INFLUENCE OF INFANTILE MORTALITY ON BIRTHRATE. 247 
representing the weighted results. Seeing that the origin 
of the curve must necessarily be the zero of both co-ordin- 
ates, the point 0, curve A, is not expressible by any simple 
relation capable of rational interpretation, much less is 
-curve B in Fig. 2. Moreover, equations (6) and (9) though 
well representing the relation within certain limits, cannot 
be regarded as quite general, forasmuch as they imply that 
when # falls to 8, there would be no infantile mortality 
whatever, whereas actually » and 6 must become zero 
together, and further it is not essential that for any value 
of 6 there must be a corresponding value for p. 
ne 5 Ch SSSR SELES GSA TTETES AMS ER SE HCCC re CaCO IROL fare 331 
Birthrates, i.e. births per 1000 of the population. 
0 50 100 150 200 250 300 250 
RATES OF INFANTILE MORTALITY. 
Fig. 2. 
Propositions (i) and (ii) in section 4 are consequently. 
true only for individual countries, and between the 
restricted limits of actually occurring rates for their entire 
populations. 
9. Subject to the defects pointed out in the preceding 
section (3rd paragraph thereof) it will be possible to get 
some idea of the way in which infantile mortality of any 
