350 Professors W. B. and R. E. Rogers 



have repeatedly witnessed, a sensible diminution of the weight 

 of the apparatus while resting on the balance. So considerable is 

 the amount of this action with some substances, where as in ca- 

 ses of nice research, much time is occupied in the counterpoising, 

 that we believe the results thus obtained can not fail of being se- 

 riously erroneous. 



II. Apparatus and processes for the more exact analysis of the 

 carbonates. 



Incited by the recent researches and impressive suggestions of 

 Dumas, in relation to the equivalents of oxygen, carbon and cal- 

 cium, we several months ago entered upon a series of investiga- 

 tions, to test for ourselves the accuracy of the received atomic 

 weights of lime, magnesia, baryta, strontia, soda and potassa, pro- 

 posing as the simplest means of effecting this object, to determine 

 the amount of carbonic acid, evolved from the pure carbonates, 

 by some process similar in principle to that above described. Re- 

 peated trials, however, convinced us that the imperfections already 

 mentioned as incident to Rose's process and our own, though of 

 but little moment in ordinary analysis, unfitted them for the 

 higher description of research, on which we were desirous of en- 

 tering. A more critical examination of the sources of irregularity 

 and error, in these and the other methods of analysis, at length 

 led us to the forms of apparatus, and modes of procedure, which 

 we have since employed with very satisfactory results. 



These instruments and manipulations, we will now proceed to 

 sketch, accompanying the description, with a reference to some 

 of the experiments used as tests of the accuracy of our process, or 

 as proofs of the errors not hitherto adverted to in the methods 

 commonly in use. Of the results of our enquiries thus far, as re- 

 gards chemical equivalents, some notice will be given under a 

 distinct head in a future number of this Journal. 



The main apparatus, that in which the decomposition is effect- 

 ed, and which is weighed at the beginning and close of the pro- 

 cess, is of two distinct forms, adapted to the different characters of 

 the carbonates under examination. Of these, one is seen forming 

 the middle portion of fig. 2, the other is delineated in fig. 3. 



The body of the instrument in both cases, consists of a light, 

 wide-mouthed bottle, having a capacity of about three cubic inch- 

 es, closed by a cork three fourths of an inch in thickness. In 

 the first form, the cork receives the tapering ends of two drying 



