No. 6. DEPARTMENT UF AGRICULTURE. 617 



the couibiniuj; ol' the two laltei' of ivplioid level-, uiplitliei-ia and 

 scarlet fever bacteria, and in one instance anthrax from deposits 

 from beh)w a railroad bridj^e. !r>ome of this sand was taken and 

 worked into mortar and plastering; the first nsed between bricks, 

 and the latter in the regular way on lathing. It was found that the 

 heat of a chimney soon made the mortar porous, and then nonco- 

 hesive; a common cause of defective flues and the destruction by 

 tire of happy homes. With the plastering, to prevent cracking on 

 the wall, it was necessary to mix horse or bullock hair. Experi- 

 mentation proved that the water and lime used did not destroy the 

 disease breeding bacteria, though in most instances the microbes 

 remained inert until the plastering had been thoroughly warmed. 

 This may explain why, in some homes, contagious diseases have ap- 

 peared — the first in the neighborhood, or even in a county — to bring 

 sorrow and woe to an otherwise happy household. 



SAND FOR FILTRATION. 



What has been stated applies with equal force to such material 

 being used for filtering water. Several large cities have used river 

 bottom sands for filtration, and though the sands were washed, 

 using 300 parts of water to one of sand, it was shown conclusively 

 that disease breeding bacteria still remained. -Nearly two years 

 ago I was employed to visit and examine, and report upon, every 

 municipal sand filtration plant in North America. The most alarm- 

 ing feature connected with the whole work was the poor quality of 

 sand used in nearly every instance. Notwithstanding, there was 

 an invariable improvement in the water supply, though the mor- 

 tality list would have been still smaller everywhere with a first 

 quality silica sand. Washington, D. C, will soon have the model 

 sand filtration plant, wherein all the defects of foreign and Ameri- 

 can filtration will be happily remedied. 



For all the evils mentioned the best sands of Pennsylvania can 

 effect a sure cure. Samples have been taken from the crests and 

 sides of hills or mountains in more than a dozen counties, with re- 

 sults as follows: Crests of Cove Mountain in both Fulton and Frank- 

 lin counties 99.53 per cent, silica; Chestnut Hill, two miles north of 

 Columbia, Lancaster county, 99.51 per cent, silica; one-half mile 

 north of Mountville, Lancaster county, 99.00 per cent, silica; Welsh 

 Mountain, one and one-half miles nort4iwest of Honeybrook, Ches- 

 ter county, 98.27 per cent, silica; three and one-half miles northwest 

 of Wrightsville, York county, 99.04 per cent, silica; three and one- 

 half miles northwest of Lewistown, Mifflin county, 98.84 per cent, 

 silica; Sinking Valley, Blair count}^, 94.89 per cent, silica; near 

 Greenville, Mercer county, 95.16 per cent, silica, and 3.22 per cent, 

 alumina; near Duncannon, Perry county, 98.65 per cent, silica; Tus^ 

 49 



