2 Mr. Talbot's Facts relating to Optical Science. No. III. 



orange-red crystals lying interspersed among yellow ones, 

 which resembled them in size and shape, and were different 

 in no respect but in colour. 



Upon repeating this experiment I found that by continuing 

 the heat a little longer I could obtain much larger crystals, 

 such that the field of viev/ of the microscope would only con- 

 tain a few of them at a time ; and with these it became possi- 

 ble to see the phaenomenon much more distinctly and advan- 

 tao-eously. These large crystals have the shape of thin flat 

 lozenges or oblique-angled parallelograms of a pale yellow 

 colour. They are very transparent, act strongly upon light, 

 and form a very pleasing and convenient object for the po- 

 larizing microscope. But the most important and singular 

 phaenomenon which they exhibit, is the sudden change of 

 colour which they are capable of undergoing, and to which 

 I do not think the science of optics has hitherto furnished any 

 parallel. The change takes place in some of the crystals du- 

 ring the process of cooling, in others shortly afterwards, while 

 in others the yellow tint remains permanent for many hours, 

 or even days. In general, a crystal which is about to change 

 colour, is known by the appearance of a red streak along one 

 of its sides or edges. 



If then the observer selects one of these and fixes his atten- 

 tion upon it, he will shortly afterwards see it change colour 

 from pale yellow to a fine and deep orange-red. The change 

 generally occupies only a few seconds, and the red tint ad- 

 vances uniformly across the crystal, i. e. the boundary of the 

 red and yellow is a straight line parallel to two sides of the 

 rhomboid, and its motion is across the ci-ystal from one of 

 these sides to the opposite one. 



The change of colour is accompanied by a visible internal 

 motion in the crystal, like a sinking or giving way of succes- 

 sive ranks of particles, one consequence of which is, that the 

 crystal after the change is generally less transparent than it 

 v/as before. This phaenomenon is, I think, the most evident 

 proof which we yet possess of the dependency of colour upon 

 internal molecular arrangement. 



As this substance sublimes very readily, I tried what might 

 be the effect of placing a hot piece of glass over the crystals 

 while they were under examination with the microscope. Im- 

 mediately each crystal became surrounded with a cloud of 

 little particles which appeared to me to be rhomboids like the 

 larger ones from which they were derived. 



The cracking of the larger crystals from heat must be an 

 effect of unequal expansion, and it appears not unlikely from 

 this experiment that the sublimation of the substance is pro- 



