FIRE TESTS OF NEW YORK BUILDING STONES 



27 



were damaged the most by cracking very irregularly around the 

 individual mineral constituents [pi. 11, Peekskill ; 15, Grindstone 

 island; and 17, Northville]. Naturally, such cracking of the stone 

 in a building might cause the walls to crumble. The cracking is 

 due, possibly, to the coarseness of texture and the differences in 

 coefficiency of expansion of the various mineral constituents. Some 

 minerals expand more than others and the strains occasioned there- 

 by will tend to rupture the stone more than if the mineral com- 

 position is simpler. This rupturing will be greater, too, if the rock 

 be coarser in texture. For example, a granite containing much 

 plagioclase would be more apt to break into pieces than one with 

 little plagioclase for the reason that this mineral expands in one 

 direction and contracts in another, and this would set up stresses 

 of greater proportion than would be occasioned in a stone contain- 

 ing little of this mineral. The fine grained samples [pi. 12, Nyack; 

 and pi. 14, Grindstone island] showed a tendency to spall off at the 

 corners. The gneiss [pi. 16, Little Falls] was badly injured. In 

 the gneisses the injury seems to be controlled by the same factors 

 as in the granites, but there comes in here the added factor of 

 banding. Those which are made up of many bands would be dam- 

 aged more severely than those in which the banding is slight. 



All the sandstones which were tested are fine grained and rather 

 compact. All suffered some injury, though, in most cases, the 

 cracking was along the lamination planes. In some cubes, how- 

 ever, transverse cracks were also developed. 



The variety of samples was not great enough to warrant any 

 conclusive evidence toward a determination of the controlling fac- 

 tors. It would seem, however, that the more compact and hard the 

 stone is the better will it resist extreme heat. The following rela- 

 tion of the percentage of absorption to the effect of the heat is 

 interesting. In a general way the greater the absorption, the 

 greater the effect of the heat. A very porous sandstone will be 

 reduced to sand and a stone in which the cement is largely limonite 

 or clay will suffer more than one held together by silica or lime car- 

 bonate. 



PERCENTAGE OF 



ABSORPTION PLATE LOCALITY 



3.084 25 Warsaw 



2.310 22 Potsdam 



1.876 24 Medina 



1. 1 18 23 Oxford 



