142 REPORT— 1871. 



closed with a cork, perforated (like the one which closed the upper vessel) 

 by a second thermal junction. A fourth aperture in the mixing-vessel, 

 closed by a stopcock, enabled it to be emptied. 



In determining with such an apparatus whether heat is generated when 

 venous blood becomes arterial, the upper vessel is disconnected from the 

 lower at a point below the glass stopcock previously described ; it is com- 

 pletely filled with water, and then the water is displaced by a stream of 

 pure hydrogen gas admitted through the upper tube. 



The lower glass tube is then connected with the vessel which contains the 

 blood to be experimented upon. The upper tube, through which hydrogen had 

 been admitted, is now connected to the Sprengel pump, which rapidly sucks 

 the blood into the vessel, Avithout the slightest j)ossibility of its coming in 

 contact with oxygen. The upper vessel is either partially or completely filled 

 with blood, but it always is ultimately left in connexion with a hydrogen 

 gasometer. 



The mixing-vessel (the lowest aperture of which has been closed by india- 

 rubber tubing and clip) is now connected to the Sprengel pump, and a va- 

 cuum is formed into which hydrogen is allowed freely to flow. The vacuum 

 is renewed three or four times consecutively, hydrogen being allowed to flow 

 into the apparatus each time. The object of this is to exclude traces from 

 the lower vessel of atmospheric oxygen. 



The stopcock which connects the upper and lower vessels is opened, and 

 venous blood is allowed to flow into the lower vessel. In actual work both 

 the upper and lower vessels are thickly covered with wadding. The upper 

 one is firmly fixed in a clamp, and constitutes a reservoir, which, except 

 when the atmospheric changes in temperature are abnormally sudden, main- 

 tains during limited periods of time a constant temperature. The lower .tube 

 being connected to the stopcock of the upper by means of a flexible india- 

 rubber tube, admits of being comj^letcly tilted, or, if necessary, shaken. 



As soon as the lower vessel contains the blood to be experimented upon, 

 the thermal junctions are bi'ought in connexion with the galvanometer. 

 The amount of deviation on the graduated scale, and the direction of the 

 deviation, at once tells the experimenter whether the upper or the lower 

 junction be the hotter. The lower vessel is thoroughly shaken, then, after 

 some time, the temperature of its contents is determined by reading on the 

 scale placed in front of the galvanometer. The tube and its contents are 

 then repeatedly tilted, a reading of the galvanometer being taken after each 

 set of five tilts. After a certain time the lower vessel has assumed a constant 

 temperature, and readings, at the interval of two or three minutes, show no 

 perceptible change. I may remark that the galvanometer which, through 

 the kindness of Prof. Tait, was placed at my disposal was so set tbat in my 

 various experiments one division of the divided scale corresponded to the 

 100th or the 120th of a degree Cent. The first observations made with my 

 apparatus were intended to determine whether such an amount of agitation 

 as would be required to communicate a thoroughly arterial colour to perfectly 

 venous blood would heat the fluid to a perceptible extent, in consequence of 

 the mechanical work expended in the agitation. 



In preliminary experiments I foimd that venous blood assumed a beauti- 

 ful arterial hue, when it was mixed with oxygen contained in the mixing- 

 vessel, by successively tilting the tube twenty times. In each tilt the tube 

 containing blood and oxygen was completely reversed. In other preli- 

 minary experiments I found that when the tulje contained thoroughly arte- 

 riaUzed blood or water, the process of tilting had no influence on the 



