| OCR Text |
Show 332 C. MAGONO and C. W. LEE the sampling point, it is desirable to distinguish them from usual mature snow crystals which fall from above. In the present classification, one group was added as the " germ of snow crystals" for the snow crystals in the early stage, and this group was divided further into six fine classes, namely minute columns: Gl, germs of skelton form: G2, minute hexagonal plates: G3, minute stellar crystals: G4, minute assemblage of plates: G5, and irregular germs: G6. The germ of snow crystals are assumed to be the next stage of ice crystals. 4. Conditions for the formation of various kinds of natural snow crystals The first purpose of the observation of the Cloud Physics Group, Hokkaido University was to examine whether Nakaya's Ta- s diagram was exactly applicable to the formation of natural snow crystals or not. The examination was made by comparing the crystal shapes with the meteorological condition of the mother cloud in which the snow crystals were formed. Because snow crystals fall from above, an exact comparison required both an observation of the shape of the ends of the branches of snow crystals and a measurement of meteorological conditions of the air parcel in which the snow crystals were sampled. The latter measurement was carried out by the use of five surface observation points distributed vertically at Mt. Teine. The result by this method was reported by one of the authors19'. However by such methods, only data at fairly lower levels, namely lower than 1,000 m height were obtained. In order to obtain data at higher levels, the data of usual radio sonde soundings which were made by Sapporo Meteorological Observatory were used only when the existence of a cloud layer was confirmed by other methods. Furthermore the results of Weickmann's observation20) of ice crystals in cirrus clouds were used. In the later observation of this work, the snow crystal sondes were used. By the methods described above, the air temperature of a mother cloud in which the snow crystals were formed as determined fairly exactly; however the determination of humidity was difficult. As well known it is impossible to measure the humidity of supersaturated air by the use of a usual hygrometer or a psychrometer, and it is also difficult to obtain the exact value of humidity in cold temperature. However the meters measuring the humidity were reliable as an indicator to determine whether the air was saturated or not with respect to an ice surface. Therefore it was possible to determine the |
