33 



periment to produce triplets. Freezing at — 1°C would be suffi- 

 cient, only such slow freezing would favor the formation of pure 

 ice which leaves the dye behind in a small volume of mother 

 liquor. The deep freezing is needed to have the dye enclosed in 

 the rapidly forming crystals, but once this has happened we can 

 maintain the triplet state even at the melting point. The same is 

 true for the other substances studied, such as acridine orange, 

 acridine, quinidine, and riboflavine. The light emission fades out 

 in these substances at different temperatures, in acridine orange 

 around — 20°C, in riboflavine around — 3°C; the blue emission 

 of pure quinidine and the yellow emission in the presence of 

 0.001 M KI fade out around — 20°C. So the real color of the 

 triplet can be seen best closer to the melting point where the ice 

 is optically less inhomogeneous. 



This shows clearly what is important for the triplet state is not 

 the low temperature but the physical state of water, the formation 

 of ice. Ice is not just solidified water. As is generally known, mod- 

 ern physics puts less emphasis on the idea of "solid" and "liquid" 

 than on "regular" and "random," and the "solid state" owes its 

 specific qualities less to the fact that its particles cannot move 

 relative to one another than to the fact that they can form regular 

 structures, crystal lattices, which is possible only if the relative 

 position of particles is fixed and regularities are not destroyed by 

 heat agitation. Glass, which is rigid and does not flow at room 

 temperature, has no regularity in its structure and so is looked 

 upon by the physicist more as a fluid of high viscosity than a solid. 

 The same is true for glycerol frozen to a rigid "glass" at low tem- 

 perature, while ice is a real solid with a regular hexagonal crystal 

 structure though it readily changes its shape and "flows," as evi- 

 denced by the motion of glaciers and as Helmholz showed. The 

 strong dipole character of the water molecules may endow their 

 crystals with specific qualities and so the most reasonable assump- 

 tion is that the transition of our excited molecules into the "for- 

 bidden" triplet state comes about by an interaction of these mole- 



