J. W. Evans — Apjiaratus for Isolating Minerals. 69 



end and inserting the piece of glass rod. After the lapse of some 

 time air is re-admitted through h^. 



The above processes may be repeated several times, if thought 

 necessary ; to more thoroughly shake the lighter fragments, the 

 tube 6, may be raised till the end is only just below the liquid. 



If an air-pump is not available, the air may be withdrawn by the 

 lungs, but the effort to overcome the weight of the column of liquid 

 is apt to render the passage of air rather violent. The tube h^ may 

 also be used (without the cork) to agitate the lighter layer by gently 

 blowing just below the surface; or to disturb any fragments resting 

 on the conical part of the apparatus, which may be best done by 

 drawing up a portion of the liquid and then letting it descend from 

 the tube close to the fragments to be moved ; but it is safer to use a 

 pipette for this purpose. 



A few lighter particles may adhere to the tube as it is removed ; 

 if it is desired not to neglect these, they may be washed back by 

 liquid of the same density. 



In order to draw off the heavier particles, the apparatus G and B 

 is used. C is a thistle-tube perforating a cork, and having at the 

 lower end a piece of india-rubber tube (ci) of less diameter when 

 unstretched than G. The india-rubber at the end of G fits tightly 

 into the top of the narrow portion of A. D is a glass rod brought 

 to a conical point at the lower end. Insert it in G so that the end 

 projects some distance beyond the india-rubber Cj, in which it should 

 fit tightly. The whole is then inserted in the funnel. If this is 

 done skilfully, few (if any) of the lighter particles are carried down 

 with it, as the conical end of the glass rod causes a centrifugal 

 movement at the surface as it passes down. To remove any that 

 may have descended, adjust the cork so that the india-rubber is a 

 little above the narrow portion of the funnel ; a little movement and 

 the lapse of a few minutes will be sufficient. If any particles are 

 caught by the lower end of the india-rubber c^, they may be dis- 

 placed by an up-and-down movement of the rod D relatively to C. 

 Now depress C so as to fit tightly in the narrow portion of the 

 funnel, tighten the cork, withdraw the rod D, and open the pinch- 

 cock ; the lower part of the liquid falls with the heavier fragments 

 into a beaker; the tube is then thoroughly washed clear of all 

 residuary particles by a stream of distilled water poured into the 

 cup of the thistle-tube G. 



After letting the tube drain till nearl}?^ free from water, the pinch- 

 cock having been closed, and the tube C removed, the density may 

 be lowered for another separation ; this is best done by gradually 

 adding the diluting fluid through &i or a similar tube, the small 

 aperture being at the bottom of the narrow portion of the funnel. 

 The thin stream of lighter fluid easily mixes with the heavier as 

 it rises, especially if the tube is being gently agitated.' 



1 I have not arranged for the calculation of density from the volumes used. 

 If required, the apparatus could he graduated (heing heightened if necessary). 

 Practically it is better to dilute with a solution only slightly lighter till separation 

 takes place and then determine the density by one of the usual methods. 



