Davy — On a New Chemical Test for CarhoUc Acid. 157 



tinned for some time, the purple colouration will be destroyed and 

 a blue produced ; moreover, where organic matters are present along 

 ■with the carbolic acid, many of them will likewise, when heated 

 with the molybdic reagent to that latter temperature, or even much 

 below it, develop a deep blue colour, which would mask more or less 

 completely the purple effect of the carbolic acid. Consequently it is 

 better in most cases to let the test act on the liquid at the ordinary 

 temperature, though the reaction may be somewhat slower in develop- 

 ing itseK. 



I have made a number of comparative experiments with this test, 

 and with the principal ones hitherto employed for the detection of 

 carbolic acid, and I find, in point of delicacy, it seems only to be sur- 

 passed by the bromine test of Dr. Landolt, which depends on the cir- 

 cumstance that when an aqueous solution of bromine is brought in 

 contact with carbolic acid, there is immediately formed the tribromo- 

 phenol (CeHsBrjO), a sparingly soluble white substance. But that 

 test could not be successfully employed, at least immediately, in many 

 cases, where the test just described might be still available, as for 

 example, in the case of different organic mixtures, where the presence 

 of the tribromophenol formed would be concealed. It possesses like- 

 wise the great advantage of being apparently not interfered with, to 

 any extent, by the presence of organic substances which mask or pre- 

 vent the reactions of many of the other tests. 



As to what is the exact nature or composition of the purple com- 

 pound which is formed in carrying out the test, I have not yet been 

 able to determine, owing to the difficulty of isolating it, or of obtaining 

 it in a condition suitable for analysis ; but I am inclined to think that 

 it is not so much an oxidation product of carbolic acid as a deoxidation 

 one of molybdic acid, and that it is a combination' of one of the oxides 

 of molybdenum, containing perhaps more oxygen than the blue com- 

 pound which is formed where the molybdic reagent acts on alcohol 

 and on some other substances ; and one circumstance amongst others 

 which seems to support this view is this, that I have failed to obtain 

 by the action of other oxidizing agents on carbolic acid a similar 

 purple reaction. 



Be this however as it may, I have satisfied myseK that the purple 

 compound formed in my test is a totally different substance from the 

 red or crimson dye termed coralline, which is obtained by the united 

 action of oxalic and sulphuric acids on carbolic ecid, and is now largely 

 used as a dyeing material ; for the red colour of the latter substance is 

 not affected by the caustic alkalies, and strong sulphuric acid changes 

 it to a reddish yellow ; whereas the purple developed in the new test 

 is changed to green by caustic alkalies, and the purple again restored 

 by strong sulphuric acid. I am, however, still engaged in this inquiry, 

 and hope to be able yet to determine the exact natiu-e of this purple 

 compound, and of the changes which occur in this new reaction. 



I shall now briefly notice some of the useful applications which 

 may be made of this test. It is well known that carbolic acid is a 



