42 Dr. Mac Cullouh on a peculiar Appearance 



each other by a space equal to the breadth of their bases ; con- 

 jecturally, about the 30th of an inch. Where they were attached 

 to the salient angle, or edge of a bar, they were at right angles to 

 its line, and at the same time equidistant at the summits from each 

 of the including planes, so as to form an angle of 45 degrees, 

 with their surfaces on each hand. This distribution was main- 

 tained in the same manner in those parts of the iron-work which 

 consisted of curved or circular parts ; so that each group, or pyra- 

 mid, was invariably placed at right angles to the tangent of the 

 curve at that part, or in the direction of the radius of curvature. 

 Hence it appeared, at first sight, that the effort of each pyramid 

 consisted in an attempt to recede as far as possible on each hand, 

 not only from the planes, but from the edge ; and thus to attach 

 itself at right angles to the Utter. The same effect also took 

 place in the interior as in the exterior of the curved parts ; and 

 thus the whole was ciliated like the leaves of some plants with a 

 regular and beautiful fringed work. 



The singularity of this appearance exciting my attention, I was 

 induced to examine it more narrowly, for the purpose of seeing how 

 the pyramids would dispose of themselves among the more intri- 

 cate parts of the iron-work. It was found, in consequence', that 

 where any two edges of a bar met at right angles, the crystals 

 formed at an apex occupied the direction of the diagonal of the 

 cube which was formed by the union of the bars, or maintained a 

 distance equally removed from the edge of the joint on the one 

 hand, and from the plane at right angles to it on the other. But 

 the crystals on the two meeting edges, where nearest to the apex, 

 did not immediately assume a rectangular position towards these 

 edges ; diverging gradually in succession from that on the angle 

 till they assumed the regular position which they held on the re- 

 mainder of the edge. 



In the re-entering, or internal angle of the same joints, the 

 crystal of the angle was also prolonged according to the diagonal 

 of the cube ; and here the crystals, intermediate between that and 

 those which stood at right angles to the internal edges at a short 

 distance, were so arranged as, in maintaining an equal distance at 



