106 RADIATION BIOLOGY 



For band I the data on compounds 2 and 7 show further that the 

 1- and 2-vectors must cancel and the 3- and 4-vectors add. Examination 

 of the a-substitution data shows that the a-vector cancels the 3-vector. 

 These two facts are consistent with the tentative theoretical predictions 

 of Fig. 2-20a if we suppose the essential nodal plane of band I passes 

 between the 1 and 2 positions, or more generally through that pair of 

 opposite rings which has the most or strongest o,p-directing substituents. 

 (These are the II and IV rings in that large group of porphyrins which 

 has a strong meta-directing substituent in the 6 position.) It seems 

 chemically reasonable to suppose that the central hydrogen atoms will 

 tend to be on this pair of rings. 



A similar, but completely independent, inference from the band III 

 intensities shows that the essential nodal plane of band III passes through 

 that pair of opposite rings which has the least or weakest o,p-directing 

 substituents. 



Consequently the nodal plane of band I probably coincides with the 

 H-H axis, and the ^A-^Q° bands I and II are probably polarized perpen- 

 dicular to this axis. The nodal plane of band III is probably perpen- 

 dicular to the H-H axis, and the ^A-'^Q° bands III and IV are probably 

 polarized parallel to this axis. 



Hydrogen addition, as in the di- and tetrahydroporphins, is sufficiently 

 different from ordinary substitution to be anomalous. It acts on the 

 intensities like a positive, or o,p-directing, moment, but it undoubtedly 

 pushes the central hydrogen atoms away to the other rings [the I and 

 III rings in the usual designations (Fig. 2-20c)]. The intensities can be 

 accounted for if the polarizations are determined with respect to the H-H 

 axis according to the rules just given, not if they are determined with 

 respect to the (apparent) strong o,p-axis. 



The deduction, from this reasoning, that band I will be perpendicular 

 to the H-H axis in tetrahydroporphin agrees with the earlier conclusion 

 of Longuet-Higgins and coworkers (1950), which was based on empirical 

 analogies with condensed-ring spectra and on semiclassical arguments. 

 [It does not agree with Kuhn's predictions (1950), but the amino nitrogen 

 atoms were neglected in his treatment.] 



The assignment of the low transition as the transverse ^A-^Qx implies 

 that the highest transition of the f-g group in porphin (with degeneracy 

 removed) and in tetrahydroporphin must also be transverse and so must 

 be the ^A-^B^, with the two longitudinal transitions ^A-^Qy and ^A-^By 

 in between. This agrees with the assignments of Longuet-Higgins et at. 

 The ^A-^Bx transition in chlorins and bacteriochlorins will be weaker 

 than its longer-wave-length companion, the strong ^A-^By Soret band, 

 because much of its intensity has been contributed to the ^A-^Qx tran- 

 sition, as we shall see. 



Empirical Determination of the Vector Directions. Returning to the 



