BULLETIN 145, U. S. DEPARTMENT OF AGRICULTURE. 



submersion was taken to indicate its corrosion. All depositions on 

 the surface of the metal were removed as nearly as possible with a 

 rubber "policeman" each time the preservative was changed. At 

 the end of the tests, where electrolytic deposition of metal had taken 

 place, the deposited metal was removed by acid and its amount 

 determined by an analysis of the acid solution. The weight of depos- 

 ited metal thus determined was added to the loss of iron, and this 

 total represented the total corrosion. The figures thus obtained were 

 then calculated to ounces of steel corroded per square foot of surface 

 exposed per week. 



PAINT TESTS. 



The treated wood was first air-seasoned for about 1 month to per- 

 mit water and light oils to evaporate, and then coated with white 

 paint (30 pounds of lead oxide to 1 gallon of linseed oil), noting the 

 color change which subsequently took place. This test was made in 

 order to determine the practicability of painting treated wood. 



The results of the tests on wood treated with water-soluble salts 

 were favorable, and the scope of the tests was therefore enlarged. 



Specimens of noble fir, 1 by 4 by 12 inches, were treated with zinc 

 chloride and sodium fluoride, and painted with white-lead paint. 

 Untreated specimens were also painted for comparison. Four speci- 

 mens were used in each set. One specimen from each series was 

 placed in running water or in the moist atmosphere in the fungus pit, 

 and two exposed to the weather. 



These tests also indicated that wood treated with these salts could 

 be successfully painted, and arrangements were then made with the 

 National Paint Manufacturers' Association for a more extensive test. 



White pine panels, 2 by 3 feet by 1 inch, were treated with zinc 

 chloride, sodium fluoride, and ammonium phosphate. Six panels 

 were treated with each preservative and six were used untreated. 

 These were sent to the Institute of Industrial Research, Wash- 

 ington, D. C, where they were painted under the direction of Mr. 

 Henry A. Gardner, representing the National Paint Manufacturers' 

 Association. After painting, the panels were divided into three sets — 

 one of which is exposed at Atlantic City, N. J., one at Washing- 

 ton, D. C, and the third at St. Louis, Mo., under the direction of 

 Dr. Hermann von Schrenk. 



RESULTS. 



The results secured from the experiments may be changed by sub- 

 sequent tests, since the field covered is new and errors in manipula- 

 tion, not at present apparent, might exist. Every feasible precau- 

 tion was taken, however, to avoid errors, and those known and 

 uncontrollable are mentioned. 



