398 A?i easy mode of washing Precipitates, 



dates, so as afterwards scarcely to admit of thorough washing. If 

 the precipitate be covered with liquid, and the neck of the funnel 

 placed so deep in water that the bottom of the filtering paper may- 

 touch its surface, the passage of the fluid, and the consequent dry- 

 ing of the precipitate, will in certain circumstances be prevented, 

 but still there will be a partial consolidation, and a great delay in 

 the process. 



In like manner, when a large quantity of liquid is to be filtered, 

 in requires constant attention in filling up the funnel, so as to 

 cause the whole to pass through in the shortest time possible. 



AU the inconveniencies now adverted to, are prevented by the 

 following very simple arrangement : — 



AB (in the Fig.) is a common wide-mouthed 

 flask or receiver of any dimensions ; CD is a tube 

 of a quarter of an inch diameter at D, and drawn 

 out to a point of one-tenth of an inch at C; EF is 

 a stronger tube, having a diameter of three or four- 

 fifths of an inch at E, and of an inch or more at 

 F. Both are fitted into a tight cork as repre- 

 sented in the figure. CD terminates at the in- 

 terior surface of the cork D, and protrudes to- 

 wards C, half an inch beyond the other tube. 

 The termination, F, of the wider tube is about half an inch from 

 the bottom of the flask. 



When a vessel thus provided is nearly filled with water, and in- 

 verted, as in the figure, the air ascends by the wide tube EF, and 

 a stream of water rushes from C with a velocity proportioned to the 

 pressure. Let a 6 be the surface of the water in a funnel, if the 

 flask be lowered upon it, the liquid will continue to flow till the ter- 

 mination, E, of the M'ider tube reach the surface AB of the fluid in 

 the funnel. The air can then no longer gain admission by E, and 

 therefore the stream at C will stop. But AB remaining stationary, 

 the liquid in the funnel will subside and leave the point E, when 

 the air will again enter, and the stream from C will flow till the 

 fluid reach its former level a h, when it will again stop, and the 

 same series of alternations wiU go on till the flask AB is empty. 



By means of this simple arrangement, in a flask of suflicient di- 

 mensions, the washing of precipitates, or the filtering of a large 

 body of fluid may be carried on, with little or no trouble, for any 

 length of time. It may be got up also in five minutes, the tube 

 EF being simply the neck of a broken retort. When empty the 

 flask is easily filled again, without extracting the cork, by insert- 

 ing a small funnel into E. 



There is but one disadvantage attending this little apparatus: 

 When the surface of the fluid comes up to the point E, the liquid 

 rises in the tube EF, till it reach the level, c d, of the water in the 

 flask. In mere filtrations this is of no consequence ; but, in 

 washings, a portion of the precipitate ascends with the fluid, and 



