Abstract
The temperature conditions for the growth of snow crystals which were observed at Mt. Teine were compared with those which were determined by Nakaya, Hallett and Mason, and Kobayashi. The agreement was much better than with the findings announced by Gold and Power, and Murai. This improvement is probably because of the fact that in the Mt. Teine observations air temperature was not supposed from the data obtained by means of radio sonde sounding, but measured just in the cloud layer in which the snow crystals grew. Thus it is proposed that snow crystals of needle, sheath and dendritic type are sufficiently useful as indicators of cloud temperature during snowfall.
Snow crystals of spatial type and snow flakes composed from far different types were also discussed from the point of view of cloud physics.
