exhibited by Hoar Frost. 43 



their bases, not to touch at their summits. Thus they all con- 

 verged for a short space round the interior diagonal crystal, as, on 

 the exterior, they diverged from it, in the manner represented in 

 the sketch. 



Where two planes met at a right angle, similar arrangements 

 took place ; the crystals, whether on the external, or within the 

 internal angle, occupying a direction equidistant from the planes 

 on each side. 



This arrangement equally occurred when the angles of meeting 

 of approximate edges or planes were greater Or less than right 

 angles, the equidistant position being regularly preserved ; while, 

 in the case of the interior angle formed by planes, any contact of 

 the summits of the crystals was invariably avoided. Hence in 

 those cases where very acute interior angles happened to exist, 

 the crystals became so shortened, for the purpose of avoiding a 

 contact between those on the neighbouring edges, that, near the 

 extreme point, they at length vanished. 



In all other more complicated cases of the meeting of the parts 

 of the iron-work, the same general rules were found to prevail. 

 In every part, in short, however intricate, where the crystals were 

 formed, they seemed endued by a repulsive power, in consequence 

 of which they tended as far as possible to recede equally from all 

 the plane surfaces and edges in their vicinity, and, at the same 

 time, to avoid any contact with each other. 



It will be suggested that this repulsive property depended on 

 some electrical condition, as we are not acquainted with any other 

 power by which it can be explained ; but it is not easy to assign 

 that modification, or mode of action of this mysterious power, by 

 which the effect could have been produced. I cannot pretend to 

 suggest any solution of this appearance, and am better pleased to 

 leave it thus recorded among the numerous insulated and inexpli- 

 cable facts in science, of which the explanation will at some future 

 time appear as simple as it now seems difficult. 



I need only add that, on the same forenoon, all the iron rail- 

 work, whicli I examined for this purpose, between Portland- 

 Place and Great George-Street in Westminster, exhibited the 



