473 



S. S. Brody and M. Brody 



formation of aggregates with a final effective size 3.8 molecules (see Table 

 II). This size is somewhat larger than the three nnolecules given by 

 Rodgrigo (47) who, however, made his measurements with solutions of chloro- 

 phyll at concentrations lower than those found in vivo . Some of the possible 

 molecular weight distributions, as well as mechanisms of aggregation - which 

 give rise to the effective sizes we have found, will be discussed elsewhere. 



D. Number of Florescing Forms of Chlorophyll in Euglena 



The broad emission band observed in Euglena most likely corresponds to an 

 envelope of several overlapping spectra. Before an evaluation can be made of 

 the contribution of each fluorescing component to the envelope it is advisable 

 to determine the number of components giving rise to the fluorescence 

 envelope. Weber (58) has described a method for enunnerating the components 

 in such complex systems on the basis of fluorescence spectroscopy. 



Weber's method essentially consists of setting up a matrix as shown below, 



where F^ 



is the intensity of fluorescence obtained at wavelength n, upon 



WAVELENGTH OF EXCITATION 



i 9 



_lO</) 



excitation at wavelength m. All possible 2x2, 3x3, 4x4, etc. , deter- 

 minants are evaluated. When all the 2 x 2's are zero, there is only one 

 fluorescent form present; when the 3 x 3's are zero, there are only two 

 fluorescent forms, and so on. In our experiments the measurements of F^^ ^^ 

 each have an estimated error of 5%, therefore, when we say the determinant 

 is zero, it means it is zero within our experimental error. The fluorescence 

 spectrum under study covers the spectral range 670 to 770 m|j.. The 

 intensity of emission is measured at 5 m|ji intervals over this range so that 

 as many as 21 excitation spectra can be obtained. The excitation spectrum 

 which corresponds to a particular wavelength in the emission spectrum is 

 punched, at 2 m|j. intervals, onto IBM cards, resulting in 125 to 175 points for 

 each excitation spectrum. 



The matrixes were calculated on an IBM 7090 coinputer. There is no 

 problem in doing all the possible 2x2 matrices. However, because of the 

 long time it takes to evaluate all the possible 3x3 and 4x4 matrices in a 

 21 x 150 matrix, some of the data had to be eliminated. 



Fluorescence excitation spectra were determined, at room and liquid 

 nitro.gen temperatures, for dark grown Euglena gracilis which had been 

 allowed to green for periods of time ranging from 6 to 80 hours. With these 



