256 HOLMES. 



estimation of the alkaloids, that the whole question is still in confusion 

 and must so remain until the nature of the precipitates and the mecha- 

 nism of the reaction under varied conditions and with different classes 

 of alkaloids have been more thoroughly studied. 



EXPERIMENTAL. 



Some years ago H. D. Gibbs, of the Bureau of Science, obtained pecu- 

 liar results while testing samples of prepared smoking opium seized under 

 Act ^^ Xo. 1761 of the Philippine Commission and submitted to this 

 laboratory for examination. 



The iodic acid test -" for morphine often gave a chloroform solution 

 which was brown in color and not at all characteristic of solutions of 

 iodine in this solvent. The failure of the test was traced to an incom- 

 plete separation of morphine from codeine. It was found that, M'hile 

 small amounts of codeine in the solution do not interfere, the test is 

 always valueless when the concentration of the morphine is not consider- 

 ably in excess of that of the codeine, or when the latter alkaloid is present 

 in large amounts, irrespective of the concentration of the morphine, or 

 when the amount of codeine present is greatly in excess of the amount 

 of iodine liberated by the morphine. It seems probable from these obser- 

 vations that the tendency which codeine has to combine with iodine is 

 so gTcat that the liberated halogen will not, as such, pass into the chloro- 

 form and impart the characteristic color to that solvent, and also that 

 codeine has a greater avidity for iodine than morphine. 



Additional evidence, which would appear to corroborate this conclu- 

 sion, will be presented later. 



An investigation of some of the relations of these alkaloids was there- 

 fore undertaken, and since heroin is so closely allied to them, that 

 alkaloid was also included in the study. 



Solutions of the sulphates of morphine, codeine, and heroin in distilled 



water were made to approximately 1 'per cent strength, and tk solutions 



of sodiimi thiosulphate and iodine in potassium iodide prepared. A 

 few preliminary precipitations gave results which, although apparently 

 anomalous, were still capable of a certain correlation and indicated that 

 the iodine content of the precipitate was in large measure dependent on 

 the concentration of iodine in the solution. 



A series of determinations was then undertaken in the following 

 manner : 



In each experiment, 5 cubic centimeters of the alkaloidal solution were 



" This act forbids the importation and sale, except for medicinal purposes, of 

 opium and cocaine and their derivatives. 



=°Blyth, Poisons: Their Ellects and Detection, London, 4 Ed. (1906), 300. 

 This t«st depends upon the fact that morphine will set iodine free from iodic 

 acid. Codeine and heroin do not do this. 



