DEVELOPMENT OF LIGHT AND HEAT. 499 



tricklings therefrom moisten tlie earth below, the Soldanella plants are aroused 

 from their winter's rest. Their little arched flower-stalks begin to elongate and 

 come into contact with the hard under surface of the snow, though the temperature 

 here is zero. Growth is carried on at the expense of the supplies of mateiials 

 obtained by the Soldanellas in the previous summer, which had been stored up 

 partly in the evergreen leatheiy leaves lying flat on the ground, and partly in the 

 short root-stock embedded in the soil. The reserves are employed as substances for 

 building, and a portion of tliem is respired, in order that it may bo possible to dis- 

 solve the rest, to bring them to the places where they are required, and to obtain 

 the force necessary for the work. The heat liberated by this respiration melts the 

 gi-anular ice covei-ing in the immediate neighbourhood of the flower-buds. In con- 

 sequence of this a cavity is formed in the ice above each bud, or rather, each bud 

 becomes overarched as if by a tiny dome of ice. But the stem continues to gi-ow in 

 height; and the flower-bud borne on it, which is respiring and giving out heat, is 

 accordingly raised up in the dome-shaped hollow space and pushed into it. There 

 it promotes afresh the melting of the ice and an extension of the cavity, and thus 

 actually boi-es a path for itself upwards through tlie ice-strata. This goes on until at 

 length the respiring and heat-producing Soldanella bud has melted an actual canal 

 through the covering of hardened snow, and makes its appearance above, the stem 

 looking as if it had been stuck into the snow. The flower-buds now open and the 

 pretty violet bells sway about in the wind. Naturally the snow will be penetrated 

 first where it is thinnest, i.e. near the margin, where also the melting from above 

 proceeds most rapidly. Consequently it is the edge of the snow-field mainly which 

 is riddled, the Soldanellas growing up through the holes. It is not at all uncommon 

 to find places where 10-20 flowers appear on the border within a stretch a metre 

 long. On looking closer and making cuttings through the ice, all the stages of de- 

 velopment described may be seen side by side. Two other phenomena, however, are 

 not a little surprising. Here and there are to be found single Soldanellas whose 

 buds have already opened before they have emerged above the ice-covering. These 

 Soldanellas actually blossom in a small cavity of the ice, and remind one of plant- 

 organs or insects inclosed in amber or small coloured splinters which have been 

 fused inside glass balls. Tliis sub-glacial blossoming of the Soldancllas is not 

 limited, strangely enough, to the opening of the corolla; the anthers actually 

 dehisce and liberate their pollen. 



What also surprises us very much on closer inspection is the fact that the holes 

 in which the flower-stalks are situated narrow like a funnel towards the base, so 

 that there the ice touches the stem, or, in other words, that the canal down below is 

 completely filled by the stem. When it is remembered that the flower-bud which 

 melted the ice and formed the canal had a diameter at least three times as large as 

 that of the stem, it would be expected that the stem would be placed in the centre of 

 a comparatively wide hole. But, as stated, this is not the case, and the phenomenon 

 can only be explained by supposing that the porous gi-anular ice forms a plastic 

 mass, and that the granules displaced by the melting sink down in accordance 



