Fune 25, 1885] 
Maals River, and supposing that this was also the case 
during the Glacial age the ice-stream must have moved 
up an incline before it could reach the depression leading 
down to the Balsfjord. This cannot, however, have 
been the case. As long as the ice-stream had perfect 
liberty to travel down an incline—here present in the 
shape of the broad Maals River, along the southern slope 
of the Mauken—it would hardly ever move in the oppo- 
site direction wf an incline, leaving, however, local 
accumulations out of consideration. It might therefore 
be reasonable to suppose that the configuration of the 
land along the Divi Valley, and especially the Overbygd, 
was very different during the Glacial age. A continuous, 
though slightly inclining, surface must under these circum- 
stances at that period have extended from the alpine 
plateaux above the Divi Valley to the depression along 
the Tag Lake, and the present configuration be caused 
by subsequent erosion. It should be stated that the out- 
let of this lake does not now follow the course of the ice- 
stream towards the Balsfjord—which might have been 
reasonably assumed—but is at the opposite, eastern, 
end towards the Maals River. This seems to indicate 
that the present declivity of the Overbygd in an easterly 
direction in any case cannot be older than the close of 
the Glacial age. 
As stated, travelled granite blocks from the Divi Valley 
are found in great numbers along the northern slope of 
the Mauken, towards the Tag Lake, upwards of 2500 
feet (784 m.); but that these should have been raised 
from lower levels to their present height seems 
improbable. The northern slope of this mountain 
does not lie transversely to the course of the ice-stream, 
but longitudinally to it. Of course the screwing-up of 
the ice may also take place in the latter case, but I should 
say only in isolated spots; this cannot have been the 
case along the Mauken. Neither is it possible that the 
bottom of the lake lay at that level in the Glacial age. 
It must then have lain lower than the alpine plateaux by 
the frontier, and even if we allow for enormous glacial 
erosions, it would be impossible to believe that the 
bottom then lay at such a height. As the blocks on the 
Mauken cannot thus have been deposited along the 
bottom of the ice-stream, nor brought thither through 
screwing-up of the ice, we must assume that they have 
been deposited from the surface of the ice-stream. The 
latter being strewn with blocks, which at the frontier was 
above 3000 feet (941 m.) high, has therefore, at 40 or 50 
miles therefrom, had a height of 2500 feet. The surface 
can, therefore, under this long journey, only have had a 
very small declivity outwards. 
From the western end of the Tag Lake the great ice- 
stream has moved forward to the Sag Valley, which, being 
then as it is at present, has been able to receive it and 
turn it in a north-westerly direction downwards to the 
Balsfjord. That the Sag Valley cannot be of glacial 
origin, produced by erosion, is clear from the very nearly 
acute angle it forms with the Tag Lake depression. It 
might also be assumed that the ice-stream here might 
have moved forward across the Slet Mountain and the 
long, narrow peninsula between the Malangen and Bals- 
fjord, but that this was not the case is proved clearly by 
the circumstance that travelled granite blocks are found 
on this peninsula, or only at low levels, which I shall 
presently explain. 
It may be probable that the ice-stream from the Tag 
Lake has met another descending from the Maartinder 
in the Sag Valley, but there is no middle moraine proving 
this. On the other hand, travelled granite blocks are 
but sparsely strewn along the north-western side of the 
Sag Valley, at the foot of the Slet Mountain. Should 
the Sag Valley, therefore, be of glacial origin, it might 
more naturally be attributed to the ice-stream from the 
Maartinder, but even then eroded before the great inland 
ce-stream entered it. If, however, this was the case, the 
NATURE 
179 
former ice-stream must have been in motion long before 
the latter, of which there is no probability. 
We therefore come to the conclusion ‘hat the basin of 
the Balsfjord, viz., the Tag Lake depression and the Sag 
Valley, cannot be the result of the erosive action of the 
inland ice, but that tt existed prior to the Glactal age,and 
that, in fact, the depression in question was the cause of 
the ice-stream taking this course. 
We will now follow the depression through the fjord 
and adjacent sounds. 
As soon as we leave the true bottom of the fjord the 
travelled blocks are differently situated to those inland. 
There are plenty of granite blocks to be found, but ¢hey 
aré everywhere confined to lower levels, viz., from the 
shore-line up to 120 feet (38m.). Above, there ts none, 
and the line of disappearance ts very marked, My re- 
searches have extended, on the eastern side of the fjord, 
from the bottom to the sea; on the western side, 
though they do not extend so far, they go to show that 
the conditions there are identical with those on the 
eastern side. It is particularly significant that neither 
here are the blocks found above a height of 120 feet along 
the low, transverse ridge which runs from the Balsfjord 
on one side westwards to the Malangenfjord, and on the 
other, eastwards to the Lyngen and Ulfs fjords. Thus, the 
outer Malang isthmus, which, rising slowly to a height 
of 400 feet (125 m.), leads from the Bals to the Malang 
fjords, is along the former strewn with blocks, but only at 
lower levels. Above 120 feet they disappear. From this 
also it is clear that the inland ice cannot have moved 
forward across the Slet Mountain and the isthmus be- 
tween the Bals and Malangen fjords, previously referred 
to. From the bottom of the Nordkjos, a short bye fjord 
of the Balsfjord, running eastwards, the Balsfjord isthmus, 
two miles long, with a height of 250 feet (78 m.), leads to 
the bottom of the Storfjord in Lyngen. Here, too, the 
blocks are confined solely to lower levels towards the 
Nord and Balfs fjords. The blocks have not reached as far 
as across the isthmus to the Storfjord. 
The blocks may in the same manner be followed along 
the Ramfjord, which as a bye fjord runs from the mouth 
of the Balsfjord eastward to the Bredvik Isthmus. From 
the southern side of the mouth of the Ramfjord the 
Anders Valley runs in a southerly direction between lofty 
mountains and with a steady incline. Here, too, travelled 
granite blocks are found to a height of 120 feet, du¢ not a 
single one above. The case is the same along the sounds 
around the town of Tromsé. Further, I have followed 
the blocks northwards, on the mainland to Tunnes, about 
five miles from the town, but whether they have travelled 
further along the Grét Sound I have not yet been able to 
ascertain. The same applies to the Kval Sound. But 
researches made on the islands ow¢s¢de this sound prove 
beyond a doubt that the granite blocks from the Balsfjord 
cannot have reached these islands by way of the Kval 
Sound. 
The greatest number of travelled blocks along the 
Balsfjord belong, judged petrographically, to the Divi 
Valley granite, blocks which might with certainty be 
referred to the coast granite not having been found. 
Along the sounds, too, the greatest number of blocks, if 
not all, may be referred to the Divi Valley granite ; but 
blocks belonging to the gray, streaky gneiss-granite of 
the Kval Island are also met with here, some of which 
may even be referred to exact localities in the island. 
Among the rocks along the Troms Island and adjacent 
sounds blocks of a coarse-grained syenite are also often 
found. In the Divi Valley no varieties of syenite appear, 
but they are often encountered combined with gneiss and 
gneiss-granite on the coast. Although I have not yet 
succeeded in finding syenite in place which with certainty 
can be said to be petrographically identical with that of 
these travelled blocks, I have every reason to believe that 
they haii from the west. 
