EFFECT OF TEMPERATURE ON HEMOGLOBIN 177 



points (this line, as stated in Barcroft and King's paper, was Bohr's 

 line (6) which nearly fitted the points), and (2) the considerable extra- 

 polated portions have been cut off, except at the top of the 38° line, 

 where we have Bohr's work to guide us. 



Now as regards the relation of these curves to one another : imagine 

 five horizontal fines drawn corresponding to saturations 90, 80, 64, 

 50 and 40 per cent. The pressures at which the curves intersect 

 these horizontals are as f oUows : 



If now each of these pressures be expressed as a percentage of the 

 pressure at 38° C. for the same saturation the table appears as 

 follows : 



The identity of the above percentage pressures for the various fines 

 is, I think, quite remarkable and certainly the variations are not 

 greater than the experimental errors involved. The greatest deviation 

 is the lower portion of the 32° line. This deviation depends upon one 

 point, an error of 1 mm. in the pressure of that point would bring it 

 into good agreement with the rest. 



The curves all seem to be of the same degree of inflection (in 

 Hill's notation n = 1-8 approximately): we may therefore proceed 

 to consider the impfication of that fact. I am indebted to Adair 

 for the following statement of the fact that identity of inflection is 



