15 
and the effect of the short season upon the development of soils and topo- 
graphy. These have already been brought out in part by E. M. Kindle 
(33, 34) and by the present writer (53). 
The predominating vegetation which gives the distinguishing character 
to the landscapes of the region is the coniferous evergreen forest. The 
determining factor in its distribution appears to be the physiological dry- 
ness of the soils, which is brought about in this case by low temperatures 
over much of the year, resulting in the rendering of existing water unavail- 
able to plants, and in the production of a high degree of acidity (70). 
The period between the final thawing of the soils in spring and the 
time of their freezing in the autumn is not recorded, nor have any data been 
compiled as to an average frostless season, if the latter can actually be said 
to exist. Absolute minimum temperatures during July commonly reach 
32 degrees 1 or fall below it, and killing frosts are likely to occur at any 
time during the summers 2 . Such low July minima were recorded at 
Chipewyan nine times during the forty-five years between 1884 and 1928, 
five times at Vermilion between 1905 and 1928, six times at Fort Smith be- 
tween 1913 and 1928, only once at Hay river between 1893 and 1928, and 
not at all at Resolution between 1914 and 1928. The first dates just given 
are the earliest from which there arc consecutive records. Low tempera- 
tures are much more common in June and August. In forty-five years' 
records at Chipewyan they are found twenty-eight times in June and 
twenty-four times in August. At Resolution, where July minima are rela- 
tively high, the June minima fell to 32 degrees or below thirteen times in 
the fifteen years recorded, but only twice in August. The sudden change 
from June to July is probably due to the fact that the Great, Slave Lake ice 
holds through June. From these few data it is evident that the season that 
is quite free of frost is very short and exceedingly hazardous. It centres 
in July and is probably more certain on the south shore of Great Slave lake 
than in other localities under consideration. 
A seasonal indicator that has been recorded for many years throughout 
the north is in the times of opening and closing of the lakes and rivers. 
The lower reaches of the Peace and Athabaska w r here they approach Atha- 
baska Lake area usually open about the first of May (50) , but the western 
part of the lake is not free of ice until about May 10. The lake ice usually 
sets again during November. The large degree of variation in these times 
is shown by the records of E. M. Kindle (35). In the autumn of 1919, he 
found ice set in Quatre Fourches channel on October 9, and much floating 
ice in Rocher river. Protected bays in Athabaska lake had an inch of ice 
on them. His party, ascending the Athabaska, found it necessary, at 
McKay, on October 13, to abandon a scow that had been nearly cut through 
by floating ice. The usual time for the freeze-up of the Slave, according 
to his notes, is in the latter part of October. The ground was frozen, in 
1919, on October 4. Great Slave lake freezes up between the latter part 
of October arid the middle of November. 
The six to six and a half months of open water in Athabaska Lake 
area is undoubtedly a longer period than that in which the upper layers of 
1 Temperatures given in this paper are Fahrenheit. 
2 Detailed meteorological data given here and in the tables below have been supplied to the writer 
through the courtesy of the Director of the Meteorological Service, Toronto, Ontario. 
91963 — 2 * 
