22 Chapter II 



haemoglobin point by point, irrespective of laws, and to find out 

 experimentally what it actually was like. 



Fig. 11 shows that the actual curve determined point by point 

 differs fundamentally from Hiifner's hyperbola. The two cross, 

 therefore Bohr's cannot be a hyperbola, and further it has a 

 double contour and therefore is not a curve of the same order 

 at all. 



At the present time it is scarcely necessary to dwell on the theory 

 which Bohr propounded to explain his curve. In two words it was 

 that when oxygen broke from oxyhaemoglobin, the haemoglobin 

 itself split into globin, and something rather like haematin which 

 he called haemochrome. Another explanation of this curve will be 

 alluded to in another place. The whole question took on a new 

 aspect when it became clear, as will appear hereafter, that the 

 affinity of haemoglobin for oxygen is profoundly influenced by the 

 nature of electrolytes (3) in the solution. 



The bearing of this discovery was at once grasped by Ff. Roberts (4) 

 who suggested making a solution after the method of Bohr* and 

 then dialysing a portion of the solution to get rid of such residual 

 salts and traces of ether as might be in it ; then comparing the 

 dissociation curves obtained from the dialysed and undialysed 

 portions. 



The experiment was performed on a solution obtained after 

 three days' dialysis in an aseptic dialyser; it was free from the 

 characteristic smell which always clung to our preparation of dog's 

 haemoglobin made by Bohr's method, it was quite neutral in 

 reaction ; Hardy kindly undertook to test its freedom from salts, 

 and showed that in saline concentration it was equivalent to a 

 004 N solution of sodium chloride. The points on the dissociation 

 curves of the two solutions were then determined, and are shown in 

 Fig. 12. The round points are those of the dialysed solution, the 

 square ones are those of the undialysed solution. It is at once 

 apparent that the latter are in very close agreement with the curve 

 published by Bohr (denoted by II in the diagram), whilst the former 

 fall so nearly on the rectangular hyperbola (denoted by I) as to make 

 it in the highest degree probable that could the salts be entirely elimi- 

 nated the coincidence would be complete. Whilst in our opinion the 

 theory advanced by Bohr to explain his own curve may be set aside, 



* I.e. precipitating the haemoglobin crystals with ether from centrifugalised 

 corpuscles, washing them, redissolving them and shaking out the ether. 



