FIRE TESTS OF NEW YORK BUILDING STONES 



17 



five minutes after which time the flame was again applied for 

 10 minutes and the cube was again allowed to cool. To deter- 

 mine the combined action of heat and water a second cube was 

 subjected, as before, to the flame for 10 minutes, then a strong 

 stream of water was turned on to the sample, along with the 

 flame, for five minutes. Then the water was turned off and the 

 flame continued for another five minutes, after which, for five 

 minutes more the flame and water together were allowed to act 

 on the sample. 



The results of these various tests are given in the sections of 

 the paper which follow and the tabulated effects are shown in 

 tables 5, 6, 7 and 8 with the separate sections. Reference to 

 the plates will show plainly the effect of these experiments on 

 the different kinds of stone. 



Thin sections of most of the rocks tested were examined under 

 the microscope with the hope that they might shed some light 

 on the cause of the variations in refractoriness of the different 

 stones. Unfortunately they did not and therefore the petro- 

 graphic descriptions are placed at the end of the paper. 



Fire tests on granites and gneisses 



The cubes, for the most part, in the 550 tests stood up very well. 

 All of the samples remained uninjured on slow cooling, with the 

 exception of the gneiss from Little Falls (n) 1 which developed a 

 few cracks. On sudden cooling but two samples seemed to have 

 been injured, and only slightly so. These are a coarse grained 

 granite from Pine Island (1) and a fine grained granite from Grind- 

 stone island (9). The gneiss from Little Falls (11) was measurably- 

 more affected on fast than on slow cooling. It will be noticed in 

 reference to the table that three of the samples, Pine Island (1), 

 Little Falls (11) and Northville (14), took on a brownish tinge. 

 This is probably due to a change in the condition of the iron present 

 from a ferrous to a ferric state. 



At the higher temperature (850 ) none of the samples remained 

 uninjured, though some suffered more than others. In all cases 

 the sudden cooling did more damage than the slow cooling. The 

 gneiss, Little Falls (11), acted very badly, especially on sudden 

 cooling, in which test it split parallel to the bands and had numer- 

 ous other cracks. The fine grained stones, Nyack (4) and Grind- 

 stone island (9), showed a tendency to spall off at the comers, 

 while all the other samples, which are coarse grained, cracked very 



'These numbers refer to samples as listed and described at the end of the paper. 



