Ferrihaemoprotein Hydroxides 121 



(Theorell and Ehrenberg, 1951; Coryell, Stitt and Pauling, 1937; Theorell, 

 1942; Deutsch and Ehrenberg, 1952). In practice, slightly different values of 

 //^ could be used in equations (8) and (9); but, in the absence of independent 

 evidence, the same value will be used for the present calculations. Thirdly, 

 the band maxima for corresponding derivatives of myoglobin and haemo- 

 globin do not occur at exactly the same wavelengths, nor are the extinction 

 coefficients identical, hence small variations would also be anticipated between 

 the high-spin forms, and between the low-spin forms, of both haemoproteins. 

 In those regions where the spectra of both fluoride and cyanide derivatives 

 do differ in this way, correction factors can be introduced by adjusting the 

 wavelength scale and/or all extinction coefficients by the appropriate amount. 

 But even if this is done, the method of calculation leads unavoidably to 

 extinction coefficients, Ej^ and e^, that represent a kind of average of the true 

 values for the two individual haemoproteins. Because of these limitations, 

 reliance can only be placed on predominant features in the calculated absorp- 

 tion spectra, i.e., the positions and extinction coefficients of well-defined 

 absorption bands. If the assumption of a thermal mixture is correct, then 

 these would resemble those for the fluoride and cyanide derivatives; more- 

 over, in no region of the spectrum should the extinction coefficients assume 

 substantially negative values. 



In all but one of the sets of calculations the value of 2-24 B.M. has been 

 adopted for /<j. This is about the minimum obtained for ferric complexes of 

 myoglobin, haemoglobin, peroxidase and catalase, and it has the advantage 

 that fi^ is a whole number, so that, on substituting for ^^^ and /%b» equations 

 (8) and (9) are simply 



26 = 5ajib + 35(1 - ajj^) (8a) 



20 = 5aHb + 35(1 - anb) (9a) 



The fractions of the low-spin forms (a) in ferrimyoglobin and ferrihaemoglobin 

 hydroxide are thus 0-3 and 0-5 respectively. To check how sensitive the results 

 are to the value chosen for fx^, 2-84 B.M. has been used in one set of calcula- 

 tions. This gives fi^ = 8, y.j^ = 0-33 and a^b = 0-55: but as will be seen, 

 this change of about 10% in a does not aftect the type of absorption spectra 

 obtained. 



In view of the approximate nature of the present calculations, a small 

 correction that should be made to the experimental values of /li has been 

 neglected. This amounts to about 0-13 for /n = 2-70, and 0-06 for /< = 5-92, 

 when the values are obtained in the usual way by taking a haemoprotein 

 derivative that has no unpaired electrons, e.g., carbonmonoxy-ferrohaemo- 

 globin, as the reference compound to allow for the diamagnetism of the 

 protein. This method entails the assumption that the paired <^-electrons of 

 the iron atom make no contribution to the magnetic moment, but recent 

 calculations have shown that an orbital paramagnetism, induced by the 



