28 RESPIRATORY EXCHANGE OF ANIMALS AND MAN 



be very rigid to withstand the water pressure. Moreover, the con- 

 nections must be made movable when the animal chamber has to be 

 lifted in and out of the bath, which certainly does not make for tightness. 

 The best general form of an animal chamber is undoubtedly that 

 described by Grafe [1909] for his Jaquet apparatus, which has 

 been adopted also by Benedict [1912] and by Fridericia [1913]. It 

 consists of a thin-walled metal box of suitable dimensions. Such a 

 box can be made of tinplate, zinc, or tinned brass, and made airtight 

 by soldering. The aperture for introducing the animal is closed by 

 means of a fluid seal which is always absolutely effective. The fluid 



FIG. 6. Benedict's respiration apparatus for small animals. 



seal can be at the top of the box (fig. 6, H, K), which is the best plan 

 for smaller animals, or around the floor, in which latter case arrange- 

 ments must be provided to lift the whole of the box. 



The volume of the animal chamber and indeed of the whole respi- 

 ration apparatus must be determined, but the accuracy need not be 

 particularly great It is best therefore to give it a simple form which 

 can be easily measured. Heide, Klein, and Zuntz [1913] have de- 

 scribed a method by which the volume can be determined by intro- 

 ducing a measured volume of a gas which can be analysed accurately 

 (carbon dioxide will usually be most suitable), mixing and analysing 

 the resulting mixture. 



As circulating pumps rotary blowers, as introduced by Atwaterand 



