70 HEMOGLOBIN 



but not of all. Adair has, I believe, obtained haemoglobin solutions 

 approaching 40 per cent, in concentration. 



For practical reasons also it is usually very desirable to have the 

 haemoglobin in as concentrated a form as possible. Whether for gas 

 measurements or thermal measurements, the greater the concentra- 

 tion of the haemoglobin the larger the proportion would the readings 

 caused by the effect under consideration bear to those due to ad- 

 ventitious causes and experimental errors. 



The solutions used by Barcroft and Roberts had, as stated above, 

 a haemoglobinometer reading of 70 — in other words they were about 

 9 per cent, haemoglobin. As time went on and technique improved 

 solutions of considerably greater concentration as well as purity were 

 used. It will be noted that the dialysed solution had been at one 

 period as dilute as 4 per cent, haemoglobin and had subsequently 

 been concentrated. 



More recently an attempt was made in collaboration with Adair (7) 

 to manufacture small quantities of very concentrated haemoglobin 

 solution, the properties of which — the conductivity, the hydrogen- 

 ion concentration, the dissociation curve and the oxygen capacity — 

 were known, and of these properties more anon. But whilst these 

 experiments had been taking place, another property of the solutions 

 had gradually been forcing itself upon us, namely, their transparency. 

 Some of our solutions were more transparent than others. We there- 

 fore subjected them to microscopical examination. We had supposed 

 that in the haemoglobin solution there might be "shadows" which, 

 owing to the viscosity and the high specific gravity of the solution, 

 it had been impossible to separate by centrifuging. One film which 

 we made rather bore out this view, for when stained with Jenner's 

 stain the solid matter did not take the stain whilst the solution stained 

 red. The nature of the material with which we were working revealed 

 itself in the following way. If a drop of a relatively transparent 

 solution of haemoglobin be placed on a slide and a coverslip be put 

 on it (thus producing a thin layer of the material) the material 

 presents a very ill-defined appearance. Clearly there is something 

 solid in it, but it is very difficult to say what, or to describe it. If, 

 however, a drop of saline is placed at the side and allowed to diffuse 

 under the coverslip a remarkable transformation takes place. At 

 once the preparation assumes the appearance of blood. The cor- 

 puscles are somewhat ill-shapen as is the case when a solution not 

 quite isotonic with it is added to blood, but it is clear that they 



