298 



On the Cri/st(flli'/ir Stnicltire of (lie 



on tlie Neva at St. Potersbiiroh, at the l)e<,nnniiig of a lliaw, wlion two 

 feet in thii'kiiess,is traversed by rows of minute air bubbles extending iit 

 straight lines, sometimes a liitle inlieeted, froni the upper surface of the 

 ice, towards (he lower: at from two to tive inches of which they terminate." 

 Other bloi ks j)resen(cd these bubbles united, softs to form cylindrical 

 canals a little thicker than a horse hair. — Observin«r slill further, he 

 says, "1 found blocks, in which the process was more advanced, and two, 

 three or more clefts struck off, in dillereut directions, from the verti- 

 cal veins, so that a section perpendicular to the veins would represent 

 in miniature (he star-formed cracks of timber. Finally, in some ])ieces 

 these cracks united from top to bottom of the veins, sei)arating (he whole 

 HKiss into vertical prisms, having a greater or less number of sides. In 

 this state a slight shock was sullicient to detach them : and the block, 

 with its scattered fragmetits, was in all respects, the exact miniature 

 resemblance in crystal, of a Giant's Causeway. The surface was like a 

 tesselated pavement, and the columns rose close, adhering, and ])arallel, 

 from the compact mas?^', of a few inches, at the lower surface — more or 

 less time is required for tlie process, which I have seen in all its different 

 stages." This, as well as the preceding class confirm a remark previously 

 made, namely that the phenomena in question depended essentially on 

 variation of temperature, and it matters not whether this variation is 

 caused, by the bodies losing, or gaining heat, as may readily be observed 

 by comparing the circumstance of (he different classes with each other. 

 In the last case, the phenomena are first observed when a thaw com- 

 mences, and their development keeps pace with the progress of this, so 

 that their connection with the variation of temperature, during the pas- 

 sage of (he solid ice into water, is thus evidently seen. In the former 

 classes the variation was consequent on the passage of fluids into solids, 

 and was therefore dependant on the obstruction of heat — in this it is 

 exactly vice versa. 



Having, in the preceding seven classes, included all (he phenomena 

 analogous to those of the trap dykes of Amboor, with which 1 am ac- 

 quainted, it now becomes necessary to consider with more care, and in 

 greater detail, the mutual connection that subsists between them ail- 

 so that, if it be possible, we may by this means be led to some more 

 definite ideas of the cause to which they are due, than we can now 

 have. 



Varied as the circumstances of the different cases have been, attention 

 has occasionally been directed to that one point, in which all, without 

 «xception, have been found to agree. Now, it is a rule in enquiries of a 



