SNOW 135 ous authors, and several interesting observa- tions are given upon the conditions of their formation. The accompanying figures, repre- senting specimens of the simple and the more complicated forms of crystals, are from Bu- chan's "Meteorology." Scoresby, who first studied these forms, classified them into : 1, thin plates (figs. 1 to 7) ; 2, spherical nuclei studded with needles (fig. 8) ; 3, three- or six- sided prisms or needles (fig. 9) ; 4, six-sided pyramids (fig. 10); 5, prisms terminated by planes (fig. 11). The conditions regulating the occurrence of each figure are probably quite definite, inasmuch as it is rare that more than three or four kinds of flakes occur at the same time. The high cirrus clouds are prob- ably generally formed of spiculse, or possibly small flakes of snow, which when the clouds are not too thick give rise to the phenomena of halos (see HALO), and the geometrical ex- planation of these latter seems generally to require that the snow flakes present in these clouds should be principally of the simplest forms. The amount of snowfall in different parts of the earth is known with less accuracy than is that of rainfall, owing partly to the drifting of the snow, but especially to the fact that a too great diversity has existed in the methods adopted by the observers to ascertain either the quantity or the depth of the snow. It is generally assumed that -fa or -^ of the depth of snow measured immediately after falling will give the corresponding depth of lelted snow. Quetelet, as the mean of many observations, says |, but for very dry or very wet snow these fractions are very uncertain. The total depth of snowfall is greatest, other conditions being the same, where the strong winds of winter are laden with moisture ; thus it averages annually 4 to 7 ft. in the interior of Maine, Vermont, New York, and Upper Can- ada, but only 2 ft. for the states in the same latitude further west. One of the heaviest snowfalls recorded in America was that which continued from Feb. 19 to 24, 1717, when the snow remained 5 or 6 ft. deep over all the set- tled parts of New England. The geographical distribution of snow at sea level is such that in general in the eastern parts of North America and Asia it is rarely seen S. of lat. 30, and in western Asia S. of lat. 36. On the W. side of North America it is rarely seen at the sea level on the immediate coast, but is quite com- mon in the interior. Falls of snow may occur in any month in extreme polar latitudes ; in New England and Canada snow falls mostly from November to March inclusive, but in the latitude of Washington, D. C., it falls mostly during January and February. The average number of days on which snow falls is, for St. Petersburg, 170; Paris, 12; Washington, D. C., 20 ; Gibraltar, ; San Francisco, ; Charleston, S. C., 1. But on ascending above the sea level we soon come to altitudes such that snow may fall and remain on the ground at any season ; the altitude at which accumulations remain throughout the year is called the limit of per- petual snow. The conditions governing this lower limit were first studied carefully by Hum- boldt in his climatology of Asia, and more re- cently has been investigated by Grad (1873) ; according to these, the limit in question has a general apparent connection with the iso- therms of 32 F., but departs therefrom to an important degree when the prevailing winds are dry or moist. Thus the limit is lower in the southern than in the northern hemisphere ; lower on the S. than on the N. side of the Himalaya mountains ; lower within the tropics than under the latitudes 20 to 35. From these latitudes it diminishes, according to Grad, to about 3,000 ft. in lat. 60 S. and 65 N. ; but only in the high polar regions is the limit below 1,000 ft., it being higher in Greenland or Spitsbergen, where it is only the glaciers that descend to sea level. (See GLACIER.) Owing to the innumerable reflect- ing facets of the minute crystals and the quan- tity of air caught between the crystals, a layer of snow is a remarkably perfect non-conductor of heat ; for this reason the covering of snow on the ground forms an almost perfect protec- tion to the plants beneath against the freezing that would otherwise follow the radiation of their heat into the atmosphere. In Ebermayer's "Influence of Forests" (1873) a case is quoted (by no means an extreme one) in which the temperature of the air was 6-8 F., and that of the surface of the earth beneath the snow + 33-8 F., while below the surface the earth was still warmer. On the other hand, the in- dividual crystals of snow have probably the same large radiating power as ice in larger solid blocks, which according to Leslie is 85, that of lampblack being 100. The consequence of this is, that during the night very hard frozen crusts are formed on the surface of the snow which has been somewhat thawed during the day; the same property, together with that of regelation, explains the peculiar struc- ture of the surface snows of glaciers, and assists in the formation of areas of colder air over snow fields than over bare land. Equally important is the great absorptive power of snow for solar heat, since by reason of it the surface of a layer of snow is melted rapidly, and a large amount of moisture is thrown into the air, giving rise to extensive fog and haze, and having a decided influence on the development of storms. Snow flakes in falling bring with them nearly all the fine dust float- ing in the air, leaving the atmosphere extreme- ly pure; thus in northern Europe Nordens- kiold has found freshly fallen snow impreg- nated with a black dust of carbon and iron such as could only have come from meteors ; at other times the dust is such as could only have come from eruptions of volcanoes, espe- cially those in Iceland. Snow is occasionally tinged black, yellow, red, or green, as was known to Pliny. These colors are due to the presence of microscopic organisms, as was
