1822.] Philosophical Transactions for 1821, Part IT. 145 



dantly deposited in short acicular crystals aggregated in stelli- 

 form groups. These crystals pressed between folds of unsized 

 paper, in order to separate the adhering acid, and then exposed 

 to heat, are readily melted : in cooling, the melted mass shows 

 evident traces of acicular crystallization, and the crystals are of 

 a yellow colour. This yellow substance is readily inflamed, burns 

 with a bright flame, emits much smoke, and leaves a consider- 

 able residuum of carbon. 



'* Of all the characters of the white concrete substance? 

 described in this section, its ready disposition to crystallize is, 

 perhaps, the most remarkable. 



^' If thrown into a red-hot crucible, a dense white vapour 

 arises from it; which, being received into a bell glass placed 

 over the crucible, is condensed round the lower part of the glass- 

 in the form of a white powder ; but in the upper and cooler part 

 of the glass distinctly crystalline plates are formed, of a beautiful 

 silvery lustre. 



" A similar and equally beautiful crystalhzation may be 

 obtained by boihng this substance in water, in a glass matrass 

 having a long neck ; in the upper part of which crystals will be 

 formed and deposited during the boihng. 



'* If exposed to a degree of heat not more than sufficient to 

 melt it under a bell glass, the vapour that rises from it crystallizes 

 before it reaches the surface of the glass, and flies about the 

 interior with exactly the appearance of a shower of minute parti- 

 cles of snow. 



^' If a piece of cotton twine be coiled up like the wick of a 

 candle, and after having been dipped in this substance while 

 melted, be set on fire for a second or two, and then blown out, 

 the vapour will soon begin to crystallize round the wick in very 

 distinct thin transparent laminae. 



^* This experiment affords one mark of distinction betweea 

 this substance and benzoic acid, and also between it and cam- 

 phor ; for under other similar circumstances, benzoic acid crys- 

 tallizes in acicular crystals, which are often grouped in a 

 stelliform ; and camphor crystallizes, or is rather congealed, in 

 globular particles having a stalagmitic appearance. 



'' The most usual crystalline form of this substance is a 

 rhombic plate, of which the greater angle appears to be from 100^ 

 to 105° : crystals at least of that form I have repeatedly ob- 

 tained from its solutions in water, in alcohol, in acetic acid, in 

 the yellow oil described in the last section ; and lastly, by melt- 

 ing and very slowly coohng the substance itself. Sometimes 

 several of these plates are variously grouped together; some- 

 times a single plate intersects another plate at nearly right 

 angles, so that in some points of view the compound crystal 

 appears simply cruciform. The only distinct modifications 

 which I have observed of the common form are a rhomboidal 

 plate, which is very nearly rectangular ; and an hexagonal plate ; 



New Sef'ies, vol. iii, l 



