94 THE ROYAL SOCIETY OF CANADA 



As at present developed several points must be noticed in the 

 construction. The spiral "c" should be as smooth as possible, and 

 free from joints, or the mercury will break and let air pass, which will 

 also happen if tube is much steeper than 1:5 {\\—}i°), or of greater 

 internal diameter than 3 mm. (slope must be rather accurately ad- 

 justed as at 10° it may stop, and at 13° it will break at the slightest 

 kink). In constructing the top of fall tube care should be taken to 

 have a rather sharp smooth bend, with connecting tube "e" very 

 slightly below highest point. This is necessary as under a vacuum 

 the mercury has a great tendency to break into drops. The drawing 

 represents conditions about as they should be. The capillary "c^" 

 should be about . 5 mm. internal diameter, and should be made by 

 first thickening the fall tube, heating a rather short part of the 

 thickened tube till quite soft, and not drawing out too quickly (to 

 shorten constricting portion and strengthen capillary). The first side 

 of the U-tube should be wide enough to prevent air bubbles being 

 carried over. The length of mercury column carried up in "a" does 

 not need to be longer than 4 — 4.5 cm., using 3 mm. tube; if much 

 longer the shock at the bottom of the spiral will be considerable. 

 There should be a slight constriction at "a" (to ^ internal diameter) ; 

 this makes a slightly greater pressure below it than that represented 

 by the column of mercury being lifted, permitting length of column 

 being regulated by tap at bottom. The bulbs between pump and 

 gauge, etc., prevent mercury being carried over on stopping pump 

 (which should be done by clamping rubber connection to water pump 

 and allowing air to enter slowly through "a"). The bottom of fall 

 tube should taper a little before the bend and be quite thin where it 

 enters the wax seal. This construction makes a sudden jet of mercury 

 through capillary which will efïectually carry quite a small bubble 

 of air through. (Wax seal is used to permit capillary to be perfectly 

 straight and smooth, which would be difficult if a glass seal were 

 used.) 



This pump (spiral 27 cms. high, 18 cm. diameter), when tested 

 with only a McLeod gauge attached (total volume to be evacuated 

 app. 50 c.c), took ^ hour to reach .001 mm. of mercury and with a 

 250 c.c. flask attached reached .0008 mm. in 2>^ hours, and then 

 stopped pumping, but the flask was connected to the pump with 

 sealing wax, which may have leaked a very little. 



I desire to acknowledge my indebtedness to Dr. W. F. Seyer for 

 many helpful suggestions regarding the construction of this pump. 

 Chemical Laboratory, 



University of British Columbia, 



Vancouver, B.C. 



