57 



ducts of decomposition, to indicate that it was variable in composition. 

 Further data are, however, needed for the final decision of this 

 question, and we shall accordingly feel very much indebted to any 

 chemist who has a specimen of narcotine of well-ascertained origin, 

 or which he believes to have a different composition from that given 

 above, if he will kindly spare us a sufficient quantity for analysis. 



II. Composition of Cot arnine. 



The combustion of cotarnine with oxide of copper and oxygen, as 

 well as the determination of the proportion of platinum in its chloro- 

 platinate, leads us to adopt the formula C 12 H 13 NO 3 for this base. 

 The formula usually adopted contains one more atom of carbon ; but, 

 independently of our analytical results, the supposition that cotarnine 

 contains only twelve atoms of carbon is supported by the simple 

 manner in which the action of oxidizing substances on narcotine. can 

 then be expressed, namely, by the equation 



C 22 H 23 NO 7 + O = C 10 H 10 O + C 12 H 13 NO 3 , 

 Narcotine. Opianic acid. Cotarnine. 



and, as will be shown hereafter, by the manner in which cotarnine is 

 decomposed by dilute nitric acid. 



III. Decompositions of Opianic Acid. 



Opianic acid is readily decomposed when heated with strong 

 hydriodic acid ; no iodine is set free, but iodide of methyl is formed 

 in considerable quantity at the same time as a non- volatile substance, 

 very easily altered by heat and exposure to air, especially if in contact 

 with alkali, the precise nature of which we have not yet been able to 

 ascertain. 



"When opianic acid is heated with an excess of a very strong solu- 

 tion of potash, it splits up into meconin and hemipinic acid. These 

 substances were found by experiment to be formed in proportions 

 corresponding to the equation 



2C io H io O 5 _ c 10 H 10 O 4 + C 10 H l O 6 

 Opianic acid. Meconin. Hemipinic acid. 



The meconin thus produced has all the characters which have been 

 ascribed by previous observers to meconin obtained by other processes ; 

 its identity was further established by analysis, and by the preparation 

 of chloro- and nitro -meconin, the former of which was analysed. 



