THE UNION OF OXYGEN WITH HEMOGLOBIN 129 



that it could be put into such a form by mathematicians as to make 

 it fit either the known curves or others. I tried to draw the curves, 

 making some rather simple assumptions, but without any success. 

 Starting from the assumption of a hyperbola for each reaction I did 

 not get an S-shaped curve, but a curve, the tangent to any part of 

 which cut the x axis to the left of the origin. 



It would appear, as Adair has pointed out, that the theory is to 

 some extent capable of experimental verification. If the haemoglobin 

 solution is in any case very alkahne, even when reduced, the pigment 

 would be present mostly in the form of sodium hsemoglobinate ; if, 

 on the other hand, there is no base in the solution, the sodium com- 

 ponent would be absent. In either case a curve of hyperbohc nature 

 should be found, whilst the double inflection would be most prominent 

 near the isoelectric point. The evidence on this point clearly needs 

 strengthening. 



Some curves worked out by Adair, Bock and myself and published 

 in the Harvey Lectures, 1921-1922, p. 156, are just of the type which 

 would fit in with Henderson's theory, but the subsequent and more 

 exhaustive researches of Adair have yielded curves which, when cor- 

 rected for progressive difference of hydrogen-ion concentration of the 

 fluid as the material becomes oxidised, show almost no change in in- 

 flection from one hydrogen-ion concentration to another. 



Adair's theory {U). According to Hill's theory the value of n as 

 calculated from the dissociation curve of a solution of haemoglobin 

 should be the same as that actually found with an osmometer. 

 It was to test this point that Adair took up the measurement 

 of the osmotic pressure of haemoglobin. In a former chapter the 

 matter has been dealt with in detail. Here it need only be said that 

 in dilute solutions at the isoelectric point Adair finds the osmotic 

 pressure to correspond to a molecular weight of 68,000, or thereabouts, 

 and thus to a molecule which corresponds to four units, in short, to 

 Hb^. It will be gathered from what has already been said that this 

 determination cuts right across all preconceived ideas of the constitu- 

 tion of haemoglobin. Clearly on the hnes of Hill's equation, if n equalled 

 4 the curve of pure haemoglobin would be even more inflected than 

 that of blood, whilst the work which had been done on the subject 

 indicated that in such a condition the dissociation curve would be 

 a rectangular hyperbola. To that point we will return. In the mean- 

 time, passing from dilute solutions, by which is meant solutions of, 

 say, 1 per cent, to 0*1 per cent., to solutions of the same order of 



