On larger glaciers snow water percolates deep into snow, forming water envelopes of snow, firn, and ice grains, and accelerates the formation of firn and glacier ice (Paterson, 1969). Such transformation of snow into glacier ice, under the greater pressure of the upper masses, takes several decades or hundreds of years (Wilchelm, 1975, p.137). The Triglav glacier is thin, therefore there are no required pressures of the upper masses and the glacier mass regenerates relatively rapidly by melting snow.
Fig. 3 shows the contact of ice surface with snowfield which is soaked with water flowing over the ice of the lower icefield on September 24,1994. In cold nights, such water-soaked snow freezes. The contact is marked with darker humus dust particles which water keeps depositing on the snow. Therefore, the stratified ice of the base is also darker.
On the uncovered darker ice whose albedo intensely lags behind the albedo of fresh, white snow, ablation absolutely prevails in summer. Water accumulated to form streams cuts deep gullies (half a meter or even more) into the ice, exceptionally also holes, the latter most probably in the crack where it reaches the rocky base.
The experiences gained by surveying the glacier in melting season when flowing water was gurgling under the rubble cover, have been supported by collapses as they were seen in September 1994. When the roche moutonee with the 2472 m point poked up through the ice, a more permanent stream began to flow onto it from the ice under the rubble, which was used as water supply to the neighboring mountain hut for several years. The collapse which occurred lower down on the scree and closer to the slope of Kredarica allowed the access to the rock over which the water was flowing with an estimated discharge of 0.1 l/sec. It undoubtedly originated from the ice under the rubble in higher locations. Namely, the rubble on the surface receives more solar energy because it is of darker colour than the snow. Cool spells in May and June, partly also later, intensely cooled this rubble well under 00C. When later on rainwater percolates through it, it freezes. Therefore, the ice under screes is not necessarily the fossil glacier ice.
As to the changes of the glacier mass under Triglav three periods can be discerned. In the second half of the past century the glacier most probably consisted of firn or glacier ice, but there are no data on the existence of darker ice. In the post-war period 1946-88, it consisted above all of darker ice in the base covered with snow; in accordance with the climate, the former occurred as an island in this snowfield, or, in "drier" years, it was some- times completely disclosed on the entire surface. In the nineties, also this ice base of the glacier disintegrated.
