CHAP. IV.] THE BLOOD. METHODS OF RESEARCH. 



213 



corresponding readings of the barometer. The original yolume of gas 

 to be analyzed is measured as before described, with this difference, 

 that the absolute pressure to which it is exposed is known without 

 reference to the barometric pressure outside at the time. The ex- 

 plosion is effected in the eudiometer, into the upper end of which 

 two platinum wires are fixed for the purpose; the arrangement of 

 these wires is the same as in Bunsen's eudiometer. As to the mode of pre- 

 paring and introducing pure hydrogen and of exploding the mixture, 

 the reader will find sufficient information in Roscoe's translation of 

 Bunsen's Gasometry." 



Description of more simple methods of gas analysis. 



For purposes of demonstration it is sometimes- convenient to employ the 

 following expeditious and far from inaccurate method. 



The tubes for collecting the gases which are strongly recommended are 

 represented in Fig. 47. They are about 250 millimetres long, and are 



FIG. 47. ABSOKPTION TUBE 1 , WITH 



DOUBLE SCALE, AS MADE BY AjLVER- 



GNIAT. (Scale about J.) 



FIG. 48. IRON SPOON EMPLOYED 

 IN TRANSFERRING TUBES FROM ONE 

 MERCURIAL TROUGH TO ANOTHER. 

 (Scale about .) 



1 These tubes are constructed by MM. Alvergniat freres, 10 et 12Eue de la Sorbonne, 

 Paris. From personal observation, the Author can testify to their accurate calibration. 



