128 On the colouring Constituent of Roses. [August> 



oxydation being aided by the addition of a little wax, the whole 

 gradually ran together into a single globule of more considerable 

 size ; which remained upon the charcoal exposed to the most 

 intense heat that could be communicated for several minutes. 

 Being then suffered to cool, it was so powerfully attracted by 

 the magnet, that from the preceding observations I had 

 little doubt of its containing iron in the metallic state ; and 

 upon submitting it to the proper tests, this fact was decided. It 

 exhibited a high degree of metallic lustre to the action of 

 the file. As I have since l'epeated the experiment, and always 

 with a similar result, I thought it might gratify your curiosity to 

 see the iron thus obtained. I have, therefore, sent you one of 

 these globules fixed into a deal splinter and filed ; exhibiting an 

 appearance like a small iron nail driven into the wood. But as a 

 philosophical query suggested by the whole of the preceding 

 inquiry, I wish to propose this question : whether the rose tree, 

 among other instances of the same nature, be not in itself a 

 living test of the presence of an acid and an alkali in the 

 plant ; the acid being manifested by the red colour of the parts 

 of fructification, and the alkali by the green colour of the leaves ? 

 The mere circumstance of oxygen existing in a greater proportion 

 to hydrogen, than is necessary to form water, in a vegetable 

 substance, is sufficient to explain the presence of an acid ; and 

 alkaline bodies, being substances negatively electrified, an excess 

 on the side of the hydrogen, explaining the alkaline character of 

 the leaves, may thence be inferred. That the iron may, perhaps, 

 exist as a phosphate in the flower seems to be warranted in this; 

 that the precipitate thrown down by ammonia from the infusion 

 of roses, when acted upon by a powerful heat, phosphoresces ; 

 and this compound is still more likely to be detected in those 

 flowers exhibiting white and blue colours, because these are the 

 hues of the native phosphates of iron in the mineral state. How- 

 ever, as the carbonated alkalies dissolve the oxide of iron with 

 ease, and this solution, which was first noticed by Stahl, is decom- 

 posed by the alkalies in a caustic state, it may be the same which 

 acts as a colouring principle in roses. That the base of it is iron 

 has been here proved. If I have leisure, I will carry this investi- 

 gation somewhat further ; and by extending it to the colouring 

 principle in other vegetable bodies, endeavour to ascertain 

 whether iron, so universally prevalent, be the only metal to 

 which the varied hues of flowers ought to be ascribed. Hitherto, 

 the nature of their colouring principle seems to have been un- 

 known. Thenard says of it, that it has not yet been obtained in a 

 separate form, although he admits that it is almost always soluble 

 in water. His words are,* " Lew principe colorant na point 

 encore etc isole ; il est presque toujours soluble dans Veay." 

 I have the honour to be, Gentlemen, yours faithfully, 



Cambridge, July 14, 1818. EDWAKD DANIEL ClARKE. 



* Traite de Chimie, tome troisieme (1716), p. 376. Paris, 1S15. 



