130 TRANSACTIONS OF THE [MAY 23, 



Moisture at 110° C 0.08 



Loss on ignition ...... 0.74 



SiO, 69.465 



Fe,03 . . 2.30 



AlA 17.50 



MnO trace 



CaO 2..'J7 



MgO 0.305 



K„0 4.07 



S 0.04 



97.07 

 NajO by difference 2.93 



Crushing-tests were made on five sample cubes with the great 

 Emory testing-machine in the School of Mines, and as preparatory 

 to this the specific gravity w-as found on four cubes at 2.688, 2.687, 

 2.684, and 2.688. After three weeks' soaking these cubes absorbed 

 water, respectively, 0.0021, 0.0021, 0.00224, and 0.0026 per cents. 

 The crushing-tests gave the following results. The cubes were first 

 ground and polished so that the faces next the jaws of the crusher 

 were parallel within a limit of error of 0.005 inches. The cushion 

 employed between the cubes and the jaws was blotting-paper. 



I. II. III. IV. V. 



Height in inches . 2.033 1.983 2.059 2.011 2.009 

 Breadth " . 2.0 2.13 2.02 1.97 2.03 

 Thickness " . 2.1 1.99 2.03 2.03 2.03 

 Area in sq. inches . 4.2 4.23 4.1 4.0 4.12 

 Maximum compres- 

 sion in pounds . 113,200 122,000 106,000 101,400 108,700 

 Crushing - strength 



per sq. inch . 26,952 28,841 25,853 25,350 26,383 



The cubes exploded without previous cracking. This strength 

 is exceptionally high, as the general run of granite is far less. It 

 does not, however, equal the elaeolite-syenite of Little Rock, Ark., 

 which was tested by the late J. Francis Williams (see Annual Re- 

 port of Arkansas State Geologist, 1890, vol. ii), where results of 

 over 30,000 pounds were obtained; but it is far beyond the re- 

 quirements of building. 



Cubes of the rock were also boiled in acid. One which was thus 

 treated for five hours in boiling dilute HCl (1 part HCl of specific 

 gravity 1.20 to 20 parts H^O) lost 0.59 per cent, in weight. A 

 second cube treated in the same way in boiling dilute H^SO^ (1 

 part H,,SO^ of specific gravity 1.84 to 20 parts H^O) gave a loss of 

 0.48 per cent. Both these results indicate a great resistance to 

 natural solvents. Two large cubes were placed in a muffle and 

 maintained at a bright red for half an hour. One was allowed to 

 cool just below redness and then plunged in cold water. It caused 

 one crack that extended half through. The other cube w'as allowed 

 to cool slowly in the air, and showed a thin, external, crumbling 

 layer. When these results are compared with somewhat similar 



