12 BULLETIN" 145, U. S. DEPARTMENT OF AGRICULTURE. 



times as much so as coal-tar creosote. Sapwood antiseptic was 

 almost nontoxic. 



Preservatives designated as copperized oil, fuel oil, Timberasphalt, 



kerosene, and N. S. Special had very low toxic properties. 



CORROSIVE ACTION OF THE PRESERVATIVE ON FLANGE STEEL. 



Of the various preservatives tested, coal-tar creosote, copperized 

 oil, the fractions of coal-tar creosote, the various carbolineums, the 

 water-gas-tar products, fuel oil, Timberasphalt, and kerosene had 

 very slight corrosive action on steel. In practical operations such 

 action can very probably be neglected. 



The metallic salts, except cresol calcium and sodium fluoride, were 

 much more pronounced in their action than coal-tar creosote, so that 

 the depreciation in plants using them would seem to be much greater. 

 The corrosive action of cresol calcium is much less than the other 

 salts tested. Zinc sulphate was much less corrosive than zinc chloride. 



The very marked corrosion of hardwood tar, Preservol, and tar 

 from Douglas fir is probably due to the comparatively large amount 

 of acetic and other organic acids which they contain, while the some- 

 what less corrosive action of Spirittine and beechwood creosote are 

 probably due to smaller quantities of these acids. 



In general, wood-tar products were much more corrosive than coal- 

 tar or petroleum products. This property may, however, be largely 

 eliminated in the future development in methods of refining these oils. 



DISCOLORATION OF PAINTED WOOD. 



All of the oils tested rendered the wood unfit for subsequent paint- 

 ing. 1 Copperized oil was least objectionable in this respect. If 

 thoroughly dried after treatment, so that excess oil would not appear 

 on the surface, it is possible that wood treated with some of these 

 preservatives could be satisfactorily painted with dark pigments. 



The water-soluble salts were all satisfactory (except sodium silicate 

 and cresol calcium) in that they caused no discoloration and appar- 

 ently no deterioration of the painted surface. From the results of 

 tests thus far made it is likely that woods treated with zinc chloride 

 and many water-soluble salts can be successfully painted. 



Cresol calcium caused discoloration of the paint; sodium silicate 

 reacted with the oil of the paint film, causing it to lose its adhering 

 properties. 



Additional tests on the effect of zinc chloride and sodium fluoride 

 on white-lead paint were made. Specimens exposed to the weather 

 were in good condition after 2 \ months, when they were accidentally 

 destroyed. Specimens placed in moist air were in good condition 

 after 1 year and 4 months. No difference was noticeable between 



1 This applies, of course, only to the method of painting herein described. 



