33$ AGRICULTURAL STATISTICS. 



no loophole of escape. They ought also to be so worded as not 

 to give offence or create suspicion. 



The public must have the positive assurance that all informa- 

 tion given is strictly confidential, and will on no account be 

 divulged by the statistician to anybody of whatever standing. For 

 that purpose the statistician ought to be supplied with a proper 

 safe, of safe receptacle for the safe-keeping of all forms and 

 papers. In the Transvaal we have already been approached for 

 information relative to the assets of individuals, which informa- 

 tion we have courteously, but most emphatically, refused. 



When a country begins collecting and compiling agricultural 

 statistics, the first year's figures can only be taken as approxi- 

 mately correct. As years go on and the work progresses, and the 

 foundations on which to work become more and more settled, the 

 figures also approximate nearer and nearer to accuracy. Work 

 for accuracy with the individuals and the sum total will be right. 

 It may seem illogical, but it is a fact, that individual members of a 

 whole group may vary considerably, whereas the whole 

 group varies very slowly. There may be a difference in 

 yield on one farm or another or even in one district or another. 

 A farm may produce considerably less than the previous year, 

 but then there will be another farm that will make up for it in 

 producing considerably more, and thus there will not be much 

 difference in the total. 



Another thing that must also be borne in mind is, that the 

 people have to be educated up to the fact that the periodical col- 

 lection of agricultural statistics is necessary. When we started 

 in the Transvaal — this was in October, 1908 — we had the farmers 

 against us. They were prejudiced against the idea and had the 

 most wonderful notions about it. I visited the different wards 

 and held meetings with the object of convincing the prejudiced. 

 When explaining to them the objects of the collection of agricul- 

 tural statistics I put the matter to them in this way: "What do 

 you think of a farmer who is unable to tell you how many head 

 of cattle he possesses, or sheep, or other stock, or how many bags 

 of grain he has reaped?" The reply invariably was, that such a 

 man was not worthy the name of farmer. To which I would add, 

 " Yes, and suppose any of you were to meet General Botha and 

 say to him. ' General, you are Minister of Agriculture. Can 

 you tell me how many head of cattle and sheep, etc., you have in 

 the country, or how many bags of mealies the country produces?' 

 and he said, ' I don't know.' ' This was generally convincing. 

 With such and other arguments I found no difficulty in persuad- 

 ing them of the value and desirability of reliable statistics. 



The proper season for the collection of agricultural statistics 

 differs. Over the greater part of South Africa the months of June, 

 July and August would be the best. For tropical fruits, tea and 

 sugar, perhaps another season will have to be selected. 



The winter months offer themselves as the best, being the 



