66 HEMOGLOBIN 



in boiled distilled water in the glass cylinder. The tube, like the cylinder, was rendered 

 aseptic with formalin and similarly washed. The dialysis was allowed to proceed 

 for 2 days during which time the cylinder stood in ice. The water was changed 

 at intervals of a few hours during the day-time. The earlier samples of water which 

 had been used for the dialysis gave a precipitate when tested with silver nitrate, 

 the later ones did not. At the end the haemoglobin solution was without odour of 

 any kind. Clearly then the cause of its original odour, as well as all suspicion of 

 formalin, had disappeared. 



During the dialysis the haemoglobin solution became considerably diluted and at 

 the end it gave a reading of approximately 30 with Haldane's standard haemoglobino- 

 meter. It was necessary therefore to concentrate it. This was done by distillation 

 over a water bath at 40° in an ordinary distillation flask fitted to a condenser. The 

 condenser was surrounded by ice; the whole was kept vacuous by the action of a 

 Tcepler pump fitted with a drying chamber which contained sulphuric acid. The 

 concentration was allowed to proceed until the haemoglobin solution gave a reading 

 of 70 on the hsemoglobinometer — ^which was identical with that of portion B. 



Portion B was kept in ice during the dialysis of ^. To start with it had a haemo- 

 globin value of 85, and before use it was filtered through a new Berkefeld filter which 

 reduced its haemoglobin value to 70. 



The solutions used subsequently by Barcroft and Hill (4) were made 

 without ether or alcohol. The corpuscles were centrifuged and washed 

 in the manner described above. The apparatus used for dialysis was 

 as described, but the mass of washed corpuscles, not that of crystals, 

 was put straight into the dialyser. The assumption was that as the 

 salts dialysed out the corpuscles burst and haemoglobin was set free. 

 That this must have been so to a large extent is clear from the fact 

 that there were often considerable masses of haemoglobin crystals in 

 the bottom of the dialyser, the solution apparently becoming super- 

 saturated. One point may be noted here, namely, that the solutions 

 made in this way never filtered so easily as those made from the 

 resolution of the crystals. 



Three other methods of obtaining haemoglobin crystals on a con- 

 siderable scale may be mentioned. Each of them has considerable 

 merits; they are (1) the method of Dudley and Evans (5), (2) that of 

 Heidelberger(6), and (3) that of Parsons. 



(1) The method of Dudley and Evans depends upon the fact 

 that oxyhaemoglobin is less soluble than reduced haemoglobin. If 

 therefore a saturated solution of reduced haemoglobin be oxidised, the 

 crystals of oxyhaemoglobin will separate out. 



In the application of the method horse corpuscles were washed 

 by centrif ugalisation until the washings were free from protein and 

 dialysed in collodion sacs against distilled water under the osmotic 



