444 



STATE BOARD OF AGRICULTURE. 



''a-c", Fig. IT. The circuljiiiiig wheel "w" jjoiiilcd out in the same figure 

 was made to revolve in llie mercury in the direct ion oi' the large arrow. 

 Each pocket in the wheel thus carried gas from the air-chamber in the 

 upper part of the pumj) down into the mercury to the position *'p", 

 when the air emptied toward the center (due to its tendency to rise in 

 the mercury) and came out into the outlet tube "o" as shown in Fig. 

 II B. The pump was therefore able to circulate air against a pressure 

 equal to the mercury-height "s-t", and the confined gas in the pump 

 could not come in contact with the pulley journal "j" under any cir- 

 cumstances, since the lowest level "s", to which the mercury could be 

 brought, was higher than the pulley-journal. As a precaution, a small 



-0 



Fig. II. Diagrammatic projections of the working parts in gas-circulating-pump (see 

 "g" Fig. 1). A = Side-projection; B = Edge-projection; a-c = level of mercury; w= 

 circulating wheel ; p = position of air pocket when it empties toward the center ; p' = air 

 pocket ; o= outlet tube ; s = bearing around the common Central outlet for the air pockets 

 in the circulating wheel ; j = pulley journal. 



cup was set beside each pump (when in use) just beneath the boxing 

 '*k". If a mercury leak should have occurred from this point, the vol- 

 ume of the leak could have been known at once from the mercury caught 

 in the cup. When the apparatus was set up, much trouble was experi- 

 enced in getting all connections sealed gas-tight. Tests were not begun 

 with the extract, however, until the apparatus was able to standi for 

 several hours, under a jDressure greater than any that might occur in 

 the regular experiments, without showing a measurable leak. Moreover, 

 test experiments were run with samples of the same extract in air in 

 both gas-containers of the apparatus until comparatively uniform results 



