Carbonic Acid in Natural and Artificial Salts. 469 



Is set free simultaneously with the water, but upon this regula- 

 tion and the gradual heating depends the success of the opera- 

 tion. 



The water bath F, of which F' is a top view, is now put into 

 position, as shown, so that one limb of the U-tube fits into the 

 curved recess in the side of the bath. A small alcohol lamp 

 keeps the bath hot so that the water driven out of the salt may 

 not condense in the upper part of the limb. By the use of this 

 bath the time required for the operation is much shortened. 



As soon as that part of the tube, which contains the boat, 

 appears free from condensed water, it should be highly heated 

 until all the moisture has disappeared from the drawn out por- 

 tion. This will generally take from twenty minutes' to half an 

 hour : when completed, the water bath is removed and the 

 apparatus allowed to cool, the air current being still kept up. 

 As soon as the tube is cool, the U-tube is disconnected and the 

 boat with its contents removed and placed in a small well 

 stoppered glass tube which has been weighed. 



The increase of weight of the U-tube gives the amount of 

 water in the sample, while the weight of the small tube, con- 

 taining the boat and the calcined salt, if subtracted from the 

 sum of the weights of the tube, the boat and the salt taken, 

 shows a loss which is the water plus the acid C0 2 of the bicar- 

 bonate. The calcined salt which should be sintered together 

 but not fused can then be used either for a determination of 

 the residual C0 2 or, as is preferred, dissolved and the alkali 

 determined by standard sulphuric acid, methyl orange being 

 used as the indicator, the same portion being then used for the 

 CI determination if chlorides are present. 



Many attempts were made to collect and weigh the C0 2 

 driven off by heat, but it was found impracticable to do so 

 without, at the same time, sacrificing the water determination. 

 No matter how carefully the heat and air current are regulated, 

 the C0 2 comes off too rapidly for proper absorption in a potash 

 bulb and if the air current is too slow, water condenses back of 

 the plug and this determination is too low. 



"When no attempt is made to absorb the C0 2 , the results are 

 very satisfactory. A specimen of Urao (Na 2 C0 3 , NaHCO s + 

 2H 2 0) from Owens Lake, Cal., gave the following results : 



Salt taken. Loss of salt. H 2 0. Difference =C0 2 . 



1*0040 29-51 20-12 9'39 



1-0490 29-58 20-09 9-49 



1-1583 29-49 20*00 9'49 



Average 1-0638 29-53 20-07 9-46 



Three closely agreeing determinations of total C0 2 gave an 

 average of 38*13 per cent; two determinations of the alkali 



