356 DR ANDERSON’S RESEARCHES ON SOME OF THE 
ordinary temperatures, 33°28 per cent. was taken up. These results are very un- 
intelligible, and certainly cannot be employed as the foundation of an atomic 
weight. It is, however, worthy of observation, that they are very nearly in the ratio 
of 1, 2, and 4, but this relation must be purely fortuitous, as I have found that 
thebaine is very easily decomposed by hydrochloric acid, and none of the results 
agree at all with the actual atomic weight, as deduced from the experiments which 
I am about to detail. 
VIL. Analysis of Thebaine. ' 
The thebaine employed for analysis was prepared by the process already de- . 
scribed, and purified by repeated crystallisation ; it was burned with oxide of cop- 
per, and is very easily combustible. 
14-675 ...  earbonic acid, and 
5-475 grains of thebaine, dried at 212°, gave 
ils 
3°500 sete Water. 
4-990 grains of thebaine gave 
Il. ¢ 13-383 ...  earbonie acid, and 
3°135 ee» water. 
5-089 grains of thebaine, of another preparation, gave 
IIT. ¢ 13-621 ... carbonic acid, and 
3°228 -». water. 
5:336 grains of thebaine gave 
3°735 ...  platinochloride of ammonium. 
6°332 grains of thebaine gave 
4515 ...  platinochloride of ammonium. 
I. II. III. 
Carbon, 3 73:10 73:14 7301 
Hydrogen, A 7:10 6:98 7:04 
Nitrogen, : 4:39 4:47 see 
Oxygen, : 15:41 15:41 


100-00 100-00 
These results correspond exactly with the formula, 
C5, Hy, NO, 

differing from that of codeine by two equivalents of carbon, as is seen from the 
following comparison of the experimental mean with the calculated result of that 


formula. 
Mean. Calculation. 
eve eS 
Carbon, : 73°08 73°31 (Cas 228 
Hydrogen, ; 7-04 6°75 a 21 
Nitrogen, - 4:43 4:50 N 14 
Oxygen, : 15:45 15:44 O, 48 
100-00 100-00 311 
