4 PHOTOCHEMISTRY OF VISUAL PURPLE. I 



tion first turns yellow, and after some time turns completely colorless 

 (wasserhellig) . The discrepancy between the two observations I 

 have not yet been able to explain. 



The velocity of visual purple bleaching varies with the intensity; 

 the rate may thus be controlled to allow a quantitative following of 

 the process. With intensities near 100 meter candles the solutions 

 may take half an hour to bleach. The purpKsh pink solution slowly 

 loses its color, at the same time taking on a yellowish tinge, which 

 increases and remains. The solutions obtained by the present tech- 

 nique do not show a reversible reaction. The bleached solutions re- 

 tain their yellowish color in the dark for as long as they can be kept 

 from bacterial decomposition. The conditions under which reversal 

 takes place are different in several respects, and will be considered at 

 length in future contributions. 



2. The most obvious means of following the process is a colorimetric 

 one. Unfortunately only a small quantity of material is available, 

 and in dilute solution at that. A method has therefore been devised 

 by means of which small amounts of solutions are manipulated ac- 

 curately, and are made to produce the same optical effects as large 

 quantities. The experiments are performed in very small test-tubes, 

 made by sealing one end of short lengths of narrow bore glass tubing. 

 The tubes are about 50 mm. long, and have an inside diameter of 2 

 mm., and an outside diameter of 2.5 mm. Each test-tube holds a 

 little over 0.1 cc. This quantity of solution produces a depth of nearly 

 35 mm. Viewed end-on, the effect is produced of a solution of much 

 greater concentration. 



The crucial point is, of course, to have equal depths of liquid in all 

 tubes of any given set of experiments. This is accomplished very 

 simply. The test-tube is placed into a wider tube which fits it snugly. 

 One end is sealed, and near the open end is a circular scratch. The 

 inner tube rests on the bottom of the measuring tube. The solution 

 is now inserted into the inner tube by means of a pipette ha\ing a 

 long capillary tip, and the depth of the solution is so adjusted that 

 the bottom of the meniscus is on a line with the circular scratch. 

 This can easily be done within 0.2 mm., and, including slight varia- 

 tions in the thickness of the sealed end of the test-tube, would make 

 the error in this part of the manipulation of the order of 1 per cent. 



