204 Chapter XIII 



to the alveolar air then the blood must have been exposed to a 

 greater pressure of oxygen in the lung than in the alveolar air. The 

 results of some of their experiments are given in the table which 

 faces this page. 



In discussing the work of the Oxford School I have intentionally 

 withheld all reference to their method of estimating the relative 

 quantities of oxygen and carbon monoxide associated with the 

 haemoglobin. Clearly the whole value of the results which they 

 obtained depends upon the accuracy of their estimation and no 

 critical account or investigation of the subject can omit to take 

 cognisance of it. 



The essence of their method is as follows. Some blood is taken 

 from the object of the research before the CO is administered : this is 

 diluted and set on one side : let us call it A. The carbon monoxide 

 is administered : some more blood B is taken from the patient. B is 

 a different colour from A since it contains CO haemoglobin, but by 

 adding carmine to A the colours of the two may be matched. The 

 amount of carmine solution added is noted down, call this x. B is 

 now saturated with CO, its colour again changes, but again A may be 

 made to match it by the addition of more carmine to the extent of 

 y c.c. From the relation of x to y (which is not a simple linear one), 

 the percentage saturation of B with CO may be calculated. 



Apart from the statement made above, that the real estimation is 

 that of the ratio of CO haemoglobin to total haemoglobin, whilst it 

 professes to be an estimation of CO haemoglobin to oxyhaemoglobin, 

 i.e. it assumes that there is no reduced haemoglobin in the blood 

 an assumption which is probably justifiable when there is no oxygen 

 want and but little CO haemoglobin the possible sources of error in 

 the above method fall into two general categories which have to do 

 with (1) the possibility in practice of matching the tints, (2) the 

 theoretical soundness of the method. Haldane and Douglas obtained 

 results which closely agree for any particular sample of blood, though 

 they have told me that the quantities of carmine they use in the 

 operation differ. The method is claimed to be accurate to 2 / . This 

 statement must be treated with great consideration. I have no 

 sympathy whatever with persons who attack a method of this kind in 

 a light-hearted way, and having failed to get consistent results after 

 a few titrations or even a few days spent in the endeavour, condemn 

 the method. The method is essentially a subjective one. The accuracy 

 which can be obtained by subjective methods with long and incessant 



