624 Scientific Proceedings, Royal Dublin Society. 



termination into a funnel of a little larger diameter, at the brim, 

 than that of the limbs of the U-tube (fig. 3) . 



The apparatus is first filled with water up to the fixed mark, 

 and weighed ; this weighing, after deduction of that of the tube 

 itself, gives the weight of the volume of water required to fill the 

 tube as defined by the mark. If it were possible to introduce now 

 a weighed quantity of the powdered mineral, the density of which 

 it is desired to ascertain, we should then simply have to draw off 

 the excess of water till the fixed mark were reached, and the next 

 weighing would give the weight of the water displaced by the 

 mineral, and from this the specific gravity could be obtained at 

 once. 



This, indeed, is the principle employed ; but the details 

 are slightly modified. The surface-tension of water is so great 

 that the powdered mineral, although well cleaned by the action 

 upon it of the mercury and potassium iodide used in its 

 separation, will not all sink through, but a small portion 

 always floats as scum on the surface ; it is therefore necessary to 

 substitute some other fluid for water, and that which I have found 

 succeed as well as any is the common paraffin used for burning in 

 lamps. The specific gravity of this having been ascertained by 

 means of the tube, the latter is filled with paraffin up to the brim of 

 the funnel, and the powdered mineral introduced from a small 

 weighing bottle. It rapidly falls through the paraffin, not a grain 

 remaining on the surface. There is no difficulty in completely 

 filling the limb of the tube from which the funnel proceeds, and 

 thus results of accuracy can be obtained at a minimum expenditure 

 of time and trouble. It may be useful to add a few suggestions 

 on the filling and emptying of the apparatus. In filling, the tube 

 should first be inverted (fig. 4), and care should be taken that the 

 end of the immersed capillary does not accidentally emerge from 

 the fluid. By sucking at the funnel end the liquid readily enters. 

 When one limb has been completely filled up to the narrow con- 

 striction by which it is joined to the other the apparatus should be 

 restored to the upright position (fig. 5), care being still exercised 

 that the immersed capillary does not leave the fluid. The empty 

 limb now fills by outflow from the full one, which is supplied 

 by the siphoning action that takes place, now that it is in an 

 upright position. By this method of filling air bubbles are 



