58 



PAUL J. ANDERSON 



the roads muddy from hard rains; (3) it does not look or feel 

 like road dust; (4) it gives an alkaline reaction with phenol- 

 phthalein, which is not true of any other dust sample collected 

 in that county; (5) its chemical composition (given below) shows 

 it to be a partially burned dust. 



A study of the mills in operation showed that there were two 

 main sources from which a large amount of dust came: (1) The 

 crushing, pulverizing and drying mills, where the raw mixture of 

 limestone and shale is reduced to an extremely fine, dry powder 



TABLE I 



Chemical analyses of dust samples 



*N umber of sample. . . 



Silica (SiOa) 



Iron and alumina 

 (FeoOs + AI2O3).... 



Lime (CaO) 



Magnesia (MgO) 



Sulfur trioxide (SO3).. 



Loss by ignition 



Total 



Carbon dioxide (CO2). 



Insoluble (in hot HCl) 



ATLAS 



POBTLAND 



CEMENT 



23 .50 



10.04 

 62.08 

 1.23 

 1.63 

 1.30 

 99.78 

 not de- 

 term- 

 ined 

 not deter- 

 mined 



* Sample book of the chemical laboratory of the Atlas Portland Cement Com- 

 pany, Hudson, New York. 



(technically called composition) ; (2) the stacks of the kilns in 

 which the composition is burned and fused to clinker (unground 

 cement). The second is undoubtedly the principal source of the 

 dust on the foliage because: (1) The strong draft in the kilns, 

 produced by the burning, powdered coal and air forced in below 

 under a pressure of 100 pounds to the square inch, must neces- 

 sarily carry some of the dry, fine composition dust out of the 

 top of the stacks; (2) the distance to which the dust is carried 

 points rather to the high kiln stacks and forced ejection rather 



