1874.] -^^1 [Koenig. 



trade, hold the inflated part of the neck (serving for a hold to the rub- 

 ber) over a Bunsen flame and let it contract slowly at a dull-red heat, un- 

 til the channel has become capilar as shown in figures la, Ic and 2a of 

 the accompanying plate. It needs hardly to be remarked, that during 

 the process, the tube has to be kept revolving, and allowed to cool slowly. 

 The glass wall has become very thick and strong, facilitating the next 

 process of grinding. This is done upon an ordinary rotary grindstone 

 in from 8 to 10 minutes. I grind off one-half of the inflation at a steep an 

 gle, as shown in the figures. The orifice is not required to have a definite 

 size and is naturally given by the points a, /3. The grinding is contin- 

 ued until the elliptic section of the channel has come with its lowest 

 point from about 1-16 to 1-8 of an inch above the lowest point of the in- 

 clined ground plane. 



A very short practice affords sufficient skill to grind a very nearly plane 

 surface. Absolute plaueity is not required. The sides and back are 

 ground next to produce a point, which is necessary for the letting out of 

 small drops of liquid. The ground face stands at right angles to the 

 graduation and may be put either on the right or on the left side, accord- 

 ing to the convenience of the operator. Fig. Ic represents a front view 

 of the ground face, with the capilar orifice at 0. The size of the latter 

 depends on the kind of work which is to be done with the burette, as it 

 influences the size of a drop. On my 20cc burette, divided into twen- 

 tieths, I have a very narrow orifice, a drop corresponding to one-half a 

 division. I use this burette exclusively for argentum nitrate solution. 

 For ordinary alkalimetric work I use a burette (50ccj graduated into one- 

 fiftbs and allow the drops to equal one-tenth cubic in. This opening 

 empties the burette in one minute and a quarter, when running at full 

 stream. 



2, The valve. Platinum in form of a smooth plate is not acted upon 

 materially by any of the solutions now in use for volumetrical analysis. 

 The valve consists of a platinum plate p of elliptical shape, | and 3-16 

 of an inch being the respective parameters. Thickness about 1-32 of an 

 inch. To the centre of this plate is soldered the platinum stem i, the 

 end of which is pierced by an eye. The spring t, made of brass or Ger- 

 man silver and platinated, is screwed to the clamp c, and has a fork at 

 its other end for the insertion of the platinum stem i, forming thus the 

 hinge h. It carries a nut n, through which the screw s passes. In order to 

 open the valve, the screw head is turned, when the screw bolt comes into 

 contact with the glass tube and forces the spring backwards. The valve 

 plate assumes then a position as represented in figure Id, allowing the 

 fuU stream to run straight downwards without the least splashing. The 

 capilar orifice being elliptical, with its long axis parallel to the stream, it 

 is evident that by reversing the screw, the orifice will close gradually, the 

 lowest point the last, allowing a most complete regulation, and when once 

 reduced to dropping a quarter of a turn of the screw will close totally. 

 The only objection to this arrangement of the valve, which has presented 



