310 PHENOMENA OF SUCTION. 



The pressure necessary for the required velocity of efflux is bes 

 obtained by means of the reservoir, described in art. 27 (page 255) 

 but a column of water, l m high, is also sufficient to produce th 

 necessary velocity in a glass tube 4 or 5 mm wide. The latter shoul 

 have walls about l mm in thickness, and is provided with a discharg 

 tube, made of sheet zinc, and just wide enough to pass over th 

 glass tube. There is no strain on the zinc tube, and it is not nece* 

 sary to solder the edges over one another ; the edges are brougl 

 close together, as in fig. 203 A, moistened with soldering water, 

 small piece of solder placed upon them and cautiously heated ; whe 

 it melts the solder flows of itself along the edges, fills the interstu 

 between them, and thus closes the joint. A narrower tube of thi 

 metal, 4 or 5 cm long, is soldered to the former, 20 ram distant fro: 

 the end at which the water is to be discharged. The narrower tul 

 must not be inserted into the wider tube, but must be fixed exte 

 nally to it ; it should therefore be made curved at the top by filin 

 so as to correspond to the surface of the tube to which it is to 1 

 soldered. The narrow tube is best prepared by using a cylindric 

 piece of wood or metal, upon which the sheet metal is hammeri 

 into a tube with the mallet ; a tube which is made by simply bendii 

 the metal is rarely smooth and round. The sheet zinc may 

 rendered softer before working it, by heating it to the temperate 

 at which solder melts. When the narrower tube has been fixed 

 the wider, the latter is attached to the glass tube ; the glass tube 

 first heated, a layer of sealing-wax placed upon the hot tube, a 

 allowed to cool ; the metal tube is then heated, and pushed over t 

 glass tube. If this order of proceeding in fixing the tubes upon o 

 another is not strictly carried out, a quantity of sealing-wax is 

 to be squeezed into the zinc tube, and the latter would beco: 1 

 narrower at the precise spot where it is to be wider than the gl 

 tube ; to avoid such an emergency it is further necessary to leave i 

 edges of the end of the glass tube which is inside the zinc tube sha 

 instead of rounding them off. This end should be very near 

 the aperture which leads from the wider into the narrow zinc tu . 

 as shown in fig. 203 B. The part of the glass tube which is b< 

 upwards is connected, at a height of l m above the orifice from wh 

 the water is to issue, with some such contrivance for supplying 

 tube with water, as is shown in fig. 141 (page 197) or fig. 155 (p:j' 

 225). It is, however, better to connect the bent tube wittji 

 separate tube, l m long, by means of a short india-rubber tube i' 

 with sealing-wax. The width of this tube should be 8 [or 10 r i; 

 such a wide tube is not easily broken by the weight of the water' 

 the funnel, or other contrivance for filling it at the top, while i<> 



