TIME SEQUENCE OF REACTIONS 



187 



If now one records the magnitudes of the rapid and slow effects 

 caused by iUumination of the bacteria as measured at various wave- 

 lengths other than the 430 m^u value used in Fig. 2, two difference 

 spectra are obtained, as illustrated by Fig. 3. In the region of the 

 Soret bands, the rapid phase shows a large peak at 428 to 429 m^ 

 with its associated trough at 405 mu. The spectrum is probably due to 

 the "CO-binding pigment." In the visible region of the spectrum, 

 the peaks are too large compared to the Soret band of the "CO-binding 



Peoks thot disoppear on illumination 



-0.010 



E 

 u 



0.005 E 



+0.005 



o. 

 O 



360 



420 



540 



580 



620 



460 500 



Fig. 3. .Difference spectra representing rapid and slow effects follow-ing infra- 

 red illumination of a suspension of R. rubrum. 



pigment" to be associated with it, and are, in fact, so large relative to 

 the total change in the Soret region that pigments other than cyto- 

 chromes may be involved. Nevertheless, there is a superficial resem- 

 blance to cytochrome bands, for example, the 550 m/z peak could be 

 attributed to the a band of a "c" type pigment, but the large 514 m/x 

 /3 band as well as the 574 m/z trough are atypical. The slow phase 

 shows two distinct Soret bands at 420 and 427 m/z attributable to 

 cytochromes of type b and c, respectively. The Soret bands of these 

 cytochromes are more clearly defined because the deep trough at 

 405 m/x noted in the rapid phase does not interfere. There is little 

 evidence for the a band of a cytochrome of type h in the slower 

 reaction, although this band clearly shows in spectra measured at 



