134 

 TABLE OF AVERAGES. BRITISH MEASURE (continued). 



METHODS ADOPTED. 



The methods adopted in the Natal tests were generally similar to those followed in 

 the admirable series of experiments conducted at the Watertown Arsenal by Mr. S. P. 

 Sharpies for the United States Government, the results of which are recorded in Prof. 

 Sargent's Report on the Forests of North America (Tenth Census U.S., Vol. IX.). 



SPECIMENS TESTED. 



The specimens used were mostly cut in May or June 1889, then stacked and 

 seasoned for some months in a well-ventilated shed in Durban, before testing in 

 January-February 1890 ; they were thus seasoned eight months on an average, and the 

 seasoning was more thorough than would usually be the case in the same interval. Dry 

 pieces of Stinkwood, Ironwood, and White Pear from the Knysna forests, and of Flat- 

 crown from Victoria County, were also tested. 



With few exceptions, the results obtained do not give a measure of the properties of 

 timber of the best quality. The material disposable was too limited in quantity to admit 

 ef a selection, and much of it was cut from small trees scarcely mature. For each 

 apecies, the pieces of wood procured from each locality recorded in the tables were 

 cut from the same tree, and the results of the various tests are thus comparable. 



DETEKMINATIONS OF DENSITY AND HARDNESS. 



Timber retains a very variable proportion of water, according to the mode and length 

 of seasoning ; and to have an accurate meaning, the density determinations should be 

 made on pieces dried artificially till they cease to lose weight, but it would have been 

 difficult to adopt this method, and the density of the air-dried wood was alone taken. 



The specimens used for the purpose were made exactly 4 centimeters square and 32 

 centimeters long ; they were weighed and the density calculated. The results are given 

 in Table I. 



