ELECTRON IMICKOSCOPY 



Fig. 2. A center nucleus and the etched surface 

 of a snow crystal. 



minute crystals of cubic nature are observed 

 outside the rounded pattern. 



Replicas of Ice Crystals. Silver iodide par- 

 ticles become active as nuclei of ice crystals. 

 The particles are produced in the laboratory 

 by evaporating silver iodide on an electric 

 hot wire. The particles are then introduced 

 into an undercooled cloud at a temperature 

 below -4°C (17, 20). Small ice crystals are 

 formed soon after the seeding. The meshes 

 of the specimen are covered with collodion 

 film and are arranged on a clean glass slide. 

 Just before the replicas are made, each mesh 

 is covered with a droplet of "Formvar" solu- 

 tion. The meshes are exposed to the falling 



crystals to collect samples (G, 11), or the ice 

 crystals are collected on a mesh by the use 

 of an impactor. Then the specimens are kept 

 in the cold box for several hours while the 

 solvent and the crystals evaporate. The so- 

 hition for preparation used is about 0.01 per 

 cent "Formvar" in ethylene dichloride. 



In general, a solid particle can be seen at 

 the center of the crystal I'eplica by the use 

 of an electron microscope. The result of the 

 particle examined by the electron diffraction 

 method shows the Debye-Scherrer rings due 

 to Agl and Ag^. Silver is produced by de- 

 composition of silver iodide as the result of 

 electron bombardment. 



Center Nucleus and Condensation 

 Nuclei of a Snow Crystal. It was found 

 that a relatively large, solid nucleus (4, 7, 8, 

 15) always exists at the central portion of 

 the crystal, with innumerable minute nu- 

 clei observed in the remainder of the crystal. 

 The former is called "center nucleus" and 

 the latter are called "condensation nuclei" 

 in this article. The material of the center 

 nuclei was estimated from the shape by com- 

 paring it with the electron micrographs of 

 known substances or by observing specimen 

 changes (3, 21) due to electron bombardment 

 in the electron microscope. Better estimation 

 was obtained from electron diffraction pat- 

 terns. 



Fig. 3a shows a dendritic snow crystal. 

 After sublimation, a solid particle remained 

 at the position where the crystal center 

 vanished by sublimation, as shown in Fig. 

 3b. Many small particles were found around 





n 



^ 



bor 



\ r 



Fig. 3. Dendritic snow crystal and the nucleus. 



256 



