Since 1862, Schillarization of moonstone has been studied extensively using various methods, such as optical and electron microscopes. The theories on the origin of Schillarization may be classified into two; diffraction and dispersion theories. The diffraction theory was first put forward in 1921 by the late Prof. S. Kozu of Tohoku Imperial University, who studied both adularia and moonstone by X-ray and found that moonstone was crypto-perthite and adularia potash feldspar, denying the long-holding misunderstandings that moonstone was a variety of adularia, showing Schillar. Therefore, it is suggested that the term adularescence which has been carelessly used among gemmologists should be abandoned. Fig.10 A shows the Laue pattern of moonstone from Shrilanka which consists of double spots due to albite and orthoclase, definitely showing that moonstone is crypto-perthite. When this sample is heated up to 1088℃, the Laue pattern changes to a single crystal pattern as shown in Fig.10 B, exsolution pattern and Schillar effect disappear simultaneously. Kozu therefore believed that Schillarization of. moonstone is due to the optical diffraction from crypto-perthite lamellae. The lamellae have been also observed under the electron microscope. The dispersion theory was put forward by Spencer (1930) and Raman (1950), though the latter used a term "diffusion", instead of dispersion. Both did not present enough evidence to support their theory. Now, let's compare the Schillar and texture of moonstone with those of labradorite. Labradorite consists of regular lamellae whose thickness changes with the composition, and color, irridescence, changes with the thickness. On the contrary, moonstone consists of fine alternating lamellae of albite and orthoclase whose thickness varies from lamellae to lamellae, depending not on the compositions. Albite lamellae show fine albite twinning. The colour of moonstone from most localities is blue or silky white. It is clear that the optics should be different between moonstone and labradorite.Fig.15 shows a sketch of a simple experiment applied to moonstone from Shrilanka. A narrow light beam transmitting through the crystal normal to the (001)(arrow) is reflected by the perthite lamellae parallel to the (8^^-01), resulting in an elongated band of light in one direction. When viewed from the upper side, blue(青), white(白) and red(赤) colours is seen on the cleavage surface as apart from the point of emergence of the light beam. Similar experiment was carried out using a cylindrical screen, a specimen being positioned at the centre of the cylinder. Fig.16 shows intensity distribution of light on the screen. A peak is observed at 60°,which correaponds to the odd times reflection from (8^^-01) perthite lamellae, whereas a broad peak appearing near 0°is due to the even times reflection from the lamellae. The colour on the screen changes with the thickness of specimens. It is bluish when the apecimen is thinner than about 1mm, and is silky white for the crystal of 1 to 5mm thick, and reddish for thicker specimen than about 5mm. Since the phenomenon is very similar to the scattering of sunlight in the sky, it is conjectured that the origin of Schillarization of moonstone is dispersion of light due to the lamellar texture. When dispersion is weak, the colour is blue, and when it is strong, it is reddish. If dispersion is medium, the colour is silky white. Figure 17 shows an electron micrograph of a thin foil parallel to the (001) of moonstone from Shrilanka. Albite lamellae with fine polysynthetic twinning and orthoclase lamellae have very wavy and irregular interfaces and varying thickness. The orientation of reflection and refraction will slightly deviates from those expected for the reflection of real (8^^-01) lamellae. Since the reflection and refraction are repeated many times in the moonstone, the light will become to show the nature of

(View PDF for the rest of the abstract.)