BEN T LEY GLASS 833 



fcrcnt exciting wa\cleiigtlis, the relative tlirections of the transition 

 moments of the chflerent electronic bands may be determined; (2) 

 by niaking measurements of changing polarization with change of 

 concentration at a given wavelength, the transfer of electronic energy 

 among like molecules may be estimated. The polarization spectrum 

 is best studied in rigid solutions, such as propylene glycol at — 70° 

 C. Molecules of the phenol series exhibit two regions of the spectrum 

 with polarization at right angles to each other. The polarization 

 spectrum of the indole-tryptophan series is more complex. A flat 

 maximum at 270 mfx and a sharp minimum at 295 m^u, indicate an 

 independent transition at a large angle. The very high maximum 

 at or just above 300 myu, is interpreted on the basis of the Franck- 

 Condon principle rather than as being an independent electron tran- 

 sition. N-glycyl tryptophan has an additional polarization band, at 

 290 ni/i, — unexplained. 



The second type of polarization measurement reveals depolariza- 

 tions by nonradiative transition. The reciprocal of the linear polari- 

 zation of the fluorescence is a linear function of the molar concen- 

 tration (so long as very thin layers, 30 fj. are examined) . The slope 

 constant R,, represents the distance between parallel molecules at 

 which there is a 50 per cent probability of transfer of the excited state 

 from one to the other. Observed values range from 15 A to 40 A, and 

 are independent of the exciting wavelength in the case of phenol, but 

 not in the case of indole, which shows transfer of excitation at 14 

 A with wavelengths below 290 m/^,, but none that is observable above 

 305 nxfj.. 



Proteins may be grouped into those containing tyrosine and pheny- 

 lalanine but no tryptophan, and those containing tryptophan as well. 

 Phenylalanine contributes very little to fluorescence, however. The 

 first class of proteins includes insulin, tropomyosin, zein, ribonu- 

 clease, and several trypsin inhibitors, all of w^hich resemble tyrosine 

 in their fluorescence and polarization spectra, although polarization 

 is somewhat reduced in amount. It is clear that some transfer of the 

 excited state between tyrosine residues in the protein must take 

 place, but the distances involved and the spatial relations are not 

 known. 



The second class of proteins includes most globular proteins. Their 

 fluorescence spectra are characteristic of tryptophan alone (with 

 maxima ranging from 332 m^ in the case of chymotrypsin to 346 rufj. 

 in the case of pepsin) . From the difference one may conclude that 



