EFFECT OF TEMPERATURE ON -HvEMOGLOBIN 175 



paper by W. 0. R. King and myself in Vol. xxxix of the Journal 

 of Physiology (X). 



Yet these early experiments, with all their crudities, have their 

 merits, which curiously enough turn on the very absence of mental 

 processes involved in their conduct, whether these processes be of 

 a high or low order, I often wonder at the frequency with which 

 I turn back to some of the pages in Vols, xxxix and XLH of the 

 Journal of Physiology, as compared with those of later papers. And 

 the reason is that the curves given in those papers were made without 

 any preconceived idea about what shape they ought to be if this or 

 that theory were true, or if the views of A were to be preferred to 

 those of B. And so it was a matter of curiosity to me to see what 

 I made of the dissociation curves of oxyhsemoglobin at different 

 temperatures when I did not know what sort of shape a dissociation 

 curve ought to be, and whether temperature had any effect at all. 

 Of course I should have known, for Paul Bert (2), with that wonderful 

 vision which so far outran the experimental methods of his day, had 

 made some observations on the subject. At room temperature he 

 found that blood at 15 mm. oxygen pressure was 90 per cent, saturated 

 with oxygen, whereas at body temperature it was but 50 per cent, 

 saturated. That observation of course contains the kernel of the thing, 

 although it was not made at any specified carbonic acid pressure or 

 anything of that kind. Hiifner(3) also, in 1889, made some observa- 

 tions which tended in the same direction. 



But to return. Brown and Hill(4)i, who studied the effect of 

 temperature on the dissociation curve of blood, have produced a very 

 beautiful series of dissociation curves which as curves for blood will 

 become classical: they do not serve our present purpose, however, 

 which is concerned not with blood but with haemoglobin. We must 

 go back and consider the earlier data of Barcroft and Hill (5). Their 

 curves do not take us any further because it was assumed that the 

 curves were hyperbolic and on that assumption it was considered 

 only necessary to determine one point on each curve, and from that 

 point to draw the curve. Therefore we get back to the original curves 

 of Barcroft and King, which I had always regarded as too crude to 



^ The essential difficulty in the application of curves such as Brown and Hill's (in 

 which the temperature varies greatly while the COg-pressure is constant) to haemo- 

 globin is as follows: as the temperature rises the amount of combined COg alters 

 but little whilst the amount of free COj becomes much reduced. Hence the higher 

 the temperature the more alkaline the blood. 



