356 Philippine Journal of Science m& 



Thus the durability or resistance of timbers to the attack of 

 termites is not effected by hardness or weight; that is, the im- 

 munity of timbers is not due to their physical properties. 



RELATIONSHIP BETWEEN THE RESISTANCE AND THE CHEMICAL 

 PROPERTIES OF TIMBER 



Snyder 7 states that the immunity or relative resistance of 

 ironwocd is not due to hardness — since Asiatic termites attack 

 the hardest wood, lignum-vitas — but to the presence in the wood 

 of a substance (oils or alkaloids) repellent or distasteful to 

 termites. He also states that the presence of tyloses or of gums 

 may be factors in determining the durability and resistance of 

 hardwood species. It has been proved that the physical prop- 

 erties of woods are not the real factor of resistance as suggested 

 by that author. However, another suggestion of Snyder, that 

 the presence of certain chemical substances in woods is the 

 true factor, is somewhat dubious. As he gives no data in detail, 

 it is hard to understand what are the principal ingredients and 

 how they act in preventing the damage. 



In order to determine the relationship between the resistance 

 and the chemical properties of woods, I made the following 

 investigation : 



The quantity of ash and benzene extract in the above-men- 

 tioned woods was measured and compared with the percentage 

 of damage obtained at Tainan. 



Sampling. — A block of wood is cut along the three planes 

 crossed at right angles, and the sawdust is collected. By the 

 diagonal method reduction of the amount of the sample is made 

 repeatedly, until about 20 grams of the sawdust are obtained. 

 This sample is exposed for one hour in an air bath at 105° to 

 110° C. Then it is kept in a desiccator as a representative 

 sample. 



Method of estimation of ash. — A 1-gram sample is heated to 

 redness in a platinum crucible over a Bunsen burner, and the 

 incombustible substances are brought to a constant weight. 



Method of estimation of benzene extract. — To estimate the 

 benzene extract 5.00 grams of the sample are put into a Soxhlet 

 apparatus and extracted with pure benzene for from six to 

 ten hours. The benzene is driven off on a water bath, and 

 afterwards the container is heated in an air bath at 120° C. 

 to a constant weight. All determinations are made in duplicate. 



' Bull. U. S. Bur. Ent. 94 (1916) 79, 80. 



