August I, 1S89] 



NATURE 



15 



and local variations become ; and since also, as a rule, 

 they are there more lasting than in our changeable 

 climate, and take effect simultaneously over vast areas of 

 country, it becomes an important object of inquiry how 

 and to what extent they react on the atmosphere, and 

 indirectly affect not only the local weather, but even that 

 of surrounding regions. In the case of a sheet of snow, 

 attention was first drawn to this question by Prof. 

 Woeikof in 1871, in a paper published in the Transac- 

 tions of the Russian Geographical Society, in which he 

 endeavoured to show that as regards certain portions of 

 Russia, the spring temperature depends very much on 

 the quantity of snow that has fallen in the previous 

 winter, a snowy winter being followed by a cold spring, 

 and vice versd. In 1878 he adduced further evidence of 

 the effects of snow on the temperature, in a notice of the 

 weather of December 1877, in the Zeitschrift fiir Meteoro- 

 logies and in 1880 Prof. Hann and Dr. Assmann, in the 

 same periodical, traced the remarkably low temperature 

 of Central Europe and the persistent anticyclone which 

 ' lay over that region in December 1879 to the effect of the 

 heavy snow that had fallen in the beginning of the month. 

 In Northern Europe, where snow had not fallen, the 

 temperature was not merely relatively, but absolutely 

 higher, and it lay beyond the limits of the anticyclone. 



Meanwhile, in 1877-79, the Indian meteorologists, 

 being unacquainted with Prof. Woeikof's earlier work, 

 had independently arrived at the conclusion that varia- 

 tions in the quantity of the winter and spring snowfall 

 of the Himalaya exercise an important influence on the 

 monsoon rains of the Upper Provinces of India ; the con- 

 sequences of an unusually heavy and especially late 

 snowfall on the mountains being the persistence of dry 

 westerly winds on the plains to the partial exclusion or 

 enfeeblement of the rain-bearing summer monsoon ; and 

 it was shown that this sequence of land-winds on the 

 plains after snow on the mountains is a common recurrent 

 feature of the winter and spring months. In consequence 

 for some years past, monthly reports on the state of the 

 Himalayan snows in the first six months of the year have 

 formed a part of the regular routine of the Indian Meteoro- 

 logical Service, and they have afforded important data 

 for forecasting the character of the monsoon rains. 



In 1886 a system of snow reporting was adopted 

 in Bavaria by Dr. Lang, and as the result of Prof. 

 Woeikof's labours, a similar system was established in 

 Russia two years later. In India, the system was estab- 

 lished under a general order of the Government in April 

 1883, and that country, therefore, and not Bavaria, as 

 stated by Prof. Woeikof, has considerable priority in this 

 matter. 



In his recent work, referred to at the head of this 

 article, Prof. Woeikof sums up the present state of our 

 knowledge of the influence of a snow-sheet on the soil, 

 cHmate, and weather. The protection of the ground 

 against frost, which is afforded by a covering of thick, and 

 loose snow has long been familiar, and was, indeed, the 

 subject of detailed observation by E. and H. Becquerell 

 in 1879-80 ; but additional observations of much value 

 are given in evidence on this head, especially an elaborate 

 series carried on during two years at the Russian Polar 

 station Segastyr, in the Lena delta, in N. lat. 73° 23'. At 

 this place, when the ground is covered with snow during 



more than tix months of the year, the temperature of the 

 coldest month at r6 metre beneath the ground surface 

 was found to be less than 1° C. below the annual mea 

 temperature of the surface (alternatively snow or soil) 

 exposed to the air, while that of the warmest month was 

 I5°'4 C. above it. The greater part of this difference of 

 the excess and deficiency must be attributed to the non- 

 conducting layer of snow and the protection thereby 

 afforded against radiation and contact with the cold winds. 



Of the differential cooling effect of a snow-sheet on the 

 atmosphere, as compared with that of a bare land surface, 

 the systematic evidence hitherto available for a rigorous 

 comparison is less extensive than could be desired, owing 

 to the fact that, until recently, but very few meteorological 

 observatories have recorded the presence or absence of 

 snow on the ground. But it fortunately happens that 

 one first-class Observatory, that of Upsala, has done so 

 for a period of fourteen years, and the discussion of these 

 observations forms one of the most interesting and im- 

 portant chapters in the book. Comparing month by 

 month the mean temperatures of all periods during which 

 the ground was under snow with those with an unsnowed 

 surface. Prof. Woeikof finds that the former are lower in 

 November by 4°7, in December by 5^"i, in January by 6"^, 

 in February by 5'''i,and in March by 5'"2 C, respectively 

 equal to 8-*5, 9°*4, lo^'S, 9''*2, and 9"'4 F. The effect of a 

 snow-sheet in lowering the temperature of the air, and in 

 helping to establish anticyclonic conditions, such as pre- 

 vailed over Central Europe in December 1879, appears, 

 therefore, to be very considerable. 



In another chapter of the work. Prof. Woeikof applies 

 the conclusion thus established to the explanation of cer- 

 tain anomalies in the winter temperatures of parts of Asia 

 and North America, and shows that the lower temperatures 

 coincide with the prevalence of snow, and vice versa. 

 The most striking instance given is that of the Armenian 

 plateau, the mean winter temperature of which, after re- 

 duction to sea-level value, would appear to be from 4^ to 

 7° C. lower than that of the lower parts of the Trans- 

 caucasian province to the north and east. On the former, 

 the snow lies for four or five months of the year ; on the 

 latter there is but little snow, and the mean temperature, 

 even of January, is from 2° to 4^ C. above the freezing- 

 point. This exceptional area is well shown on Hann's 

 isothermal chart for the month of January, in the new 

 edition of Berghaus's " Physical Atlas." 



In discussing the effect of a thick winter snow-sheet 

 on springs and rivers, a variation is pointed out, which is 

 not without importance in its bearings on some points of 

 physical geography. In latitudes where the winter cold 

 is sufficient to freeze the ground to a considerable depth, 

 if heavy snow falls early in the winter before the cold has 

 penetrated deeply below the surface, the protection there- 

 by afforded allows the ground to thaw by conduction from 

 the lower strata, and the water from the slow melting of 

 the basal snow-layer, and much of that which is produced 

 in the spring thaw, soaks into the soil and affords a sup- 

 ply which maintains the rivers more or less full through the 

 succeeding summer. But if, before snow falls, the soil 

 has been frozen to a great depth, a rapid thaw setting in 

 in the spring floods the rivers and the surrounding tracts, 

 while little or none enters the ground, and but little 

 supply is stored up for maintaining the summer flow. 



