195 



110 oxygen, and 1 eq. of ammonia more than the red salt. My 

 results lead to the same conclusion, so that its empirical formula 

 appears to be Co., Clj Ng Iljg, or Co^ Clg + 6 NHg. It also forms 

 salts with oxygen acids, and from these an oxidised base may be se- 

 parated, but has not been fully studied. 



All the three authors who have preceded me describe the crystals of 

 the red salt as regular octohedrons, and they must be very nearly so ; but 

 Sir D. Brewster informs me that they do act to a small extent on po- 

 larized light, in which case they cannot belong to the regular system. 



Fremy, differing from Genth, also describes the yellow salt as 

 forming regular octohedrons. But this is, I think, a mistake ; for, 

 as far as I have examined them, they appear to be prismatic. Genth 

 describes the crystals as rhombic or klino-rhombic. 



2. On the Acid formed when Potash acts on Oil of Bitter Almonds. 



When commercial oil of bitter almonds is mixed with an excess of 

 an alcoholic solution of potash, there is formed, instead of beuzoine, 

 a salt, crystallising in scales, which are very soluble in alcohol. This 

 salt is said in books to be benzoate of potash. And when decomposed 

 by acids, it yields an acid which, to all appearance, is benzoic acid. 

 But it is worthy of notice, that if we form benzoate of potash with 

 common benzoic acid, the salt is hardly at all soluble in hot alcohol, 

 and does not crystallise in the same way as the salt above mentioned ; 

 indeed can hardly be got to crystallise at all. I have made many 

 experiments to ascertain the cause of this strange difference, but I 

 have as yet been unable to detect it. The salt I exhibit has been 

 three times recrystallised from alcohol, and is as soluble as ever; while 

 yet the acid extracted from it appears identical with benzoic acid. 

 Its analysis, indeed, does not perfectly agree with that of benzoic acid, 

 but the diflerence is so slight, as not to affect the formula. 



Is it possible thai the presence of some foreicrn matter communi- 

 cates to the potash salt the property of solubility in alcohol, and that 

 of crystallising readily ? But if so, the more it is purified, the less 

 soluble it should become. This I have not found to be the case. I 

 rather suspect that the acid is not truly benzoic acid, and that a more 

 minute investigation will detect its true nature. Its resemblance to 

 benzoic acid is certainly very striking, but we know that homoloorous 

 compounds, although different in composition, often resemble each 

 other in as great a degree. 



VOL. Ill, Q 



