HEAT. 721 



On some of the Thimble Islands, off the shores of Stoney Creek, Conn., the walls 

 of granitoid gneiss facing the water, are peeling off in laminae, a third to a half inch 

 thick, without any apparent decomposition, or even a dimming of the lustre of the 

 feldspar or mica; and it may be owing to the heat of the day's sun, and the chilling by 

 the waters when the tide is in. Over the rocky surface of countries within the glacial 

 latitudes of the Glacial period, the scratches left by the glacier are generally, when first 

 uncovered, as fresh as when they were made. But, if the surface be open to the sun's 

 heat and light, and the rains and frosts, for a score of years, far the larger part of the 

 markings disappear; and alternate heating and cooling is an important means of this 

 obliteration. 



The sun's heat also produces cracks by drying, as in the case of 

 mud-cracks (p. 84). Such cracks in rocks are recognized by their 

 being very shallow. 



3. The Heat from a Subterranean Source. — From Totten's ex- 

 periments as data, Lyell has calculated that a mass of sandstone, a 

 mile thick, raised in temperature 200° F., would have its upper sur- 

 face elevated 10 feet ; and that a portion of the earth's crust, fifty 

 miles thick, raised 600° F. to 800° F., might produce an elevation of 

 1,000 to 1,500 feet. Cooling again, would reverse the result. 



Contraction from cooling in case of fusion, generally produces frac- 

 tures at right angles to the cooling surfaces ; and in this way, " basal- 

 tic" columns have been produced (pp. 87, 112). The columns are 

 sometimes curved, when the cooling surfaces are not parallel. In 

 contact with the adjoining rock, the cooled mass is often much frac- 

 tured in an irregular way, and finer in grain than elsewhere, in con- 

 sequence of rapid cooling. Basic rocks are oftener columnar than 

 acidic ; injected into hot rocks, neither may be so. Sandstones and 

 shales also become columnar, when subjected to a heating and drying. 



In a cooling layer of fused rock, the smallest number of fractures that can be opened 

 about a point on its surface by equable contraction is three, and hence this number is 

 the easiest to make; and since three such lines symmetrically placed make angles with 

 one another of 120° (Fig. 1109), the hexagonal prism, more or less regu- -pis. 1109. 

 lar, is the most common form of the " basaltic " column. In the case of 

 large dikes between walls of rock, that set of divisional planes which is 

 nearly or quite vertical is generally more strongly developed than the 

 others, and this occasions a laminated structure in that direction looking 

 like upturned bedding. Contraction also occasions fractures across the 

 columns. After the defining of the columns, the divisional planes between the columns 

 may be to a slight degree cooling surfaces, and the transverse fracture planes are thus 

 sometimes made concave, as explained by Mallet, giving the columns a concave top, as 

 at the Giant's Causeway Dikes sometimes have, from the same cause, fissures of great 

 length and depth parallel to the walls, which are finally filled and become mineral veins. 



4. Expansion and Contraction attending Solidification and Fusion. 

 — Experiments on the contraction attending solidification of rock ma- 

 terial have been made by Bischof, St. Claire Deville, Delesse, and 

 Mallet ; and the results of the three investigators last mentioned 

 nearly agree. In solidification, the glass state is a consequence of 



46 



