228 DIFFUSION AND OSMOSIS. 



then filled with any diffusible liquid, — for Instance, a dilute solu- 

 tion of sugar, — the manometer is attached, and the whole appa- 

 ratus is placed in pure water or an^- aqueous solution. 



and oven on all siJes, place them quickly together, lapping about 15 mm., 

 and continue the rotary motion until cool. Take a scalpel with a point bent 

 at right angles to the blade, heat it, and, inserting it in the glass tube, 

 cut away the wax at its inner end, thus exposing a shoulder of the thickness 

 of the glass. Roll out in the form of a pencil about 2 mm. in diameter a piece 

 of sealing-wax which has been made a little soft by the addition of a drop or 

 two of turpentine. A piece of this, equal in length to the inner circumference 

 of the glass tube, in a long coil, should be placed on the point of the scalpel, 

 carried in to the shoulder and pressed into it. A little heat very cautiously 

 applied fi'om without, with projier turning of the cell, will easily cause this 

 softer wax to flow and fill the shoulder with perfect smoothness. The use of 

 the softer sealing-wax makes a joint wliich will not crack under strong pres- 

 sure. Now cement the tube t very firmly into v, with the same precautions 

 as above. Unless a pressure of more than three atmospheres is desired, the 

 soft wax need not here be u.sed. 



The cell is now ready to be prepared for filling. In order to saturate the 

 porous porcelain with any given solution, the air must first he wholly removed. 

 Place the apparatus in a beaker of water which has been freed from air by boil- 

 ing, and set the whole under the bell-jar of an air-pump. Exhaust and admit 

 the air into the bell-jar repeatedly until bubbles can no longer be seen to rise 

 from the porcelain. Transfer the cell to a three per cent solution of copper- 

 suli>hate, and exhaust the air again. Four or five hours will be required for 

 this .solution to thoroughly penetrate the porous cell. At the end of this time 

 remove it from the copper-sulphate, empty it, and with some long twisted 

 strips of bibulous paper quickly diy up all moisture from its inner surface. If 

 at any time the exterior surface of the cell begins to appear dry before the 

 moisture from within has been wholly removed, dip it at once in the solution 

 from whence it came. At the moment when the moisture is properly removed, 

 fill the cell to the second joint with a three per cent solution of potassic ferro- 

 cyanide and replace it in the copper-sulphate, taking care that the surfaces of 

 the two fluids are in the same plane. An interim of at least twelve hours must 

 now elapse in order that the membrane may be properly formed. At the end of 

 this time the cell is ready to be used, either with the solution which it already 

 contains or with some other. If some other solution is to be employed, then 

 carefully empty out the potassic ferrocyanide, and after washing the cell with 

 a little distilled water, fill it with the fluid to be used. 



The cell must be so filled and sealed as to leave absolutely no air within, 

 otherwise the pressure cannot be accurately measured. Insert a perforated 

 rubber cork at g. Fill the manometer from the quicksilver to the extremity 

 of the tube with potassic ferrocyanide, or whatever other solution is to be 

 used in its place, and push it into position in the cork. Fill the cell com- 

 pletely full, and press firmly into place the second perforated cork, taking 

 great care, first, that no bubble of air remains at its base ; and second, that 

 not a particle of potassic feiTocyanide comes in contact with the outside of 

 the cell. 



A bent glass tube, drawn to a capillary point at one end, should now be 

 filled with potassic fen-ocyanide and slowly pushed into the cork. If this is 



