of Chelidonin and Jervin. 33 



centigrade; if heated then till it melts, which it does at 130 

 cent., it loses no more weight. It is soluble in alcohol 

 and aether, but not in water ; a good deal of ammonia is deve- 

 loped when it is melted with caustic potash. A solution of 

 chelidonin in diluted muriatic acid, or the watery solution of 

 the neutral muriate itself, gives with chloride of platinum a 

 beautifully yellow precipitate, at first flaky, but afterwards 

 granular, which may be washed out without decomposition. 

 This precipitate, which is a double salt of muriate of cheli- 

 donin with chloride of platinum, was used for determining 

 tlie atomic weight of chelidonin. This double salt was not 

 decomposed by boiling with nitric acid. 



First. O'MS of chelidonin dried in the air lost at the tem- 

 perature of 100* centigrade, 0*023 of water =5*13 per cent. 



Second. At 130°, when the chelidonin was melted, 0"172 

 lost 0*008 of water = 4'6"5 per cent. ; the mean of both ex- 

 periments is ^'SQ per cent. 



1. Chelidonin weighing 0*425 dried at 100°, gave 1*062 of 

 carbonic acid, and 0*215 of water. 



2. 0*2585 gave 0*64'3 of carbonic acid, and 0*1315 of water. 



3. 0*321 gave 0*793 of carbonic acid, and 0*162 of 

 water. 



In a direct nitrogen determination 0*2995 yielded 30 cubic 

 centimeters of nitrogen gas at J 1° cent., and height of the 

 barometer 27" 10'", which after reduction to standard tempe- 

 rature and pressure gave 2*88 cub. cent. = 12'19 per cent. 

 Hence the following per cent, composition is calculated : 



1. 2. 



Carbon 69*07 68*76 68*30 



Hydrogen ... 5*62 5'ii5 5*60 



Nitrogen 12*19 



Oxygen 13*12 



100*00 



1. 0*307 of chelidonin -chloride of platinum dried at 100° 

 left when burnt 0*0535 of metallic platinum = 1 7*'12 per cent. 



2. 0*3125 gave 0*055 of metallic platinum = 17*60 per 

 cent. 



From these two determinations the numbers 7076*8 and 

 7008*5, come out as the atomic weight of the double salt, 

 and therefore for the chelidonin 4'502*89 and 44'3't*9, which 

 determination, j)arlicularly the last, agrees as well with the 

 following formula as can be expected from such weighing, 

 where a dilfereuce of half a millegramme increases or di- 

 minishes the atomic weight 70 or 80 parts. 



Phil. Mag. S. 3. Vol. 18. No. 114. Jan. 1841. D 



