9 8 THOMPSON YATES AND JOHNSTON LABORATORIES REPORT 
The apparatus necessary are an ordinary alcohol lamp, a high speed centrifuge, a 
few capillary pipettes, a diluting fluid — we used a 1-5 per cent, solution of sodium 
citrate in normal saline — and a special bulb-shaped tube capable of containing -25 to -5 
c.cm. (Fig. 3). These tubes are easily drawn out in the flame of a bunsen from 
easily-fusible glass tubing having an internal diameter of a little more than five milli- 
metres. They are afterwards cut with a file to a length which fits the haematocrit arm 
of the centrifuge. 
It is very necessary for the successful use of these tubes that each step should 
be done quickly. The patient's finger must be well pricked so that the blood flows 
freely and in fair quantity. The longer arm of a perfectly clean tube — if it is greasy 
blood will not enter — is gently touched to the drop of blood and the tube allowed to 
half fill by capillarity — breathing through the tube just before using it will cause the 
blood to enter more easily. The tube is then quickly transferred to the diluting fluid and 
allowed to almost, never totally, fill itself. The filled tube is gently rotated so as 
to thoroughly mix its contents, and then, held so that there is an empty space at 
either end, the extreme tip of the longer arm is placed in the flame of an alcohol 
lamp and allowed to seal. If the mixture of blood and diluent becomes coagulated 
by heat, or if an air bubble forms in the tube before it is sealed, it will be found better 
to commence again than to attempt to remedy either defect. The best results arc 
mutrc 
est' In 
Fig. 3 
obtained from tubes so prepared if they are first ccntrifugalized rather slowly and later 
at a much higher speed, for instance, for five minutes at eight to nine revolutions per 
minute of the handle of Deland's centrifuge, and then for two minutes at sixty to 
seventy (about eight thousand revolutions per minute). At the end of this time a 
well-marked white ring will have formed above the red cells. The undesired, over- 
lying, clear fluid can be quickly removed by gently scratching the tube at about three 
millimetres above the white ring with a file and then breaking off the bulb by a 
smart tap from the same tool. Any superfluous diluting fluid is first removed, and the 
