate correlation for the 

 two towns. Yet when the rates for the country as a 

 whole are worked out these two factors will counter- 

 balance and cancel each other. That is why the operation 

 of the principle will always be more clearly seen in com- 

 parisons between large areas, particularly between different 

 countries. It will manifest itself through the law of 

 averages. And while it will be true that the same com- 

 bination of causes which leads to a high deathrate will 

 lead also to a high birthrate and vice versa, yet there 

 will always be a considerable number of factors which 

 affect one without immediately affecting the other. But 

 the law of averages will ensure that these factors in the 

 long run will balance each other, and thus we shall get 

 comparatively uniform correlations in comparisons between 

 large areas or long periods of time, coupled with many 

 irregularities in comparisons between small areas or short 

 periods. 



This is what we do find. The correlation in com- 

 parisons between countries is remarkably uniform consider- 

 ing the imperfection of our statistics, while the correlation 

 between towns, though showing the operation of the 

 law in a general way, is very irregular. The regularity 



