56 Report S.A.A. Advancement of Science. 



of burning to one of an opposite character. The secret of the trouble 

 was probably improper sampling. The seams of coal in Natal are 

 unusually intermixed with bands of shale or half-burnt coal, and 

 consequently the excellence of, say, a hundredweight or a ton sample 

 depends upon the care with which this inferior coal has been picked 

 out. The sampling is just as important as the actual analysis, and 

 before the final quantity is taken from the larger bulk sample, the 

 latter should be crushed to the size of about, say, small nuts and 

 graded according to its specific gravity. 



I am aware that shales and half -burnt coal vary greatly, and 

 results of specific gravity would not fairly compare one class of coal 

 with another, but it would be the best guide as to w^hether the sample 

 from any particular colliery was a fair one. The matter is one 

 calling for attention most urgently, at the present time. 



Consumers of coal in large quantities would view coal tests sub- 

 mitted by collieries with much greater confidence, if these tests w-ere 

 carried out in a central institution. In fact, the whole Question of 

 both coal sampling and coal testing, is one which calls out for atten- 

 tion, and is a work which should be undertaken immediately by such 

 an institution, and would provide occupation for many months of 

 exhaustive investigation. 



Natal and South African coals seem to differ in many important 

 respects from other coals, and I believe that the present methods used 

 for coal analysis in England might be subjected to considerable modi- 

 fication in their application to the mvestigations of the qualities of 

 local fuel. 



Mr. Carl Hall and Dr. J. Moir also contributed to the 

 discussion. 



