226 Dll. ORD. 



calcium salt, and that a remarkable series of gradations led 

 easily from the one set of forms to the other. At the calcic 

 end two forms were most abundant — perfect octohedra, and 

 large tabular crystals of oblong outline with rounded angles. 

 The tablets, when seen lying flat, were colourless, marked 

 with diagonal lines and faint concentric shadings ; when tilted 

 up on one side, were yellowish in colour, highly refractile, 

 longitudinally laminated, and again marked with diagonal 

 lines ; their thickness was about one third of their breadth. 

 They might be described as consisting of a number of fine 

 oblong plates bound together face to face, the outermost larger 

 in all dimensions, the innermost shorter by one twelfth and 

 narrower in proportion (PI. XV, fig. l,a, b, c). At this point 

 also were found in smaller numbers forms transitional between 

 crystals and coalescence bodies (non-crystalline, rounded, 

 calculous) ; two paths being taken. By one the dumb-bell 

 form was reached without loss of outline of the crystals, the 

 molecules undergoing rearrangement within, so that a large 

 flattened dumb-bell, with sharply cut edges, was found en- 

 closed in the outline of a tablet (PI. XV, fig. 2) ; by another 

 much more frequented path the dumb-bell was reached 

 through entire disintegration. In this second case smaller 

 tablets were found to lose their sharpness of outline and to 

 become granular, at the same time that they became marked 

 by a line of slight continuous constriction round the middle 

 of their long axis, the result being the formation of a small, 

 not very perfect dumb-bell, and beyond this of complicated 

 masses built np by the coalescence of a number of the smaller 

 ones, and taking sometimes the form of double, sometimes of 

 single, simple, or tuberculated spheres — mulberry calculi on 

 a small scale (PI. XV, fig. 3). The large dumb-bells were, as 

 is already remarked, very perfect, their outline sharp and 

 running in a bold unbroken sweep ; their substance nearly 

 homogeneous — though there were often indications of ten- 

 dency to radiating fibrillation — and very refractile; they 

 usually presented in each half a large cavity joined to its 

 fellow by a canal running through the isthmus. As succes- 

 sive sections were examined, the tablets and the octohedra 

 steadily diminished in size to the junction of the lower and 

 middle third of the plug, and mixed with them were found 

 small spherical molecules increasing in size as the others 

 grew less, until they were moulded into large bodies in which 

 the rhomboidal and spherical forms were engaged in a well- 

 balanced struggle for superiority. They were at first sight 

 oval, but were, in fact, rhomboidal with much rounded 

 angles. Each had a tiny central cavity like that of a starch- 



