418 PROCEEDINGS OP SECTION P. 



been able to subdue all difficulties, I have conquered many, and 

 others will no doubt disappear if the efforts for their removal are 

 continued. 



I trust it will not be thought for a moment that any observa- 

 tions I may make are ofiered in a captious or fault finding spirit, 

 or with a wish to dictate to others who may be quite as well 

 versed in the subject as myself. I know my own work is far from 

 pei'fect, and I am always thankful for any hints which may show 

 me the way to make it more nearly so, and glad to adopt sugges- 

 tions for its improvement so far as they commend themselves to 

 my judgment. I shall make my remarks as general as possible, 

 and shall refrain from naming any of the colonies except where 

 absolutely necessary, and then never to their disparagement. My 

 only wish is that what I say may have the effect of pi'omoting uni- 

 formity between the statistical records of the various colonies, of 

 causing them to chronicle more completely and faithfully the 

 events which are passing from day to day, and to describe with 

 greater accuracy the progress and condition of the land in which 

 we live. 



Of all the duties the official statistician is called upon to under- 

 take, the census is the most important and the most arduous. 

 The census is the foundation and starting point of all statistics, 

 the basis of legislation, and the groundwork of the calculations of 

 actuaries, financiers, merchants, and many others ; therefore, 

 every pains should he taken to make it a success. All prepara- 

 tions should be designed, thought out, and perfected beforehand, 

 and all precautions should be taken that no weak point exists in 

 any of the arrangements. A good census is an inestimable 

 advantage to a community, whilst an incorrect one is much worse 

 than none at all. 



As population is always on the move, it is essential to accuracy 

 that the census should refer to one particular day, and the opera- 

 tion ought to be performed, except in scattered country districts, 

 in three days, one day being devoted to depositing the schedules 

 and two days to collecting them. In country districts more time 

 must be taken, but except under very exceptional circumstances 

 no more than five days should be allowed for the work. Much 

 too long a time is sometimes taken over a census, which arises 

 from the districts being made too large, and too few collectors 

 being appointed. This makes it certain that some of the popu- 

 lation will be missed, to remedy which enumerators are apt to 

 make guesses and estimates, some of them being based on imper- 

 fect data and some on no data at all, and an incomplete and 

 inaccurate census is the result. 



I may here observe that although the Australian colonies are 

 far from perfect in this respect, every one of them is much in 

 advance of the United States, where one month is nominally 



