H. L. Wells — Self-feeding Sprengel jPump. 



391 



under pressure is admitted at A, it flows into the small bottle 



then up through the tube B. On the tube of 



Mj in the bottle, is a valve Y which allows a -^' 



liquid to flow into the bottle, but closes when 



the pressure is in the opposite direction. Now 



if mercury is placed in the reservoir M it will 



flow down into the bottle, provided the column 



of mercury more than counterbalances the 



column of water in B and the pressure caused 



by the friction of the current passing through 



it The conditions just mentioned are easily 



attained by making the mercury column about 



one-fifth or one-fourth the height of the water 



column and regulating the flow of water so that 



the friction is not very great. Under these 



circumstances the mercury rises in the tube B 



to a height somewhat less than its level in M, 



the valve then closes, and, if the bore of this 



tube is not too great, the mercury which has 



risen in it must be carried up to the top by the 



current of water. As soon as this mercury is 



forced out at the top of B, the valve opens 



again and the whole operation is repeated. 



This apparatus, which may be called a 

 mercury elevator, must be seen in operation 

 to be fully appreciated. It works rapidly and 

 with great efficiency, and the automatic action 

 of the valve makes it an interesting piece of 

 physical apparatus. 



Two forms of valves have been used with LM' V 

 good success. The well-known Bunsen valve 

 is shown at Y in fig. 1. For this purpose it should be made of 

 a quite heavy rubber tube so that it will not collapse under the 

 pressure. Another form of valve is shown in fig. 2, 

 which is perhaps somewhat preferable since it cannot 

 collapse under the pressure, and if properly con- 

 structed it delivers mercury more promptly than the 

 other. It is made by grinding a rather thin-walled 

 glass tube on a grindstone until a narrow opening is 

 made as shown in b ; a soft, rather thin rubber tube, 

 fitting the glass tube snugly, is cut as shown in a, 

 somewhat less than half the section of the tube being 

 cut away. The rubber tube is then drawn over the 

 glass tube so that the openings are on opposite sides. 

 Whichever valve is used, care must be taken to have it large 



Am. Jour. Sci. 

 26 



-Third Series, Vol. XLI, No. 245.— May, 1891. 



