72 PROCEEDINGS OF THE AMERICAN ACADEMY 



inch above the neck. The upper end of the cork fits the neck of a 

 glass plate ground on the under side, made as is represented at e, Fig. 

 2. This plate is about three inches in diameter, and when resting 

 upon the plate d, as is represented in Fig. 1, covers the hole com- 

 pletely, and permits sufficient lateral motion to bring the stream of 

 water on different parts of the precipitate. 



" Numerous processes in which this apparatus may be applied will 

 suggest themselves to any chemist. I will only mention one in which 

 I have found it very useful. In the ordinary process of separating 

 alumina and sesquioxide of iron from the alkaline earths, the sesqui- 

 bases are precipitated by caustic ammonia, which does not, as is well 

 known, precipitate the alkaline earths when perfectly caustic. Since, 

 however, the best liquid ammonia contains carbonic acid, and as, dur- 

 ing the precipitation and subsequent filtration, carbonic acid is ab- 

 sorbed from the air, it invariably happens that small amounts of the 

 alkaline earths are precipitated as carbonates. It becomes, therefore, 

 necessary to redissolve the precipitate, and repeat the process in order 

 to obtain a complete separation. This repetition, I think, may be 

 avoided by using ammonia gas made caustic by lime, and conducting 

 the filtration in the apparatus just described. The substitution of am- 

 monia gas for liquid ammonia in this process has been made by many 



chemists. I use for precipitating with caustic 

 ammonia gas the little apparatus represented 

 in Fig. 3. Strong liquid ammonia is placed 

 in the flask, where it is gently heated, and the 

 resulting gas passed through a chloride of 

 calcium tube into the solution. To the end 

 of the bent tube I attach, by means of a 

 caoutchouc connector, a short tube which dips into the fluid. This, 

 when the precipitation is completed, can readily be disconnected, and 

 any adhering precipitate easily removed. The fluid with the precipi- 

 tate I now throw upon the filter arranged in the apparatus above de- 

 scribed. As the process of filtering and the subsequent washing is 

 conducted in a very confined space, which can easily be entirely de- 

 prived of carbonic acid, no precipitation of the alkaline earths is pos- 

 sible, so that they are completely separated from alumina and ses- 

 quioxide of iron at the first precipitation. 



" I use two sizes of the filtering apparatus, which differ from each 

 other, however, only in the size of the bottle. In both, the diameter 



