. of Iodine and Codeine. 105 



r 7*906 grains of teriodide of codeine gave 

 II. \ 9-352 . . . carbonic acid ; and 

 [2-448 ... of water. 



8-621 grains of teriodide of codeine gave 

 111.^10-216 ... carbonic acid ; and 

 2-722 ... of water. 



r8-( 



. 9-i 

 I 2-^ 



8-042 grains of teriodide of codeine gave 

 IV. -^ 9-245 ... carbonic acid ; and 

 -413 ... water. 



The determination of the iodine was at first attempted in 

 the ordinary way by combustion with quicklime. In repeated 

 experiments, however, numbers were obtained varying from 

 49 up to 53, and on one occasion nearly to 54 percent. ; but, 

 notwithstanding every precaution in making the experiment, 

 it was found impossible, by this method, to obtain the full 

 amount of iodine. The substitution of carbonate of potash 

 for lime was attended with much more satisfactory results. 

 As carbonate of potash cannot easily be obtained absolutely 

 free from chlorine, I employed a specimen of that substance 

 containing a known amount of chlorine. A fixed weight of 

 this carbonate was mixed with the material for analysis, and 

 the combustion conducted in the usual way, a high tempera- 

 ture being sustained during the whole experiment. After 

 combustion, the contents of the tube were dissolved in water, 

 and dilute nitric acid slowly dropped in, until the solution 

 was ne.arly but not entirely saturated. Nitrate of silver was 

 then added, and after it a quantity of nitric acid sufficient 

 to render the fluid distinctly acid. The precipitate was 

 washed in the usual way and weighed, and then by subtract- 

 ing from it the amount of chloride of silver corresponding 

 to the chlorine which was present in the carbonate of pot- 

 ash employed, we have the weight of the iodide of silver. 

 The first trials made by this method gave a considerable 

 excess ; but this was traced to a quantity of carbonate of 

 silver having been carried down with the iodide, and pro- 

 tected by it from the action of the nitric acid with which the 

 fluid was afterwards super-saturated. All that is necessary 

 to avoid this source of error is to add a few drops of nitric 

 acid to the water with which the iodide is washed, and to 



