124 Prof. A. E. Nordenshiold — Geology of Spitzbergen. 



especially peculiar to basalt, and which by some geologists has been 

 regarded as a result of concretionary structure, or even compared to 

 a kind of crystallization of the whole mass of the rock. The origin 

 of the basaltic structure, however, appears to lie near at hand, and 

 the phenomenon, though operating on a smaller scale, is a very 

 common one. If, on a warm summer day, we walk across a bed of 

 clay, the surface of which has become dry, we shall often see the 

 crust of the clay broken into regular and mostly hexagonal figures. 

 Fissures of the same form may be seen on the marshy plains which 

 are met with at iho foot of the mountains in all parts of Spitzbergen. 

 In the middle of summer, when the snow has just melted, they are 

 so very marshy, that in going over them one sinks to the knees in 

 the moist, angular gravel, but towards autumn they generally 

 become quite dry, and the surface splits into regular figures. The 

 gradual contraction of a solid mass evidently constitutes the common 

 cause of the hexagonal structure of these so widely different substances. 



" When the clay or the gravel beds dry, or the plutonic rock cools, 

 a contraction takes place, beginning at the surface, or at the walls of 

 the vein, and gradually perpetuating itself to the interior, the result 

 of which evidently is a bursting of the mass. Then the two follow- 

 ing conditions must be fulfilled : 



''1. The bursting must take place in such a manner that the re- 

 sistance it meets with proves a minimum. 2. The particles must 

 not be dislocated to so high a degree as to loosen the upper contracted 

 bed from the lower on« that is yet uncontracted. This latter con- 

 dition is a necessary consequence of the contracting process beginning 

 at a certain plane and perpetuating itself gradually to the interior of 

 the mass. 



" If we solve this minimum problem mathematically on the bases of 

 the two conditions above stated, we find that the clefts which arise 

 in a solid mass when it contracts must necessarily be composed of 

 plane surfaces crossing each other so as to form regular hexagonal 

 columns, and be perpendicular to the surface where the temperature 

 is constant, or parallel with that direction in which the contraction 

 is progressing. The columnar structure pecidiar to basalt is thus a 

 simple consequence of the fact that the bursting, caused hy the contrac- 

 tion of the solid mass in cooling, tahes place along a system of surfaces 

 ivhere the resistance is at minimum; and the basaltic structure has 

 nothing in common with the concretionary structure, or loith crystalliza- 

 tion : although the regular form of the basaltic columns gives them a cer- 

 tain resemblance to crystalline prisms." 



On the west coast of Spitzbergen, where the strata are almost 

 vertical, the diabase only sparingly occurs. It is likewise nearly 

 absent on Liefde Bay and the north coast of North-East Land. On 

 the contrary, it occurs well developed on the inner part of Ice Sound, 

 Hinloopen Strait, and Stor Fiord. 



I have had several opportunities of observing the contact between the 

 diabase stratum and the underlying strata. The latter, if they were 



1 The transverse clefts which often divide the hasaltic columns seem to arise from 

 an interruption in the successive contraction, caused, for instance, by infiltration of 

 water in the solid but yet warm rock. 



