TENSION-EQUALIZER UNIT. 



27 



At times this type of absorber proved difficult to make air-tight and 

 another device was substituted in the spring of 1911, which is more 

 efficient. This is shown in detail in figure 6. It consists essentially 

 of an ordinary 2-liter wide-mouth chemical bottle, with a two-holed 

 rubber stopper in which iron pipes are inserted, one pipe, 6, extending 

 to the bottom of the container, the other being considerably shorter. 

 These pipes are of standard size, with an internal diameter of 13 mm. 

 and an external diameter of 18 mm. At the top a short pipe is fitted 

 with a metal tee, a, which is used in filling the bottle. The long pipe, b, 

 is fitted with an elbow, and hose-couplings (c and c) are attached to 

 both pipes by rubber connections. To make sure that no particles 

 of soda-lime enter the piping, the open end of the longer iron pipe is 

 protected by a wire-gauze cap, d, 9 cm. long and 2 cm. in diameter. 



FIG. 6. Carbon-dioxide absorber and accompanying water-absorber. 



a, tee for filling absorber; d, wire gauze on the end of outgoing tube 6; c, c, entrance 



and outlet of air ; /, water-absorber. 



To prepare the absorber for use, the stopper is first removed from the 

 opening in the tee, and the bottle is filled two-thirds full of soda-lime. 

 It is then laid on its side and the tubes and wire-gauze protector are 

 inserted. The bottle is again stood upright and the stopper pressed 

 down firmly; it is then filled through a funnel inserted in the metal tee. 

 After the opening in the tee has been closed by a rubber stopper, the 

 pipes and connections are tested for tightness by a water manometer 

 and then air is blown through them with the mouth to make sure that 

 they are not clogged in any way. These bottles, when properly 

 charged with 2,200 gm. of soda-lime sufficiently fine to pass through a 

 sieve with a mesh of 3 mm., should absorb about 100 gm. of carbon 

 dioxide. This suffices for 15 to 25 experiments with a resting man, 

 each experiment being 15 minutes long. 



