370 



Intelligence and Miscellaneous Articles. 



and can only be expelled by well boiling, and not always even then. 

 Hence to substances that are decomposed or fused by heating, this 

 method, in the form hitherto used, is not applicable, or at any 

 rate gives extremely inexact results. Also with substances that 

 are easily held in suspension by water or remain on its surface, as 

 finely divided sulphate of barium, the results obtained are inexact, 

 as these substances do not after the boiling sink again to the 

 bottom, and consequently are in part squeezed out with the surplus 

 water on the insertion of the stopper. 



For the filling of the pycnometer I have used a method already 

 described by me*, in which the disturbing effect of the air is com- 

 pletely excluded. A is the pycnometer, into which the ground 

 stopper a fits ; this is passed through by a glass tube, which is 

 bent back at b and ends at c inside a slip, of which the outer part 

 is attached to the piece clg consisting of narrow glass tubes ; / and 

 e are cocks, h a small funnel, and g 

 the inner part of a slip whose outer 

 part is connected with a good mer- 

 cury air-pump such as should now 

 be found in every laboratory. 



First the pycnometer and the 

 slightly greased stopper are now 

 weighed by themselves ; in doing 

 this it is best to suspend the latter 

 on its hook b by means of a wire. 

 They are then to be connected with 

 the part clg; close e, its cavity 

 having been previously filled with a 

 drop of water ; open /, and exhaust 

 to vacuum ; then let the previously 

 well boiled water flow in from h ; 

 disconnect the pycnometer at c, wipe off: the grease from it and 

 weigh it. Then dry it, introduce the powder, and repeat the same 

 procedure. The tube is bent back at b in order that the powder 

 may not be too much stirred up when the liquid flows in. If the 

 pycnometer, A, selected is a small one and we first exhaust com- 

 pletely all the space between the cock / and the pump, a three or 

 four times repeated action of the pump suffices to make the space 

 in A a vacuum. In the time required for this, even salts containing 

 water, of which the vapour does not possess at all too high a tension, 

 as, for example, the double sulphates of zinc and magnesium with 

 potassium and ammonium, lose only so small a quantity of their 

 water of crystallization that for them sufficiently accurate results 

 are obtained. 



As illustrations of the precision of the method, in the following 

 are communicated two series of measurements, made on powdered 

 glass and precipitated sulphate of barium. They were carried out 



* Wied. Ann. xvii. p. 561 (1882). 



