" MOLECULAR COALESCENCE," ETC. 127 



grain, and at the level of this a line corresponding to the 

 plane which would join the obtuse angles divided the crystal 

 into halves. Internally, the structure was apparently fibrous; 

 it was, in fact, laminar, the laminae being curiously bent and 

 twisted in a spiral way, the little central cavity marking the 

 axis of torsion (PL XV, fig. 4). Above the middle of the plug 

 these bodies, having attained their full development, were 

 suddenly lost. 



Tablets, in some respects resembling the first, next appeared 

 (PL XV, fig. 5) . They were, however, much longer, and much 

 thicker in proportion to their breadth. These, as they grew, 

 were also modified. They became longer and expanded at 

 their extremities, the middle transverse line being unaffected. 

 At first each half was split, as it were, into two crystals, 

 firmly bound together at the middle, slightly divergent at the 

 end; then, by the continuation of the same process, each half 

 became broken up into a number of flat rods bound together 

 at the middle, so that the whole mass had a great resem- 

 blance to a wheatsheaf. Further subdivision, with curving 

 and flattening of the rods, turned the 'wheatsheaf into a 

 crystalline dumb-bell ; still further, as the oxalic surface was 

 approached, the forms rapidly fell in size, acquired a sharper 

 outline, thicker waist, and, to all appearance, a central 

 cavity. In some cases, by extreme arching over of the 

 points, the two halves of the dumb-bell met round the middle, 

 and so a sort of sphere was formed. Here also the octohedra 

 which had been present throughout, and had followed the 

 variations in the size of the associated bodies, underwent a 

 change of form. They were flattened in the direction of one 

 diagonal, and expanded in other directions. They next ex- 

 hibited indentations midway between the angles, with corre- 

 sponding outgrowth of the angles until they became distinctly 

 cruciform. The rays of the cross presently lost their sharp- 

 ness, and by transverse splitting (which never reached the 

 axis of any ray) became feathered. In the central mass of 

 the crystal the octohedral form could still be made out (PL 

 XV, fig. 6) . The principal forms altogether. observed were five, 

 viz.: — 1, octohedra; 2, tablets; 3, ovoid rhombohedra; 4, 

 calculous (coalescing) dumb-bells ; 5, wheatsheaves, and 

 crystalline dumb-bells. The coalescent forms were charac- 

 teristic of the calcic end, the crystalline of the oxalic end of 

 the plug. It was further evident that certain kinds of tablet 

 "Were in distinct constructive relation with the two kinds of 

 dumb-bell and the ovoid bodies respectively. It was evident 

 also that the object originally proposed had been attained; 

 in the presence of such a colloid as gelatin oxalate of calcium 



VOL. XII. — NEW SER. R 



