250 Prof. G. Gulliver on Raphides and other Crystals. 



XXIX. — Observations on Raphides and other Crystals. 

 By George Gulliver, F.R.S. 



[Continued from p. 56.] 



Quillajea. — The crystals in the wood or bark of Quillaja sapo- 

 naria, which were long ago well described by Edwin Quekett, 

 are very characteristic of the genuine article. In a sample from 

 Messrs. Butler and M'Culloch, I find these crystals lying in 

 great numbers along the liber and mesophlceum. They are 

 commonly about xFvth of an inch long and -nnroth thick, four- 

 sided rectangular prisms, each of the faces equal, and the ends 

 tipped with short pyramids. But they vary in form. Their 

 ends may be like the edge of a chisel or wedge, and occasionally 

 as if the shaft of the crystal had been cut through obliquely 

 from one angle or face to the opposite one ; besides, the prisms 

 may be triangular. Though they are so very plentiful, they 

 occur for the most part singly, sometimes two or three partially 

 fused together, and never in bundles, in which characters they 

 further differ from true raphides, and closely resemble many of 

 the crystal prisms of Iridacese and some other Monocotyledones 

 ('Annals,' Sept. 1863 and April and May 1864). Quekett de- 

 scribes each separate prism of Quillaja as having a close invest- 

 ment or cell, but no loose one, of cellulose. 



MelastomacecE. — A species of Melastoma, at Redleaf, affords 

 an abundance of sphseraphides in the endophlocum and meso- 

 phlceum, but no raphides either in the bark or leaves. 



Crassulacea, Ficoidea, and Cactacea. — A complete examination 

 of these orders would be interesting and useful. Among the 

 few species formerly examined ('Annals,' May 1864) raphides 

 were always found abundantly in Mesembryanthemum, and never 

 at all in Crassulacese and Cactaceae, although sphseraphides and 

 short four-sided prisms were seen to abound in the last-named 

 order. These prisms sometimes appeared either abruptly trun- 

 cated, tipped with low pyramids, or with the ends as described 

 in Quillajese, &c., the tips commonly forming a part and pro- 

 jecting on the surface of the sphseraphides. Lately I have 

 again examined the plants already specified, and a few others, to 

 wit, Sedum speciosum, S. Fabaria, Epiphyllum Russellianum, Ce- 

 reiLS crenatus, and two species of Mesembryanthemum. The result 

 was still the same — a profusion of raphides in Mesembryea;, and 

 none in Crassulacese and Cactaceae. Raphides were seen abun- 

 dantly in the corolla, style, and ovary, but not in the stamens 

 and ovules, of Mesembryanthemum tricolor, and in the petals and 

 filaments of M. tortuosum. In these last two parts, and in the 

 ovary and pistil, the raphides were smaller and more fragile than 

 in the leaves and stem ; and, as I have described in other spcr 



