4 Chapter I 



Till recently there have been two schools of thought with regard 

 to the meaning of the figures given. Of these the first teaches that 

 the sources of analytical error are so great as to make more accurate 

 analysis impossible, whilst the second, represented chiefly by the late 

 Professor Bohr of Copenhagen, frankly admits that the want of uni- 

 formity is so great as to render untenable the idea of haemoglobin 

 as a simple substance. He explains the divergencies which we have 

 noted as being due to a mixture in different proportions of four 

 substances which he calls a, /3, 7 and S haemoglobins' 7 ', each with a 

 different oxygen capacity from the others. 



To the two schools mentioned above has now been added a third 

 which teaches that the combination of oxygen and haemoglobin is 

 not in the old-fashioned sense a chemical combination at all, but that 

 it is a manifestation of the physical phenomenon known as adsorption, 

 and that it therefore depends essentially on the surface conditions of 

 the molecules of oxygen and haemoglobin respectively. The pro- 

 perties of these surfaces may presumably be altered by all sorts of 

 variations in the collateral substances which are present in the 

 solution of the uniting molecules. The amount and nature of the 

 salts present for instance might be supposed to alter the charges on 

 the molecules, and in so doing to affect the amount of oxygen with 

 which a given quantity of haemoglobin would unite. 



Within recent years, partly on account of the improvements in the 

 analytical methods both for oxygen and iron and partly on account 

 of the increased importance of the subject, it has become more and 

 more desirable that some re-investigation by direct methods of 

 this specific oxygen capacity of haemoglobin should take place. I 

 mean by methods in which the oxygen is measured as such, and the 

 iron as a salt of the metal, as opposed to the indirect spectro-photo- 

 metric methods which have given uniform, and apparently excellent 

 results in the hands of Butterfield (6) . This investigation has lately been 

 undertaken by Peters (8) ; for the purpose of estimating the oxygen he 

 has used the differential method of blood analysis based upon the 

 observation of Haldane (9) that oxygen is eliminated from haemoglobin 

 quantitatively by potassium ferricyanide. The theoretical accuracy of 

 the method has been confirmed by the researches of Professor Franz 

 Miiller (10) in Berlin, while its practical details have been so far simpli- 

 fied that some half-dozen analyses can easily be performed with as 

 many cubic centimetres of blood in two hours. Peters therefore has 

 been at a great advantage as compared with his predecessors, whose 



