COBLEXTZ: REFLECTING POWER OF VARIOUS SUBSTANCES 449 



properties of the primary, viz., white lead. The combinations 

 of calcium sulfate and of barytes with white lead appeared much 

 darker than a similar combination with calcium carbonate. 



The white paints examined quantitatively for reflection were 

 samples which had been ground in .linseed oil. under standard 

 conditions. 1 There was no marked difference in the reflecting 

 power of white lead made by different processes (American, 

 Dutch, and French), the reflecting power of various samples 

 varying from 74.3 to 76.0 per cent. Similarly the samples of 

 zinc oxide paints reflected from 68 to 69 per cent. This agrees 

 with previous observations on the dry pigments which show that 

 lead carbonate has a higher reflecting power than zinc oxide. A 

 mixture of equal parts of white lead and of zinc oxide reflected 

 71 per cent. 



In view of the fact that the reflecting power of aluminum is 

 high (69 per cent) in the visible and in the ultra-violet, it was of 

 interest to determine whether sheet aluminum would be more 

 satisfactory than white paint for protecting a building against 

 warming by solar rays. The difference in reflecting power of 

 aluminum and white paint is not so marked in the visible spec- 

 trum as in the infra-red. Both materials absorb considerable 

 sunlight; but aluminum has the lower ernissivity for radiations 

 at 6 to 12fi. Hence it cannot radiate the absorbed energy as 

 rapidly as the white paint and its temperature rises above that 

 of white paint. 



This conclusion is borne out by the following experiments in 

 which the temperature of two similar plates of aluminum was 

 observed before and after one plate was painted with white lead 

 ground in pure linseed oil. Two sizes of sheet aluminum were 

 tested: A = 46 by 4S by 0.718 mm., B = 68 by 69 by 0.245 mm. 

 Thermocouples of copper and constantan wires were inserted 

 and riveted in holes drilled thru the aluminum plates, and as a 

 further attempt to obtain good contact the junctions were tight- 

 ened with Wood's alloy. The constantan wire was 0.31 mm. and 

 the copper wire was 0.40 nun. in diameter. The cold junctions 



1 Proc. Anier. Soc. for Testing Materials 11: 226. 1911. 



