102 LIGHT AND LIFE 



proof of an inter-tryptophan transfer, since each tryptophan residue 

 may be considered to He within the distance from DPNH at which 

 a considerable transfer may be expected. 



The large p-iK,/p2'o ratio, as well as its decrease in concentrated 

 urea solutions, can be best explained as the result of interactions be- 

 tween the tryptophan residues and the protein backbone (28) . In 

 future work on this problem, one should aim at answering the ques- 

 tion of the specificity of this interaction, and the participation of 

 definable structural characteristics of the protein in the production 

 of the details just mentioned. The importance of such studies for 

 the knowledge of piotein structure need not be stressed. 



Artificially Conjugated Proteins 

 The studies carried out upon dimethylaminonaphthalene con- 

 jugates of proteins by Weber (25) and by Steiner and McAllister (22) 

 have established the usefulness of fluorescence polarization in the 

 determination of rotational relaxation times. It is very satisfactory 

 that the values obtained for several globular proteins agree well with 

 previous results of Oncley (13) by the dielectric dispersion method. 

 Such an agreement can only by expected if the protein charges present 

 at the isoelectric point can be considered virtually bound to the 

 protein and incapable of showing a characteristic relaxation frequency 

 of the order of magnitude of that of the protein itself (10-" sec). 



I will consider finally two applications to which these conjugates 

 can be put, namely, the determination of the dielectric properties of 

 the medium with which the fluorescent label is in contact and its 

 variation under a variety of circumstances, and further, the determina- 

 tion of what may be called molecular expansion, by observations of 

 the change in electronic energy transfer from the tryptophan residues 

 to the attached DNS. 



The determination of the dielectric properties of the environment 

 by observations of the fluorescence spectrum is a simple application 

 of the theory of Lijjj^ert already discussed. The fluorescence spectrum 

 of conjugates of bovine serum albumin with dimethylamino naphtha- 

 lene sulfochloride is found to have a maximum at 500 ni/x, which 

 after twenty hours' digestion with chymotrypsin shifts to 520-525 nifi. 

 The shift towards longer wavelengths follows the replacement by 

 water of the less polari/ablc protein residues found in the intact mole- 

 cule in the vicinity of the attached DNS. Absorption spectra, and 

 also fluorescence-excitation spectra, show a shift in the absorption 

 maximum of some 10 m^a toward the ultraviolet after proteolysis, as 



