[ WHEELER] THERMAL EXPANSION OF ROCK 21 
A work which appeared soon after that of Reade was the “ Prelimi- 
nary Notes on the Coefficients of Thermal Expansion of Certain Rocks” 
by. Professor William Hallock,’ who treated the subject from a “scien- 
tifie or an economic point of view” for ordinary ranges ” of tem- 
perature. The first method described is that of finding the cubical 
expansion by the displacement or pycnometer method. An iron box 
was constructed with a cover which could be screwed on, the specimen 
being placed in it and the surrounding space being filled with mercury. 
The necessary calibrations were made. “Then knowing the coefficient 
of the mercury and iron, and the volume of the stone, iron and mercury, 
and the apparent coefficient of the stone and mercury in the iron, a 
simple calculation gives the coefficient of the stone alone. The linear 
coefficient is practically one-third the cubical.” It was found that the 
mercury penetrated porous stones. “There seemed also a source of 
error arising from the difference of ‘fit’ of the mercury to the stone at 
high and low temperatures,” 1.e., the mercury seemed to fit into the 
uneven cavities on the surface of the stone more perfectly when hot. 
These considerations led to the abandonment of this method, and 
the comparator method, involving the use of micrometer microscopes, 
was adopted. The specimens were heated in a water bath, the extremes 
of temperature being about 20° and 100° C. Hallock estimates that his 
results should be within about 0.1%. Chemical analyses of the speci- 
mens were also made. The results obtained were as follows: 
Standard Brass Bar CN RARE 0.00001866 
A Slate from near Hydeville, Vt. ... . .0.00000881 
Amarbletfrom Ritland V6”. CURE UX 0.00000659 
6e x (CVV UGlam ed NGS Nine, 0 .00000661 
cé cé eM iKnomville.Tenne gsc. < - 0 .00000495 
“ ce ce Ts 36 Ps SU ER 0.00000525 
cc ce “  Keowa He from Georgia 0 .00000348 
ce 7 7 ce 66 ne .0 .00000309 
cc ce ce hs Cresle, ” from Georgia. .0.0000110 
cc rT “ 4 Cherokee” “ ‘6 . .0.00000740 
ce ce ce ce Cie ..0.00000786 
«e 7 cc “ coi NC . .0.00000855 
The duplicate values are for different pieces of the same stone. 
Hallock notes that the marble samples upon which he was at work, 
“being heated for the first time to 100°C. and allowed to cool, did not 
contract to their original length, and the next two or three heatings 
resulted in continued but ever diminishing increments of length at 


1U. 8. Geol. Survey Bull. 78, pp. 109-118. 
