114 



placed at the point where the capillary tube joins the bulb. The 

 first act is to join this capillary with a tube leading from a tank 

 of mercury placed high enough above the apparatus to ensure a 

 prompt flow into the chamber when the glass cock on the reser- 

 voir is opened. The other capillary, which is permanently a 

 part of the bulb, is also brought down under mercury, and there 

 joined to a very short L, ending in a point directed upwards ; this 

 is completely immersed in mercury. After the objects in the bulb 

 have been allowed to respire for some definite recorded period, 

 samples of the gas within may be easily taken by simply opening 

 the pinchcock and very cautiously turning the stopcock connected 

 directly with the mercury reservoir. This will, of course, force 



FlG. i. a, respiration chamber; b, delivery tube; <r, tube attached to mercury 

 reservoir. Both b and c are immersed in mercury, d, place where screw compressor 

 is attached. The rubber tubing is shaded, all the rest is of glass. 



out bubbles of gas from the other end of the apparatus, which 

 may be readily collected in the tube held over the nozzle. This 

 being done, the tube is transferred to the gas analysis apparatus, 

 preferably one of the Bonnier-Mangin type and the C0 2 deter- 

 mined. All of this is very simple, but of course there are cer- 

 tain precautions which must be taken and certain procedures 

 gone through with if the results are to be quantitative. To 

 begin with, it is necessary to determine the volume of gas within 

 the apparatus, unless only a purely comparative result is re- 

 quired. This process presents, it is true, a little manipulative 

 difficulty, but nothing that is insurmountable. It is most easily 

 done by weighing the apparatus, fitted up in exactly the manner 

 .in which it is to be used, first empty and then full of water. In 



