J^lfcovery of Silex in the Epidermis of Vtgetalles. 57 



<he appearance of pulverized glafs. When the epidermis was removed, the canes no longer 

 polTcfled the property of giving out light on colli fion. 



D. 4. To afcertain, with certainty, the nature of the epidermis, I obtained from aScJ 

 grains of cane 22 grains of epidermis ; this was expofed, in a crucible, to the flrong heat of 

 an air-furnace for half an hour. It had loft three grains, was very white, infufible by the 

 heat of the blow-pipe, and infoluble in any of the mineral acids. Ten grains of it were kept 

 in fufion with cauftic potafli, in a filver crucible, for a quarter of an hour. The compound 

 was white and fcmi-peliucid. It was perfefily foluble in water, without communicating to it 

 any turbidity. Wlien muriatic acid was poured into the aqueous folution, a copious white 

 flocculent precipitate was produced. This precipitate ccJlefted, weighed about nine grains, 

 and had every property of filex. 



D. 5, To determine whether the wood and internal bark of the cane contained any filex, 

 I burnt 240 grains, carefully deprived of the epidermis, for an hour. The aflies were per- 

 fedlly white, and weighed about feven grains. When muriatic acid was poured upon them, a 

 portion was diflblved with efFervefcence. This portion was chiefly carbonate of potafli ; the 

 infoluble part, coUe6led, weighed about two grains, and was apparently filex. 



D. 6. Having afcertained, by thefe experiments, that the epidermis of the bonnet-cane 

 was chiefly compofed of flint, and that the luminous appearance above mentioned depended 

 on this compofition, I thought it probable that the other canes, particularly the fugar-cane, 

 arundo faccharifera, and the bamboo, or arundo indica *, were fimilar in their organization. 

 When two bamboos were ftruck together, I could perceive no luminous appearance. Four 

 ounces of this cane only afforded feven grains of true epidermis. This expofed to a ftrong 

 heat, left five grains of white matter, which had all the properties of flint. The reafon why 

 thefe canes produce no light, on collifion, is, that the flint of the epidermis is too fmall in 

 quantity, and too thinly diflFufed. The epidermis of the fugar-cane contained a ftill fmaller 

 proportion of flint : 200 grains of this gave five grains of white aflies, of which only one 

 grain was infoluble in muriatic acid : the four grains of foluble matter appeared to be car- 

 bonate of lime. A large piece of bamboo (the weight of which I am ignorant of) deprived 

 of the epidermis, gave a confiderablc quantity of white aflies, of which about two-thirds 

 were foluble in the muriatic acid : the infoluble part was filex. The aflies of the fugar-cane, 

 deprived of the epidermis, appeared to be chiefly compofed of carbonate of lime, and car- 

 bonate of potafli. 



D. 7. The analogy between the Englifli reeds and grafles, and the canes, and particularly 

 the fimilarity of the appearance of the epidermis, induced me to fuppofe that they might like- 

 wife contain tiiex. On this fuppofition I firfl; examined the arundo phragmites, or common 

 reed. It produced no luminous appearance on collifion with flint. Twenty-feven grains pf 

 the epidermis, expofed to a flrong heat, gave 13 grains of white earthy matter, infoluble in 



* I am aware that the ingenious Mr. Made difcovercd the tabaflicer, found in the bamiboo, to be chiefly 

 compofed of fiint. He, however, did not abfolutcly difcover it in the wood of that cane ; and no ore, to my 

 knowledge, has ever fufpefted its exiftence in the epidermis of any vegetable, — X>-, 



Vot. IIL— May 1799, I *fee 



