6 
Mr Barcroft, Apparatus for the 
according as the stopper is turned. Besides the passage F, the 
stopper is bored from end to end (. E , G). The orifice E is drawn 
off into a tube which is closed by a tap H. Perhaps the 
mechanism will be made most clear by a description of the actual 
manipulation in the measuring of a sample of blood. The whole 
burette, and its connections, are filled with mercury by raising the 
reservoir J, the tap H is opened so that the mercury may rise and 
fill the passages, G and E, in the stopper A. (The bottom of the 
stopper has been carefully ground with sharp angles at K to 
prevent bubbles of air getting caught at the sides.) H is then 
closed and the stopper turned so that F is continuous with G. 
Thus the passage F and the tube G are filled with mercury. 
By means of a small piece of india-rubber tube G is connected 
with the cannula which is already full of blood. F is now again 
made continuous with G , the reservoir J is lowered a little, and 
thus some blood, and any air that is in the cannula finds its way 
into B. Perhaps 1 c.c. of blood may be taken at this juncture. 
The stopper is now turned through 90° so that F becomes blind. 
J is raised, H is opened, and when the blood and air have been 
expelled H is closed again. After this preliminary the experiment 
may begin. Blood is drawn into B by the process just described, 
the surface of the mercury in J being kept at the same level as 
that of the mercury in B, except when some special manipulation 
is required. When about 8 c.c. has been collected, and the reading 
on the graduated scale of B taken, that stopper is turned through 
180°, so that F becomes continuous with D. As soon as this takes 
place all the blood is sucked into the vacuous receiver and some 
mercury is allowed to follow to wash the blood over. 
The reading of the blood-mercury surface on the graduated 
scale has been taken. The question arises, “Supposing this 
reading to be, say, 7 c.c., how much blood is actually delivered 
into the receiver ? What constant must be added to this reading, 
which starts from an arbitrary zero, to get the exact amount of 
blood which is subjected to the vacuum ?” The answer to this 
question is : The burette must be calibrated by taking out the 
stoppers from beneath each of the froth chambers successively, 
and allowing the blood measured in the burette to flow out of 
the apparatus through the stopperless tap into a graduated tube. 
The reading in the tube may then be compared with that of the 
burette. In such a calibration the first reading in the graduated 
tube will be fractionally lower than the successive ones. To 
correct for this, a small bulb is placed at the end of what will 
hereafter be described as the “ blood-main,” so that a film of blood 
may be made along the glass before the first sample is taken into 
one of the vacuous chambers. 
Passing from the measuring apparatus to that which repre- 
