713 



William F. Prickett, F. Dudley Bryant, and Paul Latimer 



However, we did not use thenn to attempt to account for the scattering curve 

 in Fig. 1 because the refractive index at these wavelengths is not sufficiently 

 well known. As discussed elsewhere'^', this large particle theory predicts 

 that the relation between the scattering and absorption spectra will depend - 

 amongst other things - on particle size. For some sizes (e. ^. , small parti- 

 cles) the scattering curve is S - shaped near the absorption maximum as is 

 the dispersion curved ^^K For other particle sizes, the predicted shape of the 

 scattering curve will be that of an inverted absorption curve (minimum scatter- 

 ing at wavelengths where maximum absorbance occurs). The above scattering 

 spectrum seenns to be a combination of these forms. 



While "large particle" equations successfully predicted certain optical 

 properties of red blood cells, a more severe test is obtained by applying one 

 of them to scattering by chloroplasts (diameter 3 - 4 p) near the red chloro- 

 phyll maximum, A = 0. 68 p (diameter /wavelength in water = 7). Not only are 

 the particles smaller, connpared with the wavelength, than are red cells but 

 they have more internal structure. A suitable spectral scattering curve and 

 the needed auxiliary experimental infornnation for spinach chloroplasts have 

 been reported'^' ^' and in the following paragraphs we use this to test the theo- 

 ry. The data is very similar to that which we recently obtained using a special 

 spectrophotometer and partial single layers of chloroplasts. 



All needed information about the particles is available except definite in- 

 formation about the structure of the chlorophyll a band in vivo . (Does chloro- 

 phyll a exist in vivo in one, two, or more forms with different band maxima in 

 this region?) The absorption band structure is needed for the dispersion theo- 

 ry calculations of the dependence of particle refractive index on wavelength. 

 Since the structure of this band is not known with certainty, we calculated 

 three dispersion curves and used each to calculate a scattering curve for com- 

 parison with the experimental scattering curve. Dispersion curves were cal- 

 culated from band parameters which in turn were obtained by trial fittings of 

 a corrected experimental absorbance curve. In this fitting, a basic model of 

 the band structure is assumed. Then all band parameters (position, height, 

 and width) are adjusted, automatically on a computer, to obtain the best fit. 

 In all resulting fits a chlorophyll b band was assumed at 654 m^ and also 

 minor bands at shorter wavelengths. In case I, the major chlorophyll a in 

 vivo band was assumed to be a single Gaussian shaped curve with a maximum 

 at 676 m^. In the second case, II, the band was assumed to be composed of 

 two spectroscopically distinct fornns present in about equal concentrations 

 with band maxima at 671 and 682 m^. Finally in case III a third band was add- 

 ed at 697 m|a, its magnitude being about 10% of the sum of the 671 and 682 mji 

 bands. Our scattering equation leading to Sp - values used here was a modified 

 version of Eq. 10 of Ref.(5). The three resulting theoretical scattering curves 

 and the experinnental curve^^^ are shown on Fig. 2. 



While there is qualitative agreement between theory and experiment, the 

 large particle equation predicts much less dramatic selective scattering than 

 actually occurs. Furthermore, the absolute theoretical scattering cross sec- 

 tion is of the order of 1. while the experimentally observed cross section is 





