1 6 NEW YORK STATE MUSEUM 



DESCRIPTION OF FIRE TESTS 



The samples from each locality were cut into three inch cubes. 

 Most investigators, who have studied the refractoriness of build- 

 ing stones, have selected one or two inch cubes; but these sizes 

 do not give as accurate results as the larger ones, for the reason 

 that a small piece becomes easily heated throughout the mass 

 and consequently upon neither heating nor cooling are differen- 

 tial stresses between the interior or exterior likely to be set up, 

 as would be the case if larger cubes are selected. In actual 

 fact in the burning of a building the stone does not become thor- 

 oughly heated; the heat penetrates probably but a slight dis- 

 tance into the mass, while the interior may remain comparatively 

 cold. The heating and cooling of this outer shell causes strains 

 which do not obtain in a stone which has been heated through- 

 out its entire body. One, two and three inch cubes of the same 

 kind of stone have been tested in the laboratory and while the 

 smaller cubes stood fire very well, the larger ones were more affected 

 and in some cases went to pieces. It was to avoid this error 

 and to approach more closely the existing conditions in a con- 

 flagration that the three inch samples have been employed in 

 the present series of tests. 



As far as the number of cubes would admit six tests were made 

 on the stone from each locality, four furnace and two flame tests, 

 For the first set of experiments a Seger gas furnace was used, 

 thus allowing the cube to be gradually and evenly heated. An 

 opening was cut in the cover of the furnace large enough to admit 

 the three inch cube of stone, to which a wire had been attached 

 to facilitate its handling. 



One sample was heated at a time. The heat was applied gradu- 

 ally for half an hour until a temperature of 55oC. was reached, 

 which was maintained for half an hour. The temperature was 

 measured with a thermo-electric pyrometer. The cube was then 

 takenK>ut and allowed to cool in the air. A second sample was 

 heated, as before, to 550, and this was suddenly cooled by a strong 

 stream of water. The third and fourth cubes were heated to 

 85oC. kept at that temperature for half an hour and cooled slowly 

 and suddenly as in the 550 tests. 



In order to approach more nearly the conflagration conditions 

 samples were subjected to two flame tests. In the first case the 

 cube was so placed as to be enveloped on three sides by a steady 

 but not strong gas blast. The flame was allowed to play on the 

 cube for 10 minutes, then the samples were allowed to cool for 



