106 PAINT TECHNOLOGY AND TESTS 



made on a comprehensive basis and placed under the control of 

 scientific and technical societies of renown, so that they might 

 be fair and unbiased from every standpoint. In order to secure 

 a comparison of the wearing of different paint formulas in various 

 sections of the country and under differing climatic conditions, 

 another series of tests was started in the East soon after the 

 North Dakota tests had been started. Simultaneously fences 

 were erected at Atlantic City, N.J., and Pittsburg, Pa. The 

 site of the Atlantic City fence is a strip of land running due 

 north from Atlantic and Savannah Avenues and within a short 

 distance from the Atlantic Ocean, the exposure being a severe 

 one. The site of the Pittsburg fence is back of the athletic field 

 of the Carnegie Technical Schools, the fence running east and 

 west and being exposed to the heavily charged sooty atmosphere 

 coming from the many industrial plants near by. 



Construction of Framework of Fences. At these two loca- 

 tions framework fences were built, upon which were placed a 

 series of painted panels. Heavy yellow pine posts six inches 

 square were set in the ground about six feet apart and to the 

 depth of about four feet, upon a concrete base. The posts were 

 solidly tamped and then braced at the top with supplementary 

 studding braces two inches thick. Connecting the posts was a 

 line of studding six inches by two inches, forming a solid frame- 

 work, the bottom of which was approximately fifteen inches from 

 the ground. The bottoms and tops of the fences were protected 

 by heavy boards two inches thick, so that the moisture and rain 

 might be prevented from working itself up into the wood. The 

 whole fence was sheathed with twelve-inch planed white pine, 

 thus forming a solid background for the test panels. 



Lumber for Panels. The lumber for the test panels was 

 most carefully selected, being of three grades white pine, 

 yellow pine, and cypress. A large amount of each grade of 

 lumber was secured, and after the best portion had been made 

 up into panels, the panels were inspected by an expert lumber 

 classer; nearly 40% being rejected on account of the presence 

 of knots or sappy places which appeared upon the surface. 

 Each of the panels finally passed upon as suitable for the test 

 was branded with a hot iron with consecutive numbers running 

 from 1 to 186. The grade of wood used for each panel was 

 indicated by an abbreviated mark W for white pine, C for 



