136 Oxygen in the Venous Blood. I 
cotton wool will prevent saliva from flowing down) and quickly 
introduces the needle in the proximal direction. After a little prac- 
tice it is generally possible to obtain blood at the first puncture. 
If the needle is duJl, unnecessary pain will be caused and the vein 
will probably escape the needle. If the point is too long, it will 
almost invariably hit the opposite wall of the vein before.the point 
is entirely within. The needle is introduced best at an angle of 
about 50° with the surface of the arm. The opening of the needle 
must be kept downward or to the side so that the upper wall of the 
vein will not close the opening when the blood is being drawn. In 
drawing the blood only a gentle suction is necessary, and if air 
should happen to be sucked through the blood the sample must be 
thrown away and a new pipette used. ‘The needle and rubber can 
often be kept 7m sztu while another pipette is connected with the 
rubber. When a sufficient amount of blood (6 to 12 cc.) has been 
drawn, the upper end of the pipette is closed by a finger and the 
needle withdrawn. Then the needle and the rubber are taken 
away from the pipette and the blood is discharged into a cylinder (2 
cm. wide and with a few oxalate crystals in the bottom) below a 
layer of white mineral oil at least 2 cm. deep. Just before the 
last 0.5 to 1 cc. of blood has run out, the pipette is again closed 
with a finger and withdrawn, because the upper part of the blood 
in the glass pipette has been oxidized and must be discarded. 
The bloodein the cylinder is then stirred up with a glass rod to 
secure a good mixing with the oxalate. From the cylinder 2 ce. 
samples of blood are introduced into Van Slyke’s apparatus below 
a layer of evacuated dilute ammonia as described by Van Slyke 
(12). Before a sample is taken the blood must be stirred up 
carefully to secure homogeneity. If the blood has been kept in 
an ice box for some time (see below) it is often difficult to secure 
homogeneity, even after it has gained room temperature. It gets 
more sticky and adheres to the sides of the cylinder. It is there- 
fore advisable to do the determinations on fresh blood. After the 
blood has been stirred with a rod, a 2 cc. pipette is introduced, 
the upper end being closed to prevent mineral oil from coming up, 
and after a few seconds’ additional stirring with the pipette, the 
sample is sucked up and discharged into the apparatus. The 
pipette must be calibrated for 2 cc. outflow, and may be cali- 
brated either for complete discharge or for discharge to a mark 
