EVALUATION OF WOOD PRESERVATIVES 491 



reported by Cislak in discussion following the presentation of the Snell 

 and Shipley paper,^°^ and would approach those obtained by Boving in 

 his impregnated filter paper experiments (see Table XXXI). Further- 

 more, the results of tests such as his, which involve the principle of 

 inoculation by placing "fungus on wood," instead of ''wood on fungus" 

 as in the soil-block test, require a lot of translation to interpret their 

 significance in practical wood preservation. Rabanus^^ brought this mat- 

 ter out into the debate very clearly 20 years ago. Liese et af^ answer 

 Breazzano's objections to the block test. 



There have been no further reports on the procedures followed by 

 Hopkins and Coldwell. Their methods are subject to some of the same 

 criticisms that have been mentioned in the preceding paragraphs, partic- 

 ularly if one were to consider such techniques in a search for a way to 

 speed up the culture tests. However, whether one agrees with them or 

 not, their introduction of the idea of applying strength tests leads di- 

 rectly to a discussion of strength losses, as against weight losses, as 

 criteria for establishing preservative thresholds. 



Toughness or Impact Tests for Determining Preservative Effectiveness 



Trendelenburg^^ ^ published in 1940 his scheme for testing the strength 

 of treated blocks that had been exposed to fungus attack. He was aiming 

 at a technique that would shorten the time period of fungus tests on 

 wood, but he was also looking for some other criterion than weight loss 

 as a measure of fungus attack. He used the relative impact strength 

 values of matched sound and decaying test specimens as indices of the 

 degree of decay. Boards were carefully quarter-sawed first into pieces of 

 double specimen width plus saw kerf tangentially, and of double speci- 

 men length. From these blanks four test specimens were cut that meas- 

 ured 8.5 X 8.5 X 120.0 mm. The pairs were considered to be matched 

 laterally and vertically since every effort was made to cut them from 

 the same annual rings. One piece from each pair was exposed to fungus 

 attack and the other served as a control. The fungus cultures were made 

 in Kolle flasks on malt agar. Specimens were placed radial side down 

 directly on the growing fungus surface. In the pendulum testing machine 

 the impact load was always appUed to the upper radial face, so that the 

 lower more or less infected radial face represented the tension side of the 

 specimen in the breaking test. 



Trendelenburg showed that the per cent change in strength caused 

 by the test fungus in the early stages of decay was much more pro- 

 nounced than the concomitant changes in weight or density. He called 



