470 SMITH, KUPKE, LOEFFLER, BENITEZ, AHRNE, GIESE 



debris, and the supernatant passed through a French-Mihier honiog- 

 enizer (4). The Uquid was then centrifuged for 1 hour at 10,000 X g; 

 exploratory experiments in the analytical ultracentrifuge had dem- 

 onstrated that this centrifugation removed three relatively fast- 

 moving inactive components. The supernatant was dialyzed against 

 50 times its volume of glycine buffer. In Fig. 4^4. are shown the ab- 

 sorption spectra of the dialyzed solution before and after being illumi- 

 nated. Sedimentation analysis of this solution by means of a Spinco 

 Model E ultracentrifuge consistently showed the presence of two 

 sharp boundaries with characteristic sedimentation rates (Fig. 4Z)). 



After the dialyzed solution had been centrifuged in the Spinco 

 Model L centrifuge for 0.75 hour at 40,000 r.p.m. (approximately 

 95,000 X g) most of the holochrome was left in the supernatant 

 solution, but after a 3-hour centrifugation only a small fraction of 

 the pigment was left in the supernatant. This is shown in Figs. 48 

 and 4E in which the heights of the absorption peak (Fig. 48) and of 

 the right-hand peak (leading boundary) in the ultracentrifuge pat- 

 tern (Fig. 4:E) are both greatly reduced. The small amount of proto- 

 chlorophyll in this solution is still capable of being transformed by 

 light, as the broken line of Fig. 48 shows. 



The pellet obtained by the 3-hour, 40,000-r.p.m. centrifugation of 

 the dialyzed solution was washed once with buffer solution and re- 

 suspended in a volume of buffer about one half that from which the 

 pellet came. This resuspension required thorough stirring and stand- 

 ing for several hours. The solution so obtained showed the charac- 

 teristic absorption of the protochlorophyll holochrome before being 

 illuminated and of the chlorophyll-a holochrome after being illumi- 

 nated (Fig. 4C). The sedimentation picture of the resuspended ma- 

 terial (Fig. 4F) obtained with the ultracentrifuge exhibited the two 

 peaks characteristic of the pictures presented in Figs. 4D and 4E. 

 However, the left-hand peak (trailing boundary) of Fig. 4F is much 

 reduced in size as compared to the corresponding peak obtained with 

 the supernatant solution (Fig. 4E). The greater portion, therefore, of 

 the particles forming this trailing boundary had not been sedimented 

 during the 3-hour centrifugation. In contrast, the leading peak of Fig. 

 4F is considerably increased o\er that of Fig. 4E, showing that the 

 particles represented by this boundary had ])een concentrated in the 

 pellet. The correlation between the sizes of the leading boimdaries 

 in the sedimentation pictures and of the intensities of the proto- 



