ON OFFICIAL STATISTICS. 431 



perhaps more than one, to go into the subject fully. ' Suffice it to 

 say that I believe I have made it clear that a marriage rate 

 suitable for purposes of comparison, can only be oljtained by 

 finding the relation between the number of marriages and the 

 number of single men at marriageable ages ; a similarly suitable 

 birth rate can only be obtained by linding the relation between 

 the number of births and the number of married women at the 

 child-bearing age ; and a similarly suitable death rate can only be 

 obtained by linding the relation between the number of deaths 

 at different ages, and the number of persons living at the same 

 ages. As the requisite data can only be got at or near a census 

 taking jjcriod, conclusive results cannot be arrived at at any other 

 time, and this is one reason out of many why the census should 

 be taken oftener than it generally is. 



It is much to be desired, that the causes of death should be 

 returned by medical men with more precision than they often are. 

 For example, in the case of women dying in childbed, it is very 

 common to certify that the death occurred from debility, ex- 

 haustion, blood-poisoning, pytemia, septictemia, phlebitis, embolism, 

 peritonitis, or hemorrhage ; the fact that the circumstance entered 

 was consequent upon child-bearing being kept entirely out of 

 sight. Dropsy, which is rather a symptom of a complaint than a 

 complaint itself, is often given as a cause of death, without any 

 mention of the primary disease. Such ambiguous terms as atrophy, 

 debility, old age, etc., are often set down, which should not be 

 done when it has been possible to detect the true disease. Brain, 

 chest, heart, liver, or stomach disease is frequently entered, when 

 the nature of the complaint might just as well be given. The 

 statistics of cancer and hydatids ai'e no doubt affected by returns 

 being made in this indefinite manner. Gunshot wound, poison or 

 drowning is sometimes set down without reference to whether the 

 death was accidental, homicidal, or suicidal. Such instances 

 might be multiplied. The officer charged with examining the 

 death registers should be instructed to refer back all doubtful 

 cases for more complete information. 



There are other points I should have liked to touch upon, did I 

 not feel that I was already trespassing too much upon your 

 patience. My Address is much longer than I intended it to be 

 when I commenced to write it, but I have been led on from sub- 

 ject to subject, and I was at any rate anxious not to miss any 

 subjects which might be really important. 



Victoria first, and New South Wales afterwards, have wisely 

 established departments, whose province it is to deal exclusively 

 with statistics, thereby following the example of France, Ger- 

 many, Italy, Holland, and especially Belgium. Thei'e are many 

 reasons why statistical work can be better done by the Govern- 

 ment than by private persons. Facts most important in statistical 



