DIALYTIC SEPAEATION OP GASES BY COLLOID SEPTA. 433 



0*05 cub. centim. Oxygen. 

 1'50 „ „ Carbonic acid. 

 1*85 „ „ Carbonic oxide, &c. 



2. Of the original comettes of fine gold, from gold assays conducted several months 

 before, 93*3 grms., having a volume of 4'83 cub. centims., were submitted without 

 any further treatment to aspiration at a red heat. The gold gave up in the first half 

 hour 9-45 cub. centims. of gas, and in the second half hour 0*8 cub. centim., making 

 together 10'25 cub. centims. Hence 1 volume of the gold comettes appears to hold 

 2*12 volumes of gas. This gas consisted of 



6-70 cub. centims. Carbonic oxide. 



10-25 



The comettes do not appear ever to assume again so much gas as they first acquired 

 in the assay muffle. It follows that the weight of a gold cornette is increased about 

 2 parts in 10,000 by the weight of occluded gas. As the gold also retains 7 or 8 parts 

 of silver in 10,000, it follows that the absolute quantity of gold in a cornette is less than 

 the weight of the cornette as indicated by the balance, by 1 part in 1000. This does 

 not disprove the accuracy of the usual gold assay, which is always made in comparison 

 with gold of known composition as a check, and is therefore relatively true. 



3. The same volume of gold comettes, amounting to 4*83 cub. centims., heated again 

 in carbonic oxide gas, gave up afterwards 1-6 cub. centim. of occluded gas, composed of 



1"4 cub. centim. Carbonic oxide. 

 0*2 „ „ Carbonic acid. 



re 



4. The same mass of gold comettes heated in hydrogen gas, gave up afterwards in 

 one hour 2*7 cub. centims. of gas, which appeared to consist of 



2"34 cub. centims. Hydrogen. 

 0-36 „ „ Nitrogen, &c. 



2^ 



The power of this metal to occlude hydrogen gas is very sensible. The metal here 

 appears to hold 0*48 volume of hydrogen gas. The same gold, when dissolved and pre- 

 cipitated, was also found capable of holding 0*44 volume of hydrogen. 



5. The same mass of comettes, heated in carbonic acid gas, gave up afterwards in 

 one hour 1"05 cub. centim. gas, in which baryta- water showed the presence of 



0"78 cub. centim. Carbonic acid. 

 3o2 



