430 PROCEEDINGS OF SECTION F. 



larger than it really is. As an illustration, I may mention that 

 observing the returns, year after year, showed the infantile death- 

 rate to be much higher in one of the colonies than in any of the 

 others, I drew attention to the fact in a paper which was read 

 before the leading scientific society in that colony. This occa- 

 sioned much excitement and some alarm, and my paper was 

 followed by several othei's on the subject — conti'ibuted for the 

 most part by medical men — in which various unconvincing theories 

 were propounded to account for the evil, and the matter was also 

 the occasion of editoiial articles, as well of correspondence in the 

 public journals. This occurred ten years ago, and it has only 

 recently been discovered that the high infantile mortality sup- 

 posed to exist had no being in reality, but that the erroneous 

 supposition had arisen from the fact that the births were so 

 imperfectly registered that the proportion of infants that died 

 appeared much larger than it really was. More attention has 

 since been paid to the registration of l)irths in the colony referred 

 to, and although it is not yet perfect, tlie infantile death rate 

 derived from the recorded data has declined in a very sensible 

 degree, and now differs but little fi'om that of the other 

 colonies. This is one of many instances which might be adduced 

 to show how defective statistics may cause a wrong impression to 

 be formed respecting a country, and may lead to its being con- 

 demned as a field for emigration, or may possibly injure it in 

 other ways. 



It is the practice in all countries in which marriages, births and 

 deaths are registered, to find the projoortion of each of these to 

 the total population in order to compute what are termed the 

 marriage, birth, and death rates ; yet in making comparisons 

 between different countries or between the same country and 

 itself at different periods, results ai"e arrived at which are often 

 seriously misleading, for it must be evident that the rates will be 

 largely affected by the component parts of the populations Mdiich 

 are being dealt with. Thus, other things being equal, the mar- 

 riage rate will be low in a country where the proportion of 

 children, or of persons already married is great ; the birth-rate 

 will be low in a country where the proportion of women at the 

 reproductive period of life is small ; and tlie death rate will be 

 low in a country where the bulk of the inhabitants are in the 

 prime of life. I may remark that new countries, in regard to the 

 component parts of their populations, being subject to rapid 

 changes, and in quite a different condition from older countries, 

 are those in which the mari'iage, birth, and death rates are the 

 least reliable for purposes of comparison. 



I have written much on this subject elsewhere, and have referred 

 to it under the head of Vital Statistics in the last issue of my 

 " Victorian Year Book," but it would take a whole paper, or 



