5l8 REPLACEMENT OF THE ANCIENT E. AFRICAN FOREST 



to grass-burning as the wish to cuhivate, and here (especially, 

 of course, in the less settled parts) great ring fires are lit 

 annually by the natives for the express purpose of hemming in 

 the contained game, which either breaks through and is shot at 

 or speared — or is burnt. I have known both koodoos and sables 

 to be caught and killed by the flames, in one case quite a num- 

 ber ; and these fires, as also fires lit for purposes of rat-hunting, 

 often involve considerable pieces of country. I should expect 

 that hunting man will have used this method ever since he knew 

 the use of fire and found himself in grass-veld. 



The pa])er on the Melsetter trees and shrubs, read in 1916, 

 and several times referred to here, remains unpublished owing 

 to my indisposition. In it I suggested the terms " pyrophyte " 

 and " pyrophobe," here used, for trees adapted to withstand fire, 

 and not so adapted, respectively, and gave in detail the succession 

 of pyrophyte associations that here follows the destruction of 

 forest. 



(Read, July 4, 1917.) 



NOTE ON THE MICRO-TITRATION OF ARSENIC. 



By Henr> Hamilton Green, D.Sc, F.C.S. 



(Abstract.) 



(Printed ill Ainntal Report of Director of Veterinary Research, 



Pretoria.) 



The difiiculty of determining small quantities of arsenic in 

 physiological material with any real ai)proach to percentage 

 accuracy is emphasized, and it is pointed out that 'for quantities 

 ranging from a milligram or two down to one-twentieth of a 

 milligram a micro-titration method is much more serviceable 

 than the commonly used Marsh mirrors or Gutzeit papers. A 

 method is described in which the arsenic is brought over as 

 arsine in the conventional way, collected in dilute silver nitrate, 



N 

 and titrated directly with iodine (i c.c. = one-tenth of a 



495 

 milligram AS2O3) after addition of a little bicarbonate and 

 sufficient potassium iodide to keep all excess silver salt in solu- 

 tion. Comparison is made with the reports of referees in the 

 most recent trials of the methods favoured (for food-stuffs) by 

 the Association of Official Agricultural Chemists in America, 

 and it is maintained that micro-titration is more reliable and 

 more rapid than colorimetric determination ; that it requires less 

 personal attention to detail, and is applicable in a great many 

 cases where most chemists now adopt a modified Gutzeit 

 method. 



(Read. .Inly 4. 1917.) 



