226 A CONTINUOUS RECORD OF ATMOSPHERIC NUCLEATION.

its origin is in part to be traced to a more deep-seated cause. At least
the results obtained are sufficient to warrant the extension of similar ex-
periments over a wide area of territory, including regions remote from
the habitations of man — researches for which I am now making extensive
preparation.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE

PART OF VOLUME XXXIV

GLACIERS OF THE CANADIAN ROCKIES

AND SELKIRKS

(SMITHSONIAN EXPEDITION OF 1904)

BY

WILLIAM HITTELL SHERZEK, Ph.D.

(No. 1692)

CITY OF WASHINGTON

PUBLISHED BY THE SMITHSONIAN INSTITUTION

1907

Commission to whom this Memoir has been referred

THOMAS CHROWDER CHAMBERLIN
HARRY FIELDING REID
GEORGE PERKINS MERRILL

Unlcfterbocfcer prees, «ew Both

ADVERTISEMENT.

DOCTOR WILLIAM H. SHERZER, Professor of Natural Science at Michigan
State Normal College, has brought together in the present memoir the results
of an expedition undertaken by the Smithsonian Institution among the glaciers
of the Canadian Rockies and Selkirk's in the year 1904. The general objects of
the research were to render available a description of some of the most access-
ible glaciers upon the American continent, to investigate to what extent the
known glacial features of other portions of the world are reproduced in these
American representatives, and to ascertain what additional light a study of
similar features here might shed upon glacier formation and upon some of the
unsettled problems of Pleistocene geology.

A systematic survey was made of the Victoria and Wenkchemna glaciers in
Alberta and of the Yoho, Asulkan, and Illecillewaet glaciers in British Columbia,
located about two hundred miles north of the boundary of the United States.
The largest of these is the Yoho Glacier, extending more than three miles below
the ne"ve" field, and a mile in width for two-thirds of its length. Doctor Sherzer
investigated various surface features of each of these glaciers, the nature and
cause of ice flow, the temperature of the ice at various depths and its relation
to air temperature, the amount of surface melting and the possible transference
of material from the surface to the lower portion; their forward movement and
the recession and advance of their extremities, and the general structure of glacial
ice.

In summarizing the most important results Doctor Sherzer discusses the
indicated physiographic changes in the region during the Mesozoic and Pleis-
tocene periods; the question of precipitation of snow and rain, and the effect of
climatic cycles on glacial movements, the structure of the ice as to stratification,
shearing, blue bands, ice dykes, glacial granules, and the possible methods of
their development. In discussing the theories of glacial motion the author
expresses his conviction that the nature of the ice movement can be "satis-
factorily explained only upon the theory that under certain circumstances and
within certain limits ice is capable of behaving as a plastic body, that is, capable
of yielding continuously to stress, without rupture," but "the plasticity of ice,
a crystalline substance, must be thought of as essentially different from that
manifested by such amorphous substances as wax or asphaltum. "

Doctor Sherzer also discusses the cause of the richness and variety of coloring
of glaciers and glacial lakes.

In accordance with the rule of the Institution this paper has been referred
to a commission consisting of Professor T. C. Chamberlin, of the University
of Chicago; Professor Harry F. Reid, of Johns Hopkins University, and Doctor
George P. Merrill, of the United States National Museum, and upon their
favorable recommendation is published in the series of "Smithsonian Contribu-
tions to Knowledge."

RICHARD RATH BUN,

Acting Secretary.

SMITHSONIAN INSTITUTION,

WASHINGTON, D. C., January, 1907.

PREFACE.

THE five glaciers selected for investigation are located in Alberta and
British Columbia, along the line of the Canadian Pacific Railway. They repre-
sent the great snow-ice masses which accumulate, season after season, upon the
higher slopes and within the amphitheaters of the Selkirks and Canadian Rockies,
the slow downward movements of which prevent indefinite accumulation and
bring these great ice bodies to a level where complete melting may occur and
the waters again be put into circulation. The observations here described
were begun by the writer in the summer of 1902 and continued through the
seasons of 1903, 1904, and 1905; the entire field season of 1904 being devoted
to the surveys and more detailed studies.1 Camps were established in the
immediate vicinity of the glaciers selected and they were kept under almost
continuous observation during the hours of daylight. Beginning with the nose
of each glacier, surveys around either side to the neVe field were made with
plane-table, transit, or compass; the measurements being with a steel tape.
It was found impracticable and unnecessary to traverse the neVe" areas and
those portions mapped were drawn from field observations and original photo-
graphs together with maps and illustrations from the Canadian Topographic
Survey, and other sources. The writer was ably assisted by Mr. DeForrest
Ross and Mr. Frederick Larmour, to whom he desires to make grateful acknowl-
edgment for intelligent and faithful service, rendered often under trying cir-
cumstances. During the latter part of the season of 1905 very efficient assistance
was rendered by Messrs. E. W. Moseley and O. K. Todd.

Being the most accessible glaciers upon the American continent it was
desired to render available as complete a description as time and facilities
would permit and to ascertain to what extent the known glacial features of
other portions of the world are reproduced in these American representatives.
It was hoped that a study of the same features, produced under somewhat
different conditions, might shed additional light upon their method of formation
and upon some of the unsettled problems of Pleistocene geology. A dispro-
portionate amount of time was devoted to the Victoria Glacier, at the head of
the superbly beautiful Lake Louise Valley, since this glacier is geologically the
most interesting and may well be taken as a type by students of glaciology.
A delightful camp site lies under the lee of the outer massive block moraine and
a still more picturesque one farther up, on a low shoulder of Mt. Whyte, over-

'A preliminary report upon the expedition appeared in May, 1905, in the " Smithsonian Miscellaneous
Collections," vol. 47, Quarterly Issue, pp. 453 to 496.

VI PREFACE.

looking the small lakelets. Students who may carry this report into the field
generally desire an explanation which they can put to the test, upon the spot,
and so an attempt has been made at interpretation of the various phenomena
described. The value of such interpretation will be known only after others
have passed judgment upon the same features and more extended observations
are available.

Numerous forest fires in the season of 1904 prevented distant photography,
on account of the smoke or haze, but through the courtesy of the Dominion
Topographic Survey and of the Detroit Publishing Company we are permitted
to reproduce some of their general views, obtained under favorable conditions.
In addition to the views so used the writer is indebted to Captain Eduard Deville,
Surveyor General of Dominion Lands, and his Chief Topographer, Arthur O.
Wheeler, for a series of maps and photographs and much information concerning
the regions under study. To the Director, R. F. Stupart, of the Canadian
Meteorological Service, to the Assistant Director, B. C. Webber, and to Mr.
N. B. Sanson, very grateful acknowledgment is made for meteorological data
relating to British Columbia and Alberta and for the use of instruments kindly
placed at the disposal of the expedition. Very sincere thanks are hereby ten-
dered also to Prof. Joseph B. Davis, of the University of Michigan, and to Prof.
Elmer A. Lyman, of the Michigan State Normal College, for the use of surveying
instruments. The writer further desires to express his deep gratitude to the
officials and employees of the Canadian Pacific Railway, who permitted the use
of their Swiss guides for the necessary higher climbing and in many ways rendered
very substantial assistance to the expedition. Finally to the packers and
outfitters, Messrs. Robert W. Campbell and George W. Taylor, with their indis-
pensable though often unwilling cayeuses, the writer desires to gratefully
acknowledge the most generous and courteous treatment.

W. H. SHERZER.

THE MICHIGAN STATE NORMAL COLLEGE,
YPSILANTI, MICH., December, 1906.

TABLE OF CONTENTS.

PAGE

Advertisement ............ iii

Preface . . . . . . . . . . . . . . v

List of Illustrations ........... xi

CHAPTER I.
INTRODUCTION.

1. Geographical Data ............ i

a. Physiographic features ..'........ i

b. Streams . . . . . . . . . . . ' . a

c. Glaciers selected for study . . . . . . . . . .3

2. Historical Data ............ 3

3. Geological Data ............ 5

a. Stratigraphy ............ 5

b. Physiographic changes .......... (,

c. Lakes .........'.... 7

d. Alterations in drainage .......... s

4. Climatic Data ............ 10

a. Geographic distribution of moisture . . . . . . . .10

b. Chinook winds ... ....... 12

c. Oscillations in climate . . . . . . . . . .15

CHAPTER II.
VICTORIA GLACIER.

1. General Characteristics . . . . . . . . . . .19

2. Nourishment ....'......... 20

3. Double Tributary . . . . . . . . . . . .22

a. Mitre Glacier; the host .......... 22

b. Lefroy Glacier; the parasite . . . . . . . . .22

4. Drainage ............. 24

a. Surface ablation ........... 24

b. Surface drainage ........... 25

c. Marginal drainage . . . . . . ... . . .27

d. General drainage brook .......... 28

e. Water temperatures . .......... 29

5. Forward Movement ........... 29

a. Measurements . . . . . . . . . . .29

b. Frontal changes ........... 32

c. Shearing . . . .. . . . . . . .34

d. Crevasses ............ 36

vii'1 CONTENTS.

CHAPTER III.

VICTORIA GLACIER (continued).

PAGE

1. Glacial Structure ... • 38

a. Stratification ... -38

b. Dirt zones

c. Granular structure . 39

d. Capillary structure . 4i

e. Melting features 4 1
/. Blue bands . 42
g. Ice dykes . 45

2. Surface Features . .46

a. Superficial de'bris . 4<>

b. Lateral moraines . -47

c. Medial moraine • 49

d. Terminal moraine . . 49

e. Dirt bands

f. Dirt stripes . 54

g. Dust and pebble wells 54
h. De'bris cones .

i. Glacial tables . 5 6

j. Surface lakelets

k. Rock reflection 57

3. Former Activity

a. Terminal moraines

b. Lake Louise basin

c. Lake Louise delta

d. Lake Louise Valley . .61

e. Ancient till sheet

CHAPTER IV.
WENKCHEMNA GLACIER.

1. General Characteristics . .62

2. Piedmont Type . • 63

3. Nourishment ..... 64

4. Drainage ....." . .64

5. Moraines ....... -65

6. Crevasses ..... -67

7. Movement about the Front ..... -67

8. Former Activity .... .68

a. Bear-den moraines ....

b. Moraine Lake ...... .69

c. Valley of Ten Peaks -7°

CONTENTS. IX

CHAPTER V.
YOHO GLACIER.

PAGE

1. General Characteristics . . . . . . . . . . .71

2. Nourishment . . . . . . . . . . . . .72

3. Distributary . . . . . . . . . . . • • 73

4. Moraines ............ 74

5. Crevasses . . . . . . . . . . . . -75

6. Ice Structure ............ 76

7. Drainage . . . . . k . . . . . . - 77

8. Frontal Changes ............ 78

9. Former Activity . . . . . •• . . . . . . .78

a. Moraines ............ 78

6. Plucking action ......... 79

c. Yoho Valley . . • . . . . . . . . . .80

CHAPTER VI.
ILLECILLEWAET GLACIER.

1. General Characteristics . . . . . . . . . . .81

2. Nourishment . . . . . . . . . . . . .82

3. Moraines . ........... 83

a. Surface ddbris . . . . . . . . . . . .83

b. Left lateral moraine ........... 83

c. Terminal moraines .......... 84

d. Right lateral moraine .......... 85

0. Boulder pavement . ....... 85

4. Crevasses ... ......... 86

5. Ice Structure . ............ 86

6. Drainage ............. 87

a. Surface and marginal drainage ......... 87

b. Terminal drainage ........... 88

c. Temperatures ............ 89

7. Forward Movement ........... 89

8. Frontal Changes ............ 31

a. Recession data .......... .91

b. Ice waves ............ 93

9. Former Activity

a. Rock scorings

b. Bear-den moraines

95
95
96

X CONTENTS.

CHAPTER VII.
ASULKAN GLACIER.

PAGE

1. General Characteristics . -97

2. Piedmont Characteristics . • 98

3. Nourishment .

4. Moraines . I0°

5. Crevasses

6. Ice Structure .

7. Drainage

8. Frontal Changes . .106

9. Former Activity

a. Development and decadence

b. Bear-den moraines .

c. Rate of retreat • Io8

CHAPTER VIII.

SUMMARY AND CONCLUSIONS.

1. Physiographic Changes in the Region • I09

a. Mesozoic peneplain .

b. Pre-pleistocene erosion

c. Pleistocene erosion

d. Pleistocene deposition

2. Precipitation .

a. Geographic distribution

b. Climatic cycles . . l 1 2

c. Ice waves

3. Piedmont Type of Glaciers JI4

4. Parasitic Glacier . T 1 4

5. Bear-den Moraines .

6. Surface Features . .118

a. Dirt bands, zones, and stripes

b. Differential melting effects .

7. Ice Structure .... I2°

a. Stratification .

b. Shearing

c. Blue bands .

d. Ice dykes

e. Glacial granules

8. Theories of Glacial Motion . .129

9. Color of Ice and Glacial Water . . . T3'

LIST OF ILLUSTRATIONS.

FACING
PAGE

PLATE I. — Mt. Victoria and Lake Louise ..... Frontispiece i

PLATE II. — General view of Canadian Rockies from Mt. Balfour .... 6
PLATE III. — Map of Victoria Glacier . . . . . . . . .19

PLATE IV. — i. General view of Victoria Glacier, looking southward. 2. Lefroy tribu-
tary and Victoria Glacier . . . . . . . . . . .20

PLATE V. — i. Debris-covered nose of Victoria Glacier. 2. Neve' field of Victoria

Glacier ............. 21

PLATE VI. — •!. Path of avalanche along Victoria ne"ve". 2. Hanging glacier upon Mt.

Victoria . . . . . . . . . . . . .22

PLATE VII. — i. Hanging glacier upon Mt. Lefroy. 2. Double neVe1 field of Mitre

Glacier .............. 23

PLATE VIII. — i. Parasitic Lefroy Glacier being carried by Mitre Glacier. 2. Western
face of Mt. Aberdeen. 3. Surface drainage stream upon Victoria Glacier. 4. Ob-
lique front of Victoria Glacier .......... 24

PLATE IX. — i. First stage in formation of a moulin, Lefroy Glacier. 2. Marginal
lakelet on west side of Victoria Glacier. 3. Inner end of abandoned drainage
tunnel, Victoria Glacier. 4. Mouth of abandoned drainage tunnel, looking outward;
Victoria Glacier . . . . . . . . . . . .27

PLATE X. — i. Drainage from Victoria Glacier after a day of much melting. 2. Refer-
ence boulder A, Victoria Glacier. 3. Stratified ice front, Victoria Glacier. 4. Front
of Victoria Glacier, showing irregular stratification and shearing . . . .28

PLATE XI. — Time of contact between two dirt zones, Lefroy Glacier. 2. Dirt zones

upon Lefroy Glacier ........... 39

PLATE XII. — i. Glacier capillaries, Yo ho Glacier. 2. Glacier capillaries, Illecillewaet
Glacier. 3. Stratification on wall of ice tunnel, Victoria Glacier. 4. Blue bands
in Lefroy Glacier ............ 41

PLATE XIII. — i. Blue bands near nose of Illecillewaet Glacier. 2. Contorted blue
bands, Yoho Glacier. 3. Ice dyke filled with horizontal ice prisms. 4. Crevasse
in Victoria Glacier ..... .42

PLATE XIV. — i. Stony till, left lateral moraine, Victoria Glacier. 2. Sharply crested

left lateral moraine, Victoria Glacier . . .' . . 47

PLATE XV. — i. Ground moraine, Lefroy Glacier. 2. Right lateral and medial mo-
raines of Victoria Glacier .......... 48

PLATE XVI. — i. Formation of Forbes's dirt bands, Deville Glacier. 2. Forbes 's dirt

bands, Victoria Glacier ........... 50

PLATE XVII. — i. Forbes's dirt bands, Asulkan Glacier. 2. Dust wells, Victoria
Glacier. 3. Small dirt cone, Victoria Glacier. 4. Same cone, dirt veneering
removed ............. 53

Xll LIST OF ILLUSTRATIONS.

PACING
4 PAGE

PLATE XVIII. — i. Glacial table, Victoria Glacier. 2. Dethroned glacial table, Vic-
toria Glacier . . 56

PLATE XIX. — i. Boulder mound, Wenkchemna Glacier. 2. Surface lakelet, Vic-
toria Glacier . . . . -57

PLATE XX. — Map of delta, head of Lake Louise .61

PLATE XXI. — The Continental Divide, Canadian Rockies ...... 62

PLATE XXII. — Map of Wenkchemna Glacier ........ 63

PLATE XXIII. i. — Drainage brook from Wenkchemna Glacier. 2. General view of

eastern end of Wenkchemna Glacier ......... 64

PLATE XXIV.- — i. Front of Wenkchemna Glacier. 2. Front of glacier showing
forest invasion. 3. Disintegrated blocks of bear-den moraine. 4. Melted area
on north side of surface block of quartzite, Victoria Glacier ..... 68

PLATE XXV. — General view of Moraine Lake and eastern end of Wenkchemna Glacier . 70
PLATE XXVI.— Map of Yoho Glacier . . 71

PLATE XXVII. — i. Ice distributary from Yoho Glacier. 2. Ice plucking upon a .
mountain peak, head of Yoho Valley. 3. Yoho 'Glacier, head of Yoho Valley.
4. Three-hundred-foot ice arch, Yoho Glacier . . . . . . -74

PLATE XXVIII. — i. Knoll and ridges of ground morainic material in front of Yoho

Glacier. 2. Yoho Valley, showing Wapta and Waputik snowfields ... 78

PLATE XXIX.- — i. Formation of Forbes's dirt bands from crevasses, Yoho Glacier.

2. Hanging Valley, head of Yoho Glacier ........ 80

PLATE XXX. — Map of Illecillewaet Glacier . .... .81

PLATE XXXI. — Tongue and moraines of the Illecillewaet Glacier, August, 1899 . . 82

PLATE XXXII. — -i. General view from Roger's Peak. 2. General view of Illecillewaet

Glacier . . . . . . . . . . . . .84

PLATE XXXIII. — Map of the Selkirk snowfields and glaciers ..... 85

PLATE XXXIV. — i. Beginning of subglacial fluting by pressure melting, Illecillewaet
Glacier. 2. Subglacial fluting, Illecillewaet Glacier. 3. Roches moutonne"es
near nose of Illecillewaet Glacier ....... .86

PLATE XXXV. — i. Regelation of ice blocks at foot of ice cascade, Illecillewaet Glacier.

2. Stratification in upper part of Illecillewaet Glacier ..... 88

PLATE XXXVI. — i. Illecillewaet Glacier in 1888. 2. Illecillewaet Glacier in 1905 . 92

PLATE XXXVII. — -i. Bear-den moraine made conjointly by the Illecillewaet and

Asulkan Glaciers. 2. Illecillewaet Glacier in 1897 ...... 96

PLATE XXXVIII. — Map of Asulkan Glacier . . -97

PLATE XXXIX. — i. General view of Asulkan Glacier in 1902. 2. The Asulkan

glaciers and snowfields from Mt. Avalanche ..... .98

PLATE XL. — i. Left Asulkan moraine. 2. Nose of Asulkan Glacier, 1904 . 100

PLATE XLL— i. Stratification of Asulkan Glacier. 2. General view of Asulkan Glacier

in 1898. .102

PLATE XLII. — i. Development of seracs from glacial blocks, Asulkan Glacier. 2. Dis-
rupted quartzite blocks illustrating plucking power of glaciers . . . .108

GLACIERS OF THE CANADIAN ROCKIES

AND SELKIRKS

(Report of the Smithsonian 'Expedition of 1904.)

Bv WILLIAM HITTELL SHERZER, Ph.D.

CHAPTER I.

INTRODUCTION.

i. GEOGRAPHICAL DATA.

a. Physiographic features. — The Canadian Pacific Railway crosses the
Rockies and Selkirks between north latitude 51° and 51° 30', working its way
up the left bank of the Bow River and its small tributary Bath Creek, to the
Kicking Horse Pass, attaining an altitude of 5,329 feet above sea-level. Upon
the more abrupt western slope of the Rockies the road follows the left bank of
the Kicking Horse River to its junction with the Columbia, crosses this great
waterway of the mountains, and slowly ascends the eastern slope of the Selkirks
along the left bank of the Beaver. The summit of the Selkirks, Rogers' Pass,
is crossed at an elevation of 4,351 feet, whence there is rapid descent along the
swift -flowing Illecillewaet to the Columbia again, which has encircled the system
to the north, forming the "Big Bend." These transverse mountain valleys
are lined with most majestic peaks, many of them rising a mile above the valley
floor and furnishing some of the grandest of mountain scenery upon the American
continent. The highest peak in the Rockies, seen from the railway, is Mt.
Temple, with an elevation of 11,627 feet, and m tne Selkirks, Mt. Sir Donald,
10,808 feet. Numerous peaks range from 10,000 to 11,000 feet and are believed
to culminate to the northward in latitude 52° to 53°.

The Rockies and Selkirks, together with the Gold and Coast ranges to the
west, make up the Great Cordillera in this part of Canada. North of the inter-
national boundary this great system is much narrower than in the United States,

2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

having a total width of about four hundred miles, and the component ranges
are straighter and more regular. The systems are progressively higher from
the coast eastward, culminating in the Rockies proper, which stand as a lofty
buttress along the western margin of the great central plains. Between the
Coast and the Gold ranges there lies an interior plateau a hundred miles wide
with an average elevation of about 3,500 feet above sea-level. The Gold,
Selkirk, and Rocky systems are separated by the Columbia and the Columbia-
Kootenay valleys, made by the action of water and ice along the strike of the
geological formations, assisted probably by some dislocations of the strata.

The Rocky Mountains, or as formerly called, the Stony Mountains, consist of
an imposing array of parallel ranges with a general trend in this region of north-
northwest to south-southeast, separated by longitudinal valleys and attaining
a total breadth of 40 to 50 miles. Compared with the systems to the west
they are strikingly rugged in character and free from vegetation. Skirting the
eastern border, and a part of them both geologically and structurally, are the
"foot-hills, " consisting of folded parallel ridges, reaching out 15 to 20 miles and
merging into the "plains" at an elevation of about 3,300 feet.

b. Streams. — The main streams occupy the longitudinal valleys for a
portion of their course, leaving the mountains by the transverse valleys, which
extend into the foot-hills. According to Dawson the base-level of the streams
upon leaving the mountains to the eastward is about 4,360 feet, while to the
west it is about 2,450 feet above sea-level. Upon the eastern slope of the Great
Continental Divide the waters are gathered into the Saskatchewan and reach
the Atlantic Ocean by way of .Hudson Bay; while those to the west drain into
the Columbia River and work their way to the Pacific Ocean. As pointed out
by Dawson, the actual water parting does not correspond entirely with the
highest crest line of the mountains, but lies to the eastward, in which direction
it seems to be moving. Between the international boundary of north latitude
49° and 52°, the Rocky Mountains are sharply separated from the Selkirks to
the west by the Columbia-Kootenay Valley, which maintains a considerable
breadth and a remarkably straight course through more than three degrees of
latitude. This valley is filled with drift materials to a considerable depth and
is undergoing but little erosion, the river simply cutting tortuous channels
through the loose deposits. The eastern side of the valley is generally steep
and escarpment-like, while the western is rounded and wooded. The Columbia
starts within a mile and a half of its southward-flowing tributary, the Kootenay,
and moves northwestward in a great sweep as though intent upon capturing
the drainage .of the region before starting for the sea. In this great fold it en-
closes and sharply limits the less rugged, but picturesque, Selkirk System,
with its subdued outlines and forested slopes. Some of the eastern ranges are
continuous and have the same general trend as those of the Rockies, but, in
general, there is less regularity and continuity in the arrangement of crests
and peaks, and they do not attain as great a height. The drainage is all into
the Columbia River, and the streams are unable to develop any considerable

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 3

size. Being so largely glacier fed here, as well as' in the Rockies, the streams
maintain themselves during the summer months, but reach their highest stage
in the late spring, or early summer, from the melting of the snows, when the
Columbia may rise 30 to 40 feet above its usual level. The streams are generally
turbid with glacial sediment that gives them a milky, or yellowish, appearance,
changing to green and, upon the loss of the sediment, to a blue color wherever the
water is of considerable depth. The lakes of the region owe their origin mainly to
former glacial action, consisting either of rock-basins, or of depressions in the
glacial or fluvio-glacial deposits. Certain ones have been dammed back by
morainic material deposited either beneath or at the extremities of glaciers of
greater extent than at present. Those lakes which receive glacial sediment,
or which are shallow, have a greenish cast, while those free from sediment
and of moderate to considerable depth are rich blue.

c. Glaciers selected for study. — The glaciers selected for study lie close to the
main crests of the systems above described, between north latitude 51° and 52°,
and west longitude n6°and 117° 30', from 160 to 200 miles north of the inter-
national boundary between the United States and Canada. They are but a
few of a series available for study, those being selected which are most easily
reached by well established trails. They are at such low altitudes that one may
comfortably ride almost to the nose of each and none require climbing except
to reach the neVe" regions. The two most easterly of the glaciers here discussed,
the Victoria and Wenkchemna, lie east of the Great Divide in Alberta, the other
three are west in British Columbia.

2. HISTORICAL DATA.

The establishment of the international boundary to the south, along the
49th parallel, and the opening of the railway in 1885 called for geographic,
geologic, and topographic work which was started by the various Dominion
departments concerned and is still in progress. Dr. George M. Dawson began
his work in 1874 along the boundary and extended it northward to include the
region pierced by the railway, where he was assisted by R. G. McConnell.
Topographic work of a preliminary nature, along the line of the railway, was
begun in 1886 by J. J. McArthur. Photographic methods were introduced into
the survey in 1889 by Director Deville and the accompanying triangulation of
the "railway belt" placed in charge of W. S. Drewry, D. L. S. The same
year Mr. St. Cyr made a survey along the upper Columbia, between the Selkirks
and Rockies, and in 1896 he and McArthur continued the work from Revelstoke
down the Columbia and Arrow Lakes, with the view of connecting the surveys
of the railway belt with those of the boundary commission.1 Two topographic
maps, upon a scale of two miles to the inch, were issued in 1902 by the Depart-
ment of the Interior, under the direction of James White, geographer. These
are the Banff and Lake Louise sheets and are issued by the department at

1 The reports of the work of McArthur, Drewry, and St. Cyr will be found in the Annual Reports of the
Canadian Dept. of the Interior for 1886, 1888 1889, 1890 1891, 1892, and 1893.

4 GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

Ottawa. The topographic work of the mountains is now in the hands of Mr.
Arthur O. Wheeler and there is being issued an enlarged map (scale 5,000 feet
to an inch) of a section of the mountains lying between the railway and the
Great Divide and extending from the Kicking Horse to the Vermilion Pass.
This map includes completely the regions surrounding the Victoria, Horseshoe,
and Wenkchemna glaciers, with corrected elevations, essentially the same
territory covered by Wilcox's map, 1896, on the scale of an inch and a half to
the mile. Based upon work done during the seasons of 1901 and 1902, there
will be issued with vol. n of Wheeler's Selkirk Range an admirable piece of
mountain mapping, extending from the Columbia to the Columbia, across the
Selkirks along the line of the railway.

The opening of the railway and the wonderful attractions of the region
brought in a body of non-professional explorers and mountaineers, among the
first of whom was the Rev. W. S. Green, Carrigaline, Ireland. He spent the
working season of 1888 in the Selkirks, using Glacier House as a base, and gathered
material for an interesting volume, Among the Selkirk Glaciers, Macmillan &
Co., 1890. The map accompanying the volume, originally published in the
Proceedings of the Royal Geographical Society, vol. xi, 1889, was the first attempt
at detailed mapping in the Selkirks. One year earlier than Green, in 1887,
Messrs. George and William Vaux, Jr., of Philadelphia, visited Glacier House,
secured a valuable collection of photographs and began a series of observations
upon the glaciers to which frequent reference will be made in the later chapters
of this report. During the closing decade of the last century, and the opening
decade of the new, there has been much work done in the region of an exploratory
and mountaineering character. There should be mentioned especially the names
of Wilcox, Fay, Parker, Collie, Stutfield, Allen, Habel, Topham, Thompson,
Huber, Sulzer, Noyes, and the English ladies Benham, Tuzo, and Berens.
Besides three superbly illustrated and attractively written volumes by Wilcox,
Wheeler, and, conjointly, by Stutfield and Collie,1 there have been prepared
a number of descriptive papers for the scientific societies and magazines. A
bibliography of the region, full but not complete, will be found in Appalachia,
vol. x, 1903, pp. 179 to 186. The Canadian artist, F. M. Bell-Smith, of Toronto,
has spent many seasons in the mountains and, based upon the various maps,
photographs, and original sketches, has prepared relief maps of the best known
areas of the Rockies and Selkirks. Copies of these maps are placed in the hotels
operated by the railroad.

It is not likely that these mountain valleys ever supported anything more
than a scant Indian population, owing to the scarcity of fish, game, and available
pasture. Providing food, en route, has always been a precarious matter for
exploring parties. Aside from the marmot and rock-rabbit and an occasional
porcupine, there is a strange and impressive feeling of desertion. The few

1 Camping in the Canadian Rockies, Wilcox. G. P. Putnam's Sons, N. Y., 1896. The Rockies of Canada,
Wilcox. Putnams, 1903. Climbs and Explorations in the Canadian Rockies, Stutfield and Collie. Longmans,
Green & Co., N. Y., 1903. The Selkirk Range, Wheeler. Government Printing Bureau, Ottawa, 1905.

GLACIERS OF THE_CANADIAN ROCKIES AND SELKIRKS. 5

birds that one meets seem awed into silence by the grandeur of their surround-
ings. The mountain Crees had possession of the region at the coming of the
white traders and trappers, but within rather recent time have been assimilated
by the Stoneys, a tribe of Assiniboines, from the plains to the east.

3. GEOLOGICAL DATA.

a. Stratigraphy. The first work of a geological nature in this region was by
Dr. Hector in 1858 to 1860, as a member of the Palliser expedition, his observa-
tions being confined mainly to the Rockies and the region to the east. A
geological map and numerous sections were prepared to accompany a paper
presented to the Geological Society of London, in advance of the publication
of the results of the expedition.1 For detailed knowledge of the geology of the
Rockies and Selkirks we are indebted mainly to Dr. George M. Dawson and his
assistant R. G. McConnell, the former of whom began his work in 1874, as geol-
ogist of the boundary commission. The Bow River region was entered in 1881
and in the Annual Report of the Geological Survey for 1885 there was published
a preliminary report upon the geology of the Rockies lying between the boundary
and north latitude 51° 30'. The report was accompanied by a large scale
geological map, which was followed the next year with a geological section by
McConnell, approximately along the line of the 5ist parallel of latitude. Work
was extended westward into the Selkirks and, at the Washington meeting of the
Geological Society of America, Dr. Dawson, in 1890, presented the results of his
observations amongst these ranges.2 The present Geological Survey, under the
directorship of Robert Bell, is still at work upon the detailed study of portions
of the region.

The Selkirks and Rockies consist of an enormous complex of sedimentary
strata, 50,000 to 60,000 feet in thickness, underlain by crystalline rock. In
age they range from the Archaean to the Laramie, at the close of which the
final stages of upheaval were accomplished. The rock strata graduate in age
from the west, eastward, and were folded and faulted by pressure from the
west, by which they were forced against the resistant layers underlying the
"great plains." The crystalline rocks of the series, of presumable Archaean
age, consist of gray gneisses, passing into schists, and occur only along the
western margin of the Selkirks, where they constitute the Shuswap series. No
trace of them has yet been discovered in the Rockies. Overlying the series occurs
the Nisconlith, with an estimated thickness of 15,000 feet, consisting of dark
colored argillite-schists and phyllites, showing various stages of alteration from
true argillites to micaceous schists. Interbedded layers of dark limestone and
quartzite are seen in certain sections. Although the beds yielded no fossils
they were referred to the Cambrian by Dawson, because of their relation to the

'"On the Geology of the Country between Lake Superior and the Pacific Ocean," James Hector,
M.D., 1861, Quart. Journal Geol. Society, vol. xvn, pp. 388 to 445.

2 "Note on the Geological Structure of the Selkirk Range, " Geol. Soc. of Amer., vol. 2, 1891, pp. 165 to
176. An extract from this paper is given in Wheeler's Selkirk Range, vol. i, pp. 405 to 409.

6 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

crystallines and their" supposed equivalency with strata of known Cambrian
age in the Rockies to the eastward. The crest of the Selkirks, the region in
which occur our snowfields and glaciers, consists of the folded Selkirk series,
with an estimated thickness of 25,000 feet and believed to be of Cambro-Silurian
age. The strata have been forced into a synclinal fold, which terminates to
the eastward by a thrust fault, produced by the western half of a sharp anticline
being thrust upward with reference to the eastern half. The rocks consist of
gray schists and quartzites, passing into grits and conglomerates which weather
to pale yellowish or brownish colors. The latter are often more or less schistose
from pressure and other metamqrphic agencies, silvery mica, or sericite, being
developed. At times the strata are wrinkled and contorted.

Passing into the Rockies, to the eastward, we find them made up very
largely of the representatives of the Nisconlith and Selkirk series just described,
but known in the report of Dawson as the Bow River and Castle Mountain series.
In the western part of the Rockies, adjacent to the Columbia, the upper and
younger of the two is continued from the Selkirks, showing crumpling and folding,
with metamorphism, but without faulting. The rocks are dipping eastward
and have their "strike" parallel with the mountain ranges. Along the center
of the range the folds are broad and sweeping, while eastward, for about 25
miles, there is a succession of thrust faults, running parallel with the ranges,
the maximum vertical displacement being estimated at 15,000 feet. McConnell
made out seven of these faxilts, giving rise to a series of massive mountain
blocks Cresting in succession upon one another and forming escarpments to the
east and relatively gentle slopes to the west. It was to this type of mountain
that Leslie Stephen applied the suggestive term "writing-desk."

The ranges making up the central portion of the Rockies are of the Castle
Mountain series (Selkirk series) and of Cambro-Silurian age. They are more
regular and depart less from the_ horizontal than the strata to the east and
west. Mt. Stephen, on the line of the railway, gives a 5,ooo-foot section of the
series, one shaly band being "remarkably rich in Cambrian trilobites. The total
thickness of the series is estimated at 10,000 feet, as compared with 25,000
feet in the Selkirks, and consists of limestone and dolomite, with calcareous
schists and shales. These rocks give the steep-sided, massive, block-like cliffs,
typically shown in Castle Mountain, which has furnished the name for the
series. These rocks extend lengthwise of the central ranges to the Yoho
Glacier at the north and the Victoria and Wenkchemna glaciers to the south.
At the base of Mt. Stephen, at the head of Lake Louise, and in the Bow River,
there has been brought up from below by an anticlinal fold the "Bow River
Group, " or Nisconlith series of the Selkirks. :This is of Cambrian age, estimated
.at 10,000 feet in thickness, and consists of quartzites and conglomerates, with
dark gray, purplish, and greenish .argillites.

b. Physiographic changes. When viewed from a high elevation the rough
ridges and jagged peaks appear to blend, as far as the eye can reach, into a great
plateau with ^ a notably even sky-line (pi. n), giving the appearance of an

1

O

f
—

*— CJD

3 S.

-J O

UN

.•<S1.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 7

uplifted peneplain, similar to that observed by Gilbert farther north in Alaska.1
The upheaval of such a mass, by lateral pressure, would give rise to troughs and
open gaping crevices parallel with the previously formed mountain folds and
into these the drainage streams would, for the most part, be diverted. Deepened
by stream action, widened by atmospheric agencies, and still further modified
by Pleistocene glaciers, we have the longitudinal valleys of the Rockies and
Selkirks. The transverse valleys, noted by Dawson as extending into the
foot-hills and due to "causes not now apparent," probably mark the location
of drainage streams developed while the peneplain was being formed and antedate
the final upheaval of the region. That these valleys were occupied by extensive
glaciers, presumably in Pleistocene time, is everywhere evidenced by the
morainic accumulations, rounded rock contours, glaciated surfaces, extensive
plucking, truncation of mountain spurs,~amphitheaters, rock basins, and hanging
valleys. The valleys generally were filled with ice to a depth of 2,500 to 4,000
feet during the maximum period of glaciation, the height, as pointed out by
Wilcox, rarely falling below 7,000 feet above sea-level. Either because the
mountains were so completely enveloped in ice and snow, or because of the
nature of the final retreat, extensive terminal moraines were not formed in
the main valleys. Ground-morainic deposits, however, hundreds of feet
thick, occur in places favorable for their lodgment beneath the ice. Near
Banff, in the Bow and Cascade valleys, Wilcox discovered evidence of two
distinct till-sheets, the older highly charged with pebbles, with little clay, the
younger consisting mainly of very hard clay.2 In the extension of the Bow
Valley to the eastward of the mountains, McConnell and Dawson found three
separate till-sheets. The lowest and oldest of these appeared to have been en-
tirely derived from the mountains and to pass eastward gradually into the
"Saskatchewan gravels" of the plains. For this formation Dawson suggested
the term "Albertan, "3 which he regarded as of pre-Kansan Age.

The upper two boulder-clays contained rock fragments of both eastern and
western origin, each variety preponderating in the direction of its origin, showing
a commingling of the deposits of the Cordilleran and Hudson Bay ice sheets. The
middle of the three till-sheets Dawson correlated with the Kansan and the
upper with the lowan (p. 509). In the light of our present knowledge the upper
would be referred to the Illinoian, which succeeded the Kansan in the upper
Mississippi Valley and was of much wider extent than the lowan. The corre-
lation of these sheets with those observed at Banff has not yet been made.

c. Lakes. In the very interesting paper above referred to, Wilcox recog-
nizes four types of lakes in those portions of the Canadian Rockies which came
under his observation:

First — Lakes lying in depressions of the valley drift, often in chains of

i Harriman Alaska Expedition, vol. HI, Glaciers and Glaciation, Gilbert, p. 183.

2"A Certain Type of Lake Formation in the Canadian Rocky Mountains," Jour, of Geol., vol. vn,
1899, p. 249-

»"Note on the Glacial Deposits of Southwestern Alberta," Jour, of Geol., vol. HI, 1895, p. 510. See
also note on page 384, vol. in, Geology, by Chamberlin and Salisbury.

8 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

three, or four, and especially numerous near the summits of the mountain
passes. These lakes show no regularity in form, or location, are usually shallow,
and frequently have neither inlet nor outlet.

Second — Lakes dammed by terminal moraines. In the same class may be
included those dammed by alluvial cones, or deltas formed in preexisting lakes.

Third — Lakes lying in rock basins, excavated by former glacial action.

Fourth — A special type of lake, of which Lake Louise is an example, formed
just within the mouth of a tributary valley. These lakes are leaf-shaped and
from three to ten times longer than they are wide. They owe their existence
to the presence of a ridge of ground-morainic material, thrown across the mouth
of the tributary valley from the up-stream side and curving gently out into the
trunk valley. These ridges have apparently been formed beneath the icei
when the valleys carried glaciers, by the joint action of the tributary and trunk
glaciers. The lakes may be shallow and so filled with silt that they are reduced
to swamps, or, where the ice was especially active, as in the case of Lake Louise,
the depth may still be surprisingly great. As recognized by Wilcox, these
lakes may present a combination of the rock-basin and morainic-dam types.

d. Alterations in drainage. In a region of the character above described,
exposed for countless ages to the effects of weather, water, and ice, marked
alterations in the drainage would be expected, such as reversals, stream capture,
and the migration of divides. A study of the Columbia-Kootenay Valley has
shown that the upper 200 miles of the Columbia flowed at one time to the south,
instead of encircling the Selkirks to the north as at present. This must neces-
sarily have been the case so long as that portion of the valley known as the
"Big Bend" was in possession of the ice. The withdrawal of the glaciers into
the tributary valleys would permit the present northward flow from the Columbia
Lakes while the Kootenay, flowing in the opposite direction in the valley to
the east, enters the main valley, approaches within one and one-half miles of
the head-waters of the Columbia, but completely skirts the Selkirks to the south
before joining it. Upon the opposite side of the Rockies attention has been
called by Dawson, McConnell, and Ogilvie to. changes in the Bow and its tribu-
taries. The long, slender Lake Minnewanka Valley, near Banff, was evidently
the course of a prepleistocene valley that was occupied and modified by an ice
stream during the maximum period of glaciation of the region. Dawson con-
sidered this to markr the former course of the Bow, which was deflected to the
southeastward, along the strike of the soft Cretaceous shales, when the lake
valley was occupied by ice.1 From barometric observations made by Dawson,
his assistant McConnell noted that the Ghost River opposite the Devil's Gap,
the mouth of the Lake Minnewanka Valley, is considerably higher than the valley
floor and concluded that the Ghost River turned and entered the mountains
through this valley, joining thus the Cascade and with it forming a tributary

1 Annual Report of Canadian Geological Survey for 1885, "Preliminary Report on the Physical and
Geological Features of that Portion of the Rocky Mountains between latitudes 49° and 51° 30', " 1886,
p. 1418.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 9

of the Bow.1 The topographic map of the region, issued in 1902, shows that
the river opposite the Gap is fully 100 feet higher than the level of Lake Minne-
wanka and about 400 feet higher than the present level of the Bow in the
vicinity of Banff. According to this view of McConnell the ice-filled Minne-
wanka Valley compelled the Ghost River to find for itself a new course across
the foot-hills to the eastward, which it deepened sufficiently, assisted probably
by the ice, to prevent the return of the river into its former course upon the
withdrawal of the glacier from the valley. In 1904 Dr. I. H. Ogilvie examined
the region and reached the conclusion that the upper Bow and Minnewanka
valleys were formerly continuous and that the Bow Valley below Banff was
occupied by a stream which cut back into the soft shales until it tapped the
upper Bow and effected its capture. She concluded2 that this had been ac-
complished in prepleistocene time and that the Lake Minnewanka Valley was
occupied by a glacier, fed by hanging glaciers, which moved westward, rather
than to the east, deepening the western end of the valley and forming certain
morainic deposits about the western end of the lake. The lake itself and its
southwesterly drainage would then date from the withdrawal of the ice from
the Cascade and Minnewanka valleys. Dr. Ogilvie holds that the drainage in
the Lake Minnewanka Valley before the advent of the glaciers was eastward.

The bed of the Bow River at Banff is approximately 4,500 feet above tide,
while that of the Ghost River, opposite the Devil's Gap, is not far from 4,900
feet. The present Bow, in the same distance as that from Banff to this portion
of the Ghost River, drops 300 feet, so that the bed of the Bow at Banff is some
700 feet lower than it should be in order to have the upper Bow leave the
mountains by the Devil's Gap and thence by the lower Ghost River. All
will grant that this is too much cutting to expect of the Bow in postpleistocene
time and that Dr. Dawson's theory of the diversion of the Bow by the Pleistocene
glaciers is untenable. Noting that the Ghost River has also been deepening
its bed, with a much steeper gradient and presumably for as long a time as the
Bow, the above 700 feet must represent the excess of cutting by the upper Bow,
when compared with the Ghost, since its capture by the lower Bow, upon Dr.
Ogilvie's hypothesis. We have no knowledge of the depth of the drift deposits
opposite the Devil's Gap, but there is no reason to think that they would be
any deeper there than in the valley of the Bow. The explanation that lies
nearest at hand is that of McConnell, viz., that the Ghost River upon reaching
the foot-hills made a sharp turn and reentered the mountains by the Devil's
Gap, but was diverted eastward when the Minnewanka Valley became ice-
filled. According to this hypothesis the valley of the Ghost from the Gap
down should show less maturity than that above the Gap, except so far as it
may have been modified by ice action. During the period of maximum glacia-
tion the ice movement in the Minnewanka Valley must have been eastward,

1 Annual Report of Canadian Geological Survey for 1886. Report on the Geological Features of a
Portion of the Rocky Mountains, 1887, R. G. McConnel!, p. gD.

^"Geological Notes on the Vicinity of Banff, Alberta," Jour, of Geol., vol. xn, No. 5, 1904, pp. 408
to 414.

10 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

any tendency towards a westerly movement being checked by the ice in the
Cascade and Bow valleys. The westerly movement noted by Dr. Ogilvie was
simply a minor episode toward the close of the Pleistocene glaciation.

4. CLIMATIC DATA.

a. Geographic distribution of moisture. The climatic conditions of this
section of country are peculiarly dependent upon the physiographic features
above outlined, combined with its proximity to the Pacific. The centers of
the areas of low pressure commonly move in from the ocean to the northward
of the region, give rise to westerly winds which convey the warmth and moisture
of the Pacific currents across British Columbia and Alberta. At Nanaimo,
upon Vancouver Island, separated from the mainland by 30 miles of strait,
the precipitation records available show a rain- and snowfall combined ot 41.36
inches, only 5 per cent, of which falls as snow. Opposite, upon the mainland,
the total precipitation at Vancouver rises to 63.06 inches, with 4 per cent,
falling as snow. This increase is due to the Coast Ranges, having here a north-
west-southeast trend, which compel the westerly winds to ascend their westerly
slopes, by which rise the air is cooled and its capacity for holding moisture
thereby diminished. At Agassiz, in the lower Frazer Valley, the precipitation
is but slightly less, although it is located some 70 miles from the Strait of Georgia,
up the broad open valley, and about 50 feet above sea-level. Records are
available here since 1890, with the exception of the years 1891 and 1899, and
give for the 14-year series an average precipitation of 62.02 inches, 6 per cent.
of which falls as snow. Over the broad interior plateau which lies between the
Coast and Gold ranges, the temperature is colder and the precipitation relatively
slight. At Kamloops, with an elevation of 1,160 feet and in latitude 50° 41',
the combined rain- and snowfall averages but 10.66 inches. Passing eastward
the air currents impinge upon the westerly slopes of the Gold Range, are again
compelled to ascend to still higher altitudes, with the attendant loss of moisture.
In consequence, the station of Griffin Lake, located in this range at an elevation
of 1,511 feet, and 90 miles east of Kamloops, receives an average precipitation
of 34.37 inches, or over three times as much as the latter place. Crossing the
Columbia a still higher barrier is encountered in the impressive Selkirk system
of ranges, and a correspondingly increased amount of moisture extracted from
the still laden air currents. The records in the Selkirks are unfortunately
meager, but they indicate a precipitation almost as great as that of Vancouver
and Agassiz, and considerably in excess of Nanaimo and Victoria out in the
Pacific. The station of Glacier House is located just west of the main crest
of the Selkirks, in latitude 51° 16' and at an elevation above tide of 4,093 feet.
The average total precipitation here is 56.68 inches, of which 77 per cent, falls
as snow. If the entire amount were precipitated as snow, as is practically the
case upon the peaks and elevated ne"ve" fields, this would represent an average
fall of over 47 feet. This heavy snowfall in the Selkirks and Gold Range has

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

II

necessitated the erection of numerous snow- sheds over the tracks of the Canadian
Pacific Railway, to guard against the frightful avalanches which come crashing
down the mountain slopes.

Although Donald is located but a few miles east of Glacier House, it is upon
the lee slope of the Selkirks and in the Columbia Valley some 1,500 feet lower.
The precipitation here drops to 25.39 inches. The still higher Rockies im-
mediately follow, but the Selkirks have proven greedy and there is relatively
little left in. the way of moisture. Full precipitation data in the region of the
crest ranges of the system are wanting. The snowfall at Field, however,
averages about 27 feet, at practically the same elevation as that of Glacier
House. If we assume that the same ratio holds here, between snow and rain,
as at the latter place the precipitation at Field, lying just west of the main
crest, would be about 42 inches. Passing the continental crest the currents
are drawn to lower levels, they become warmed by their descent, and their
capacity for retaining moisture increases. At Banff the precipitation for the
13 complete years available averages 20.14 inches, of which 39 per cent, falls
as snow. Beyond the foot-hills, at Calgary, the precipitation is reduced to 16.64
inches, of which 28 per cent, is snow. The following table furnishes a summary
of the climatic data of special interest in connection with this report.

TABLE I.
CLIMATIC DATA, FROM RECORDS OF THE CANADIAN METEOROLOGICAL SERVICE.

II

i .

, .

||

B

cd u

Stations, ar-
ranged from West

2

V

•

p

V C

kc"l

*rt p

§5

C TT!

II

11

3 .

o> cd

c3 c

>,.£
"So

Remarks concern-
ing location.

to East.

T3

fl

3.S

II

a B

» °

*J O

M"S

•g.9

S *

2

•a

• 3

If

a £

li

wg

o£
2g

l~

S «

H

>-) ca

S S*

v "cd

11

i-i*o

11

Nanaimo

49°io'

i23°57'

-3°

117

48.7°

90.30

-3-9°

41.36

5

S

East side of Vancouver

Island.

Vancouver

49°i7'

I23°5'

10

136

48.0°,

89.0°

6.0°

63.06

4

4

West slope of Coast

Ranges.

Agassiz

49°i4'

I2I°3l'

70

52

47-7°

103.0°

-13-0°

62.02

6

14

In open valley of Fra-

zer River.

Kamloops

50*41'

I20°20'

270

1160

47-i°

101.0°

-27.0°

10.66

24

10

Interior plateau.

Griffin Lake

5°°S8'

ii803i' 400

1511 44-6°

110.0°

-23.0°

34-37

37

10

Gold Range.

Glacier House

5i°i6'

1 1 7°3O/ 460

4093

36.9°

89.0°

-21.0°

56-68

77

5

Selkirk System.

Donald

SiW

ii7°n' '< 470

2580

37-4°

97-0°

-45-0°

25-39

51

10

Columbia Valley be-

tween Selkirks and

Rockies.

Banff

51*10

"5°3S'

55°

4542

35-3°

88.7°

-48.8°

20.14

39

13

East of Continental

Divide about 15 miles.

Calgary

5I°2'

II4°2'

600

3389

37-2°

95-o°

-49-4°

16.64

38

19

Just beyond the foot-

hills of the Rockies

Quite in contrast with the still higher mountains of this same system to
the south in the United States, the following factors conspire to yield the neces-
sary meteorological conditions for extensive perennial snowfields and glaciers;—

12 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the narrowness of the system, its proximity to the Pacific, the higher latitude,
the arrangement of the Cordilleran ranges with respect to height, and the position
of the region with reference to the ordinary paths of the great cyclonic areas.
If these areas of low pressure commonly crossed the continent so that the path
of their centres lay well within the United States, the prevailing winds would be
easterly over this region. Such winds would be colder in the winter and warmer
in the summer than those which now prevail and capable of supplying relatively
little precipitation. There can be but little doubt that any great shifting of
the paths of the cyclonic areas, either to the north or the south, would lead to
a great reduction in the size of the glaciers of this region, and perhaps to their
complete extinction.

b. Chinook winds. One of the most interesting and important climatic
factors of this section is connected with the prevailing westerly winds and the
north to south trend of the main mountain ranges, giving rise to the so-called
"Chinooks. " These winds must have been long familiar to the aborigines
and voyageurs, but were first noted by Mackenzie about the year 1790 in the
region of the Athabasca and Saskatchewan rivers, bringing clear, mild weather
in the winter and spring. l He ascribed their warmth, very naturally, to the
nearness of the warm currents of the Pacific, their progress over the snowfields
being assumed to be too rapid to permit of their being cooled. The same
winds are known in Montana and still farther south in Colorado, where they
are known as ' ' zephyrs ' ' and ' ' snow eaters. ' ' They also occur in South America
with the passage of westerly winds over the Andes, having been described by
Bishop in the vicinity of San Juan, Argentine Republic, under the name
"zonda. "2 Here they were supposed to derive their warmth from volcanic
sources. The chief characteristics of these, and similar winds are their
warmth and dryness and consequent accompaniment of bright sky. They
are most conspicuous in the winter and spring, but occur also in the summer and
fall. Standing upon the Victoria Glacier, in midsummer, opposite the nose of
Mt. Lefroy, one frequently notes gusts of balmy air sweeping down from the
elevated snowfields and puzzles over the source of the warmth. The following
graphic description of these winds will serve to get before the reader their
climatic importance.

' ' The extreme severity of the winters in certain parts of our northwestern
states, among the Rocky Mountains and along their eastern base, is much
tempered by the prevalence of a mild westerly wind, locally called the chinook.
Its name is derived from that of the tribe of Chinook Indians, living near Puget's
Sound. It is said first to have been applied by the early Hudson Bay trappers
and voyageurs, who, meeting the wind while travelling towards the Pacific
coast, and finding it particularly strong and warm as they approached the lands
ot this particular tribe, called it the chinook wind.

"It is described as a soft, balmy wind, varying in velocity from a gentle

1 Voyages on the River St. Lawrence and through the Continent of North America to the Frozen and Pacific
Oceans in the Years 1789 and 1793, London, 1801, p. 138.

2 A Thousand Mile Walk across South America. Boston, 1869. N. H. Bishop.

GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS. 13

breeze to a steady gale. Though its temperature rarely exceeds 5o°F, yet coming
as it does when one is accustomed to a low temperature, it seems warm by con-
trast with the preceding weather. The thermometer often rises from below
the zero point to 40° or 45° in a few hours, and the maximum temperatures of
the winter months in the Rocky Mountain region nearly always are coincident
with the occurrence of a chinook.

"The sky is usually clear while the warm wind blows, though observers
often note a few leaden-colored clouds of a kind seen only during the chinook.
These clouds are described as pancake-shaped, with peculiarly smooth, rounded
edges, and stand apparently motionless, high above the mountain ranges.

"The continuance of a chinook is as uncertain as its coming. It may
last a few hours or for several days. With a change of wind the temperature
falls rapidly, and winter weather once more sets in.

'The chinook wind possesses to a remarkable degree the power of melting
snow, for it is not only warm, but appears to be dry. Although a foot or more
of snow may lie on the ground at the beginning of a chinook, it disappears
within a very few hours, often seeming rather to evaporate than to melt. For
this reason the chinook is most welcome to the cattlemen on the plains of
Montana and Wyoming. In fact, without it, stock-raising would be almost
impossible, as the dried grasses of the plains, on which the cattle subsist, would
otherwise be buried the greater part of the winter. To a few, however, this
wind, instead of being hailed with delight as a break in the cold of the winter,
is a source of much discomfort. " '

The scientific explanation of this type of wind was first given, in part, by
the American meteorologist Espy,2 and later completed by Helmholz, Tyndall,
and Hann. Simply stated and adapted to the region under discussion, this
explanation may be of interest to many into whose hands this report may fall.
The presence of an area of low barometric pressure to the north gives rise to a
system of rotary air currents, moving counter-clockwise about the center and
constituting a great "whirl." The interposition of mountain barriers, such as
those already described, with a north to south trend, compels the winds, which are
westerly in the southern portion of the cyclonic area, to mount these obstructions,
from the crests and through the passes of which the air is again drawn to lower
levels through the suction of the great rotating mass. As the air rises upon the
windward slope of the mountain range, it experiences less pressure from the
surrounding air, is permitted to expand, and is, in consequence, cooled at the
rate of i° F. for each 180 feet of ascent. This cooling reduces the ability of the
air to hold moisture and when the dew-point is exceeded precipitation must
result. These facts account for the relatively large precipitation upon the
western slopes of the Coast, Gold, and Selkirk ranges. When vapor is condensed
there is liberated the so-called latent heat, which disappeared when the water
was originally evaporated. It so results that while the air is being cooled by its
own expansion, the consequent condensation of its moisture is supplying it
with heat, and the actual cooling experienced after condensation begins may be
i° for every 300 to 500 feet of ascent. The amount of heat thus supplied to the

' H. M. Ballou: "The Chinook Wind," American Meteorological Journal, IX, 1892-93, pp. 541-547.
1 Fourth Meteorological Report, Washington, 1857, p. 147.

14 GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

ascending currents will obviously depend upon the amount of moisture con-
densed and the rate at which this condensation takes place. The air will reach
the mountain crest at a higher temperature than if no moisture had been
condensed; it is capable of retaining more of its moisture in consequence, to be
carried to the leeward of the mountain range. In being drawn down the lee-
ward slope of the mountain barrier the air is compressed, as it descends, owing
to the greater weight of superincumbent air, and is still further warmed at the
rate of i° F. for every 180 feet of descent. As it is warmed its capacity for hold-
ing moisture is thereby increased and it becomes, relatively, more and more
dry, although it may actually possess considerable moisture. Such a warm,
thirsty wind is our chinook.

The temperature of the air about the peaks and mountain passes differs
less in the winter and spring from that of the valleys and lower plateaus, so that
the chinook is a more conspicuous feature during these seasons. It is also
during these seasons that the cyclonic disturbances are the most pronounced.
In the summer and fall the temperature about the peaks and passes is generally
sufficiently below that of the lower levels so that, although the heating effect,
due to the condensation of vapor and the compression of the air in its descent,
may be actually as great, it becomes much less perceptible. During the most
favorable season for the chinooks we may, theoretically, account for a sudden
rise in temperature of 20° to 25° F., but occasionally it is much greater than this,
sometimes amounting to 50°. Mr. E. B. Garriott mentions a rise of 43° in 15
minutes occurring at Ft. Assiniboine, Montana, Jan. 19, 1892.' In order to
account for such a phenomenon we must postulate a correspondingly high tem-
perature about the crests of the mountains, brought about by excessive and
rapid condensation upon the windward slope, or by some other agency. As
is well known the temperature about the crests of mountains is often considerably
higher than that in the adjoining valley, giving rise to what are known as "in-
versions of temperature. " Since the establishment of the meteorological station
upon Sulphur Mountain, October, 1903, with an elevation of 7,459 feet, there
have been some 300 such inversions noted up to the close of June, 1906. The
following list of some of the most pronounced cases, with dates, is taken from
data kindly supplied by Mr. N. B. Sanson, of the Banff station. The upper
station is 2,917 feet above the lower and i-J miles to the south-southwest. Of
the 206 instances sent, 163, or 79 per cent, of them, were noted in the morning;
26, or 13 per cent., in the evening, and 17, or 8 per cent., at noon. Of this number
101, or 49 per cent., occurred during the winter months, giving the most pro-
nounced cases of temperature inversion. The spring and fall were each repre-
sented by 27 cases, or 13 per cent., while the remaining 51, or 25 per cent., were
noted during the summer months.

' Monthly Weather Review, 1892, p. 23.

GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

TABLE II.

INVERSIONS OF TEMPERATURE, BETWEEN THE BANFF AND SULPHUR

MOUNTAIN STATIONS.

Data supplied by the Canadian Meteorological Service.
(Elevation Banff Station, 4,542 feet; Sulphur Mt., 7,459 feet.)

Date.

Time.

Banff
Station.

Sulphur
Mt. Station.

Difference.

Nov.

I?. I903

6.00 A.M.

— 22. 6° F.

— 14.0° F.

8.6°

Jan.'

5. 1904

6.00 A.M.

- 7.4°

16.0°

23.4°

"

"

6.00 P.M.

- 6.6°

4.0

10.6°

Jan.

17.

6.00 P.M.

— 1.6°

20.0°

21.6°

Mch.

2,

6.00 A.M.

-19.1°

4.0°

23-3°

Aug.

26,

6.OO A.M.

38.5° .

44-0°

5-5°

Sept.

2,

6.00 A.M.

33-8°

40.0°

8.2°

Oct.

25,

6.00 A.M.

26.6°

35-8°

9.2°

Nov.

24.

6.OO A.M.

- 1.4°

6.8°

8.2°

Dec.

24,

6.00 P.M.

- 5-6°

3-5°

9.1°

Jan.

29, 1905

6.00 A.M.

- 7.3°

43-°°

50.3°

Jan.

3°.

6.OO A.M.

-22.2°

1.8°

24.0°

Feb.

3.

6.OO A.M.

- 9-7°

4.2°

13.7°

Feb.

5,

6.00 A.M.

— 10.2°

io.5°

20.7°

Feb.

Hi

6.OO A.M.

—14.0°

36.2°

50.2°

Aug.

23.

6.00 A.M.

30-4°

36.0°

5-2°

Sept.

4,

6.00 A.M.

33-8°

46.0°

12.2°

Nov.

10,

6.OO A.M.

26.6°

32.2°

5-6°

Dec.

i,

6.00 A.M.

- 9-5°

2-3°

11.8°

Dec.

3i.

6.00 A.M.

— 6.2°

4-0°

10.2°

Apr.

22, 1906

6.00 A.M.

24.2°

35-8°

n.6°

June

24, "

6.00 A.M.

35-2°

52.0°

16.8°

Geologically the chinooks must exert an influence, the accumulated effect of
which may be considerable. They remove much snow from the eastern slopes
of the mountains, which might otherwise be available for glacier formation,
raising the lower limit of the snow-line and rendering it more irregular. That
portion of the melted snow which is not evaporated will course down the moun-
tain slopes as water, accomplishing a different work than if it had remained
solid. The alternate periods of thawing and freezing during the winter and
spring will accelerate the disintegration of rock upon the eastern slopes of the
ranges. By this means glaciers lying to the east of the mountain crests will
receive a heavier load of rock d6bris from neighboring cliffs, talus deposits will
be larger than otherwise, and soil formation go on at a more rapid rate. At
Innsbruck, in the Alps, where the foehn winds blow upon an average 42.6
days in the year, the mean annual temperature is raised 1.08° P., or the equiva-
lent effect of one degree of latitude.

c. Oscillations in climate. The problem of precipitation has been discussed
thus far with reference to its geographic distribution over the region, and although
a full discussion cannot be here introduced, there still remains the question of
its distribution in time. Here again our data are far too scant for satisfactory
generalization, but, so far as the evidence goes, it is in harmony with the theory
that precipitation occurs in cycles, there being a series of years in which the total

i6

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

amount is in excess of the normal, although it may fall below for some particular
year of the series. This damp phase of the cycle is followed by a series of
years during which the total amount of precipitated moisture is less than the
normal for this number of years, although it may be in excess for some
particular year. Based upon the meteorological data from 321 stations
distributed over Europe, Asia, Africa, Australia, North and South America,
Bruckner discovered the length of the cycle to be 35.5 years,1 the
dry phase averaging slightly longer than the damp one. At neighboring
stations the cycles may partially overlap and, in the case of coast regions, the
phases may be completely reversed, as recognized by Bruckner. Coincident
with the damp phase there is a general reduction in temperature, an increase
in the level of lakes and rivers, a rise in the ground-water level, a halting or
advance of the front of glaciers. The following table shows at a glance such
data as are available along these lines, compiled from publications of Heim,
Richter, Bruckner, and Hess. The glacial data refer mainly to the Alpine
region. Constructed for any particular region the figures would necessarily
differ more or less from those given. So far as the United States is concerned
the dry phase through which we have just passed appears to have closed and
we are entering upon another series of damp years. In the Canadian region
under study the damp phase seems to have started about three or four years
earlier than in the Great Lake region and the preceding dry phase about as long

before.

TABLE III.

PERIODIC OSCILLATIONS OF CLIMATE, WITH THEIR EFFECTS UPON LAKE LEVELS

AND ALPINE GLACIERS.

PRECIPITATION.

TEMPERATURB.

LAKE LEVELS.

GLACIERS.

Damp.

Dry.

Cool.

Warm.

High.

Low.

Advancing.

Retreating.

1591-1600

1591—1600

1595-1610

1601-1610

1611—1633

1611-1635

1630

1635-1645

1646-1665

1645-1665

1677-1681

1665—1690

1691-1715

1691-1705

1710—1716

1716-1735 |

1706-1735

1720

I736-I7SS

i73I~I745

1740

1735

1756-1770

1746-1755

1760

1750-1767

1771-1780

1756-1790

1777-1780 I

1760-1786

1781-1805

1791-1805

1798-1800

1800—1812

1806-1825

1806-1820

1820

1811-1822

1826—1840

1821-1835

1835

1822-1844

1841-1855

1836-1850

1850

1840-1855

1856-1870

1851-1870

1865

1855-1875

1871-1885

1871-1885

1880

1875-1893

1886-

1886-

1892

1894-

Klima-Schwankungen sett 1700, Edward Bruckner, Wien, 1890, pp. 133-193.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 1 7

Numerous factors conspire to prevent the movements of glaciers from
being exactly coincident with the corresponding climatic phases. For the
Swiss glaciers Heim found that, ordinarily, an advance began from 3 to 6 years
after the opening of a damp-cool phase and reached its maximum in 4 to 10
years, but in the case of the longest glaciers the maximum position might not
be reached until the close of the cycle itself. These facts partially explain the
anomalous behavior often noticed in neighboring glaciers. When the Illecille-
waet Glacier was first visited by the Messrs. Vaux, in 1887, it was standing
close up against a small moraine, which it had just formed, or more probably
assisted in forming during this period of halt. Since 1887 this glacier has been
in constant retreat at an average rate of 33.2 feet per annum. In 1905 the
retreat was found to have been reduced to 2 feet and the inference is that the
glacier is preparing to advance. So far as we may judge from a study of this
one glacier, the best known of the series, a damp-cool phase of the precipitation
cycle closed in the early 8o's, and was followed by a dry- warm phase, which lasted
for 1 6 to 1 8 years, ending somewhere near the close of the century. The follow-
ing quotation from Dawson furnishes confirmatory evidence of the existence
of the preceding wet phase of the cycle, which was itself preceded by a dry phase. l

"Evidence of a remarkable character has been found, which tends to show
that a somewhat rapid increase in the total annual precipitation, has taken
place during late years, and deserves to be recorded here. The evidence referred
to is that afforded by the abnormal height of small lakes, without outlets, occur-
ring in regions characterized by moraine hills. These serve as natural gauges,
but instead of measuring the actual rainfall give a result, dependent on this and
the counteracting effect of evaporation. The abnormal character of the rise
of the water in these lakes is shown by the facts that it has killed a belt of trees,
some of large size, and at least fifty years in age, along parts of the margins of
some of these lakelets. Both the Douglas fir and the yellow pine — the latter,
never naturally growing even in damp soil, — have been found in numbers thus
killed. The condition of the trees shows that they have been killed within a
few years, and their size indicates that the waters of the lakes in question have
not been for any considerable time during a period of 5 o years or more, at the present
high level. These observations were made in both 1883 and 1884. The lakelets
observed to be so affected were numerous and scattered over a belt of country
along the western part of the range [ Rockies] for a length of about 140 miles. "

Looking to the records of the Canadian Meteorological Service for still
further evidence of the periodicity of the climate of the region, we find that only
three of the stations have records sufficiently continuous and reaching far enough
back to be of help. These stations are Agassiz, Banff, and Calgary and their
averages are more nearly the real, but still unknown, normal. Average annual
temperatures and precipitation for the mountain stations, based upon observa-
tions made between 1880 and 1897, will certainly be found later to be too high
for the temperature and too low for the precipitation, while those averages
based upon observations taken since 1897 will prove to yield too high a pre-
cipitation and too low a temperature. In table iv we place side by side the

1 Geological Survey of Canada for 1885, p. 32 B.

18

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

precipitation data for the above three stations, so far as such data are obtainable.
In the first column for each station is given the total precipitation, in inches,
the snowfall being reduced to rain upon the supposition that 10 inches of snow
are equal to one of rain. In-the second column is given, for each year, the actual
deficiency, or excess, when the amount is compared with the normal, or average
for the entire series of years. In the third column the actual precipitation
appears as a percentage of the normal, while in the fourth there is given the
accumulated excess, or deficiency, to or from the beginning of the year 1897.
So far as we may be permitted to draw conclusions from all available data,
the break between the damp and the dry phases of the precipitation cycle in
this region occurred about 1897 in the Rockies and 1898 in the Selkirks and we
may venture to predict that for another decade the precipitation will be in excess
of the true normal for the various mountain stations. An examination of the
Agassiz data, from the lower Frazer Valley, shows that while Banff and Calgary
were deficient, this station was accumulating an excess and that since 1897
there has been a marked deficiency. The inference is that we have here an
example of one of Bruckner's exceptional coast regions, in which, although the
precipitation is also in cycles, the crests of the great waves correspond with
the troughs over the general surface of the earth. It is very unfortunate that
the data from Vancouver, Nanaimo, and Victoria are not full enough to show
whether or not they are included in this region.

TABLE IV.
PRECIPITATION DATA, BY YEARS, FOR CALGARY, BANFF. AND AGASSIZ.

CALGARY.

BANFF.

AGASSIZ.

Year.

Total
precipitation.

O O

fl
1-3

Accumulated ex-
cess or deficiency.

Total
precipitation.

Excess or
deficiency.

H

fjj

Accumulated ex-
cess or deficiency.

Total
precipitation.

Excess or
deficiency.

}i
(£"8

Accumulated ex-
cess or deficiency.

1885

12.91

—3-73

78

.8

.g

1886

11.32

-5-3«

68

8

.£'

1887

15.69

—°-95

94

<g

8

1888

17.41

+ 0.77

105

£

&

»

1889

"•59

—5-°5

70

5

1

I

1890

iS-47

—1.17

93

1

'9-54

— 0.60

97

1
u

56.43

—5-59

91

$

1891

"O

13

16.48

-3-66

82

•0

T)
V

1892

7.91

-8-73

47

1

I

67.78

+ 5-76

109

3

1893

11.05

-5-59

66

3

10.88

—9.26

54

3

+ T4-93

124

e

1894

11.63

—5.01

70

3

1

78.01

+ 15-99

126

I

1895

16.00

— 0.64

96

<

«i

54-5°

—7-52

88

1896

16.05

—0-59

96

15-86

—4.28

79

68.25

+ 6.23

no

1897

20.58

+ 3-94

124

a

23.40

+ 3-26

116

r"°

63-43

+ i .41

102

1898

te

20.58

+ 0.44

IO2

.3

w

So-31

-11.71

81

3

1899

26.15

+ 9-51

'57

o>

26.34

+ 6.20

J3I

x.

g

1900

17-57

+ 0.93

106

CO

23-29

+3-15

116

»

72.00

+ 9.98

116

CJ

1901

22.31

+ 5-67

134

8

19.27

—0.87

96

M

52.98

—9.04

85

&

1902

34-57

+ 17-93

208

0

3°-S9

+ 10.45

152

V

•8

54-68

-7-34

88

0)

1903

22.77

+ 6.13

137

%

24.82

+ 4-68

123

E

57-74

—4.28

93

1

1904

10.82

-5-82

65

3

14.80

-5-34

73

|

54.67

-7-35

88

1

1905

14-32

—2.32

86

|

16.00

—4.14

79

|

60.59

—1-43

98

Normal

16.64

20.14

-

62.02

^

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE III.

Mitre
9,470

Abbots Pass
9388 ft.

Lefroy
11,320

1000

Scale
1000 2000

Little ;

Mitre

3000 4000 feet

Map of Victoria Glacier, Lake Louise Valley, Canadian Rockies. Surveyed and drawn by W. H. Sherzer, July, 1904. Field assistants De
Forrest Ross and Frederick Larmour. Adjacent regions based upon maps of A. O. Wheeler and I). W. Wilcox.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 1 9

CHAPTER II.

VICTORIA GLACIER.

i. GENERAL CHARACTERISTICS.

THE VICTORIA GLACIER originates at Abbot's Pass, upon the crest of the
Great Continental Divide which separates British Columbia from Alberta,
flows due north for a mile between the precipitous walls of Mts. Victoria and
Lefroy, makes an abrupt turn to the northeast and pursues a straight course
for another two miles before wasting away in the Lake Louise Valley. The
collecting area at the Pass is much restricted and narrows down to 600 to 700
feet, where the rocky cliffs upon either side begin to frown at each other from
beneath the snow. These cliffs become higher and separate gradually, allowing
the glacier to broaden to about one-third of a mile as it approaches the bend in
its course. Rounding the nose of Mt. Lefroy the glacier receives its double
tributary from the southeast, attains its greatest breadth of one-half mile,
and for the last mile narrows regularly to its debris-covered nose. In this
lower third of its length it lies between Mt. Aberdeen (elevation 10,340 feet)
and Castle Crags upon the east and Mt. Whyte (9,776 feet) upon the west. A
general view of the lower two-thirds of the glacier, with the tributary entire,
is shown in plate iv, figure i, while plate iv, figure 2, shows the upper portion
and gives a lengthwise view of the tributary.

A Watkin mountain aneroid was carried to the crest of the Pass, July 22,
1904, and gave an elevation, when corrected, of 9,370 feet above sea-level. The
more accurate methods of the Canadian Topographic Survey gave Wheeler
9,540 feet elevation for this same Pass, from which the descent through the
so-called "Death -Trap" is rapid, requiring the cutting of steps in the snow
when it is hard from freezing. Owing to the north-south direction of this part
of the valley and the height of the bounding cliffs, the sun has little direct
power and the glacier is permanently covered with snow which assumes a granu-
lar form, owing to the partial melting of the flakes, and constitutes the ne"ve".
This condition of the snow causes it to resemble granular tapioca and may be
seen in the snows which, in more southern latitudes, linger until late in the
spring. The ne"ve"-covered portion of the Victoria reaches an altitude of about
7,500 feet, or about 2,000 feet below the Pass, when, as the glacier turns to the
northeast, the ice makes its appearance through the snow. The line of separa-
tion between ice and snow, as seen at the surface, is irregular and uneven,
shifting with the season and from year to year. Winters of scanty snowfall,
followed by bright warm summers, will send the neVe" line up the glacier; while
winters of heavy snowfall and cool summers will cause this line to move toward
the nose. Plate v, figure 2, shows the upper third of the glacier, leading to the
Pass through the "trap, " and the ne"ve" line in the foreground, as it appeared in
July, 1904. Rounding Lefroy, the glacier descends rather abruptly some 400
to 500 feet, owing apparently to a sudden change in the inclination of its bed,

20 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

giving rise to a series of transverse crevasses. The lower one and one-half
miles of the Victoria presents a remarkably even surface slope, suggestive of a
correspondingly even bed, so that it may be ascended with entire safety by the
most inexperienced. The nose reaches an altitude of about 6,000 feet, so that
the average slope is at the rate of about 1,200 feet to the mile. For the lower
half the surface slope is but 650 to 700 feet to the mile, or at the average rate
of 7° to 8°.

In the neVe" region the surface of the glacier is concave, owing to the accu-
mulation of snow along the base of the cliffs, this being permitted by the
relatively small amount of heat radiation and reflection. For a short distance
opposite Lefroy the cross sections of the ice stream have a horizontal surface
line, while in the lower portion the. surface is flatly convex, owing to the lateral
melting caused by the rock walls and, to a greater or less extent, by the lateral
drainage streams. • (See the cross-section of the Victoria along the line of the
steel -plates, page 30.) It does not seem probable that the glacier attains
any considerable thickness, the thickest portion, apparently, lying opposite the
tributary where the ice may be 500 to 600 feet in depth. The nose is rounded,
completely veneered by rock debris so as to conceal the ice, and perched up
above the valley floor upon an old moraine which it has partially overridden,'
but has not had the strength to push aside (plate v, figure i). The front here
is steep, the angle being about 38°, and about 90 feet in height, with a series of
gradually rising crests from the medial and right lateral moraines. Back from
this nose some 2,000 feet, there is exposed a steep ice wall, 35° to 50°, which
attains a height of 125 feet, and continues for about 800 feet. This now appears
as the side of the glacier, but the position and form of the older moraines show
that the front has been gradually swinging around into this oblique position,
the cause of which is apparent from an inspection of the map. The eastern
side of the glacier is much better protected by the right lateral moraine and by
the much broader and closely placed medial. This portion of the front is thus
prevented from melting, while the less well protected western half has been
rapidly receding. Between this oblique ice wall and the real nose other smaller
faces are developing and being enlarged with each season's melting.

2. NOURISHMENT.

The main glacier is nourished in four ways, which may be separately recog-
nized, as follows:

a. By the moisture directly precipitated into the valley between Mts.
Victoria and Lefroy. The great bulk of this is in the form of snow, which
probably amounts to about 25 feet per annum. The lesser amount in the form
of hail, rain, dew, fog, and frost would also contribute to the substance of the
glacier.

b. The funnel-like form of the valley, with its opening to the north, enables
it to catch and retain large quantities of snow drifted southward by the north
winds,- as well as that which collects in the lee of Victoria when a west wind is

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE-SHERZER.
Mitre. Lefroy.

PLATE IV.

Victoria.

FIG. I. — General view of Victoria Glacier looking southward.

Copyrighted, 1902, by the Detroit Photographic Co.
Mitre. Lefroy. Abbot's Pass.

Victoria.

FIG. 2.— Lefroy Tributary.

Victoria Glacier.

Copyrighted, 1902, by the Detroit Photographic Co.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

Aberdeen.

PLATE V.

Lefroy.

Lefioy.

FIG. i. — Debris-covered nose of the Victoria Glacier. July, 1904.
Abbot's Pass.

Victoria.

2. — Neve lield of Victoria Glacier, looking southward through "death trap." July, 1904.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE VI.

FIG. r. — Pafh of an avalanche along Victoria neve. Photographed by De Forrest Ross, July,
1904. Note ice levees along the margin of the path.

FIG. 2. — Hanging glacier upon Mt. Victoria as seen from summit of Mt. Aberdeen (elevation 10,340 feet). Photo-
graphed, 1903, by Arthur O. Wheeler.

DIVERSITY

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 21

blowing, or in the lee of Lefroy with an east wind. Thus because of its shape,
position, and depth the upper valley is able to capture much more snow than it
would ordinarily be entitled to.

c. Upon the opposing faces of Victoria (elevation 11,355 feet) and Lefroy
(11,220 feet) large beds of snow accumulate during the fall, winter, and early
spring. During the late spring and early summer much of this snow is pre-
cipitated into the valley as avalanches, with the roar of thunder and the blast
of a tornado. It is the danger from this source that has suggested the name
" Death -Trap, " for the narrower portion of the valley, although no fatalities
have yet occurred here. These avalanches may shoot directly across the
valley, or they may turn and move along it lengthwise as shown in plate vi,
figure i . They must bring down numerous rock fragments, which are distributed
completely across this portion of the glacier and incorporated into its ne"ve".

d. A considerable portion of the snow which clings to the eastern shoulder
of Mt. Victoria is compacted into stratified ice, and over an area of about one
square mile there arises a true glacier perched up on the mountainside, with a
slope that appears too steep to give it a foothold. Such a glacier is known
as a "cliff glacier," or "hanging glacier" (plate vi, figure 2). It moves down
the slope, probably with considerable velocity, and is avalanched into the
valley upon the back of the Victoria Glacier, filling the air with ice dust. At
the crest of the precipice the ice has an estimated thickness of 200 to 300 feet,
from which great blocks, sometimes as large as a city square of buildings, are
detached and fall vertically 1,200 to 1,500 feet. At certain places where the
falls are more frequent there are built up debris cones of coarse granules upon
the western margin of the glacier. This avalanched ice spreads over the glacier
and is incorporated into its body along with more or less ground-morainic material
which has been manufactured between the hanging glacier and its bed. When
the weather is cool and cloudy these ice avalanches are infrequent, but upon
a warm bright day, with much melting and more rapid forward movement of
the ice, they occur every few minutes from some portion of the long front. On
August 25, 1903, during the mid-portion of the day avalanches were noted as
follows:

10:39 A-M- moderate. 12:29 P.M. moderate.

10:43 slight. 12:49 moderate.

11:05 slight. 12:52 slight.

11:27 slight. 1:11 moderate.

11:28 heavy. 1:18 moderate to heavy.

11:40 slight. 1:35 moderate.

12:20 P.M. slight.

12:27 moderate. 1:52 moderate.

2:10 moderate.

Owing to its western and southern exposure the opposite face of Lefroy
does not support a hanging glacier, although there is a suitable collecting area

22 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

for the snow. That which is not precipitated into the valley as avalanches
melts away in the course of the summer and this water, along with that from
Mt. Victoria, forms slender cascades, which are partially absorbed by the
ne"ve" and, in part, work their way to the bed of the valley and are incorporated
into the subglacial drainage. In these various ways the upper Victoria receives
the precipitation from about two square miles of collecting area. Through
pressure, rain, and surface melting, due to the intense solar action, as well as
the chinook winds, this mass of granular snow is compacted into a very fine
granular ice. Powerful winds sweep over the bare peaks and ridges and spread
over the neVe" more or less fine matter, organic as well as inorganic. This
material is concentrated at the surface, when sufficient melting has taken place,
and gives a rather sharp line of demarcation between the older snow and that
which falls freshly upon it. In consequence of the melting and the presence
of the foreign matter, the neVe* acquires a characteristic stratification, which in
the case of the Victoria persists to its nose. Owing to the avalanches of snow
and ice from Mts. Victoria and Lefroy these strata are rendered more or less
irregular. The great weight of this snow and underlying ice forces the entire
mass valleyward and thus prevents indefinite accumulation.

3. DOUBLE TRIBUTARY.

a. Mitre Glacier; the host. Upon either side of the small peak known as the
Mitre (elevation 9,470 feet), there descend two steep snow slopes, which give rise
to two ne"ve"-covered streams of ice (see plate vn, figure 2). That to the right is
intersected by a great crevasse, caused by the glacier drawing away from the
snow and ice which adhere to the rocky slope, and forming what is known as
the " bergschrund. " This schrund renders this stream impracticable, but the
other may be safely ascended with a guide, to the Mitre Pass leading over
into Paradise Valley. Here from an elevation of 8,480 feet the descent is
very rapid for about 1,200 feet, when the two streams unite into a single glacier,
for which the name Mitre, first proposed by Allen for the entire tributary, may
best be retained. For a very short distance the glacier is permanently covered
with ne"v6, but in midsummer this soon disappears and discloses a very weak and
poorly defined medial moraine (plate vn, figure 2). It flows lazily down
the straight valley, one and one-third miles, between Mts. Lefroy and Aberdeen,
attains an average width of one-third to one-half a mile, and joins the Victoria
with a breadth of 3,200 feet.

b. Lefroy Glacier; the parasite. Upon the eastern and northern slope of
Mt. Lefroy, because of its exposure and other favorable conditions, there has
arisen another hanging glacier similar to although smaller than that just de-
scribed upon Mt. Victoria. Plate iv and plate vn, figure i, give views of this
elevated glacier, clinging to the steep mountain slope, the latter view being
taken from the summit of Mt. Aberdeen, looking westward and from an elevation
of 10.340 feet. From the steep, vertical ice face great blocks are avalanched
2,000 feet into the valley, much of the ice being ground into dust and shot

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE VII.

FIG. I. — Hanging glacier upon Mt. Lefroy as seen from summit of Mt. Aberdeen (elevation 10,340 feet). Photo-
graphed, 1903, by Arthur O. Wheeler.

Aberdeen.

Mitre Pass.

Mitre.

Litde Mitr

Lelroy.

Kio. 2. — Double neve field of Mitre Glacier, July, 1904. Faulted neve strata are seen in left tributary and a bergschrund in the right.

In foreground a snow-filled crevasse; beyond which lies the neve line.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 23

out beyond the base of the cliff. Upon the western or Lefroy side of the upper
half of the Mitre Glacier, there is thus heaped up a mass of pulverized ice, which
is compacted by freezing into strata. This mass constitutes a new glacier,
since the structure of the original one must have been destroyed by the plunge,
with the exception of the granules and their fragments. Such a glacier is said
to be "reconstructed," or "regenerated." Furthermore, this new glacier
rests upon the back of the Mitre; is nourished differently; is of a different form;
has its own distinct set of strata, unconformable with those of the Mitre; moves
across the valley instead of lengthwise of it, and is accomplishing a totally
different geological work. This glacier, for which the term Lefroy is appro-
priate, is one of the best known examples of what Forbes termed a "parasitic
glacier,"1 far better, indeed, than the type itself. It is parasitic in the sense
that it is carried by its host, the Mitre, and in that it is nourished entirely by
snow and ice which would be otherwise available for the host.

Just what is the structural relation of the Lefroy to the Mitre, the parasite
to its host, can only be conjectured, since the contact was not observed and the
plane of separation may not be at all distinct. There is, however, a very
evident, deep-seated motion down the valley and an equally evident
superficial motion across from the base of Lefroy to the foot of Aber-
deen. The result of the latter motion is to carry most of the ground-morainic
material, such as clay, sand, bruised and scratched pebbles and boulders, which
has been manufactured beneath the hanging glacier of Lefroy, entirely across
the valley and dump it in a great heap upon the eastern or Aberdeen side
of the Mitre (plate vin, figure i, and plate xv, figure i). The front of the
parasitic Lefroy being parallel with the side of the Mitre, some of the ground-
morainic deposit is arranged in ridges parallel with the side of the latter, in
which form it is being dealt out to the Victoria. Until the above stated relations
were discovered it was a serious puzzle to ascertain how a glacier could get its
ground moraine upon its own back and arrange it in ridges parallel with its side
(see plate xv, figure i ) . That the material could not have come from Mt. Aberdeen
was evident from the fact that it does not support a hanging glacier upon its
western face, as shown in plate vui, figure 2, although there is a buried mass of
stagnant ice upon the northern shoulder. Avalanches of snow and the ordinary
processes of weathering have brought down considerable angular material from
Aberdeen which covers most of the ground-morainic deposit from the opposite
side of the valley. While this ground-morainic material is being moved east-
northeast by the Lefroy for a distance of i ,800 to i ,900 feet, it is also being carried
north -north west for a distance of about 3,800 feet by the underlying Mitre and
the resultant motion is somewhat east of north. This will be made clear from
an inspection of the map, plate in.

Opposite the large debris cone seen in plate iv, figure 2, upon the western
side of the Lefroy Glacier, there is a marked depression in the surface of the
ice and also across the entire tributary where it joins the Victoria, giving good

1 Travels through the Alfs of Savoy, James D. Forbes, Edinburgh, 1845, p. 201.

24 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

exposures of the outcropping edges of the strata, to be noted later. These
depressions furnish confirmatory evidence that the general movement of the
superficial layers is across and not parallel with the valley. The position of the
left lateral moraine from the debris cone, above noted, to the Victoria, shows,
however, that along the base of the Lefroy cliff the movement is normal and
due to the Mitre, although the strata belong to the Lefroy. In consequence
of this a relatively small amount of morainic material is captured from the
Lefroy and delivered to the medial moraine of the Victoria at the nose of Mt.
Lefroy. This double tributary joins the Victoria at an elevation of about
6,670 feet, having an average surface slope of 1,360 feet to the mile, and is at
once compressed to about 600 feet, as compared with 3,200, or as 5^- is to i.
There being no corresponding increase in the height of the ice, the inference
is that the tributary delivers relatively little ice to the Victoria and that its
movement is correspondingly slow.

4. DRAINAGE.

a. Surface ablation. The drainage supply of the glacier originates from
the rainfall over the glacier and adjacent mountain slopes, from the melting of
the snow and ice in the region of the hanging glaciers, and from the general
melting of the glacier itself and its tributaries. No definite data are available
concerning the rainfall over the glacier and adjacent slopes. Owing to its greater
altitude it would be much less than at Field and Banff and would practically
all fall during June, July, August, and September. After heavy showers the
streams from the mountain slopes are in many cases highly charged with
sediment, those originating from the melting of snow being generally clear.

The temperature of the ice during the summer was found to be either
just at the freezing point, or so near it that any addition of heat was
sufficient to start the process of melting. In the abandoned drainage tunnel,
to which reference will be made later, holes were bored into the ice wall, 140
feet back from the entrance, and a standard minimum thermometer inserted
its full length. Owing to the course of the tunnel the point of observation
was estimated to be 70 feet from the foot of the oblique ice wall and about 17
feet from the actual ice face (plate vin, figure 4). During the week from
July 31 to August 7, 1904, the readings were 31.8° F., 31.6°, 31.8°, 31.9°, 31.7°,
and 32°. The maximum temperature of the air in the tunnel during the week
ranged from 31.4° to 33.0° F. Owing to the warmth of the body and that of
the candle used, it was found impracticable to get the temperature of the air
at the same time that the temperature of the ice was taken.

In the rarified atmosphere at these high altitudes the midsummer sun
strikes with surprising force and the surface ice, so near its melting point, is at
once converted into water without changing its temperature. In the case of
scores of observations made upon the surface streams of the series of glaciers
the temperature was almost uniformly 32° F. In rare cases it was found to be
a small fraction of a degree above. In order to secure some data for an estimate

o
£

>
'B

O

in

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 25

of the rate at which surface ablation was taking place over the lower portion
of the Victoria, accurate elevations were taken upon the series of steel plates
used in determining the forward movement. The results are shown in table
V, page 31, giving the total melting from July 13 to August 4, 1904, for a
period of 22 days of midsummer. The maximum lowering of the ice surface
occurred at plate No. 13, nearly two-thirds of the way across when measured
from the west side, and amounted to 3 . 8 feet, or a daily average of 2.076 inches.
This plate was located on a portion of the glacier least protected by rock d6bris.
Although this lowering of the surface is due, in the main, to the sun's action and
general effect of the atmosphere, usually above the freezing point in summer,
there are other minor agencies which would tend to give the same result. One of
these is the rain, 1.506 inches of which fell during the period of the observations.
Other agencies are subglacial melting and subglacial erosion, longitudinal
stretching, or a lateral spreading of the ice, all of which would tend to lower
the upper surface. It should be noted, further, that in the 22 days this plate
moved forward 60 inches and with the average surface slope of 7° to 8°, the
plate would be lowered by this agency alone about one-third of an inch daily.
Making this correction the actual ablation, from sun, atmosphere, and rain,
would amount to about 1.74 inches daily and for the two main months of July
and August would give a total of about 9 feet. Observations upon the lower
Lefroy showed that the ice surrounding certain morainic heaps had been
lowered during the season by about this amount. No glacial tables of this
height can be found, however, because of the undercutting effect of the sun's
rays and the consequent destruction of their pedestals. The broad medial
depression lying just west of the medial moraine (see plate iv, figure 2, and cross
section page 30) has been produced by the greater surface melting and this
has been permitted by the thinner covering of rock debris, the ice of this portion
of the glacier coming from the Lefroy side of the upper Victoria. This depres-
sion continues down the glacier for 2,200 feet, where it thins out, apparently
from surface melting. With a forward motion here of about 64.5 feet annually,
it would require 34 years for the ice to move from the line of plates to the oblique
ice face, during which time some 306 feet of ice might be melted away. If
the rate of melting and rate of forward movement remained constant, this
number would represent the approximate thickness of the ice beneath plate 13.
It is very probable that the rate of melting becomes less, owing to the concen-
tration of rock debris at the surface, but it is also very probable that the rate
of forward movement becomes also less as the ice diminishes in thickness.
The work with the spirit level indicated that this plate was originally 393 feet
above the lower margin of the ice in this depression, the difference of 87 feet
representing the rise of the valley floor in this distance. If this latter figure is
approximately correct the inclination of the bed is much less than that of the
surface of the ice itself

b. Surface drainage. Over the entire neVe" area the water from the melting
snow, as well as that from the rainfall, is absorbed into the body of the glacier

26 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

and refrozen, binding the granules together into an ice conglomerate. Where
the ice itself is exposed and crevasses absent, the melted ice and rainfall are
concentrated into more or less well denned channels, which persist from season
to season. In the early morning the glacier is impressively quiet, the ice is dry,
and many of the small pools are frozen over with a thin layer of ice. When the
summer sun enters the valley the exposed ice becomes moist, small trickles of
water unite into rills, that grow larger and larger from the union of innumerable
others, and these form still larger brooks which empty their clear, ice-cold waters
into the main drainage channels and, after a day of rapid melting, we have here
roaring torrents. These streams slowly cut their way into the solid ice by
mechanical erosion, assisted by the rock fragments which the water is able to
move along, and by melting. The water being apparently at the melting point
of the ice, 32° F., it is incapable of imparting heat to that over which it flows
and the melting must arise from the conversion of its kinetic energy into heat.
Such heat would be imparted to the ice, rendered latent in the process of lique-
faction and the temperature of the water would not be sensibly raised. The
question arose in the mind of the writer while studying these ice streams
whether water at 32° is capable of dissolving ice at the same temperature, as
it might dissolve rock salt over which it was flowing. So far as he has been able
to learn the question has not been investigated, but if water does have any such
effect upon ice under these conditions, it would help to explain the formation
of these ice channels.

In the upper part of their courses these stream beds are generally free from
debris and quite straight, but as the bed is broadened, boulders, too large for
the stream to handle, slide in from the surface and the stream is compelled to
go around. In this way a system of meanders is formed, as shown in plate
vin, figure 3, and the ice banks are rendered steep and, here and there, undercut
by the rushing water. In the lower portions of the course the bed may contain
considerable rock debris, but this has simply slid down from the surface and
nowhere suggests an aggrading action of the stream. When the stream channel
is contracted for any reason, the level of the water is raised, its velocity increased
in consequence, and an ice basin cut out upon the down-stream side, filled with
more quiet water. This is suggestive of the manner in which the Lake Louise
rock basin, to be later described, may have originated when the entire valley
was ice-filled.

In portions of the glacier intersected by crevasses it is obvious that surface
streams, of any considerable size, cannot develop. The water escapes by an
englacial or subglacial tunnel, to reappear at or near the nose. When a stream
encounters such a crevasse, from which there is drainage beneath, it forms a
small cascade and begins to cut a channel in the vertical face of the crevasse
wall. If the velocity and volume of the water are sufficient, a corresponding
channel may be produced in the opposite wall. As the lips of the crevasse
are subsequently brought together by movements of the ice body, and the
crevasse is healed, this small vertical channel persists and still furnishes an

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE IX.

FIG. 2. — Marginal lakelet west side Victoria Glacier.

FIG. i. — First stage in formation of a inoulin, Lefroy Glacier.

— Inner end abandoned drainage tunnel, Victoria Glacier, July, 1904

FIG. 4.— Mouth abandoned drainage tunnel, Victoria Glacier, looking outward. July, 1904.

if UN;

GLACIERS OF THE CANADIAN ROCKIES AND, SELKIRKS. 27

•

escape for the surface stream. This is well shown in plate ix, figure i, the small
stream entering the nearly healed crevasse from the right. As the ice moves
valleyward the drainage area of the stream may be enlarged, the amount of
water correspondingly increased, and our glacial well, or "moulin, " may be
indefinitely enlarged both in diameter and depth. Since the water is introduced
from above and escapes below, they are more like wells, turned wrong-side up.
They may be found in various stages of development, the younger in the region
of the open crevasses, the mature examples in the lower course of the glacier,
where they have been carried by the general forward movement of the ice,
and into which the surface torrent plunges with a sullen roar to unseen
depths.

The main drainage stream of the Victoria Glacier starts near the nose of
Mt. Lefroy, to the west of the medial moraine, and passes somewhat obliquely
downward to a moulin, opposite the oblique ice face (plate iv, figure i, and
plate in). This stream drains that portion of the glacier over which the
surface melting is the greatest. A second drainage stream originates near the
above, but in the depression between the medial moraine and the Lefroy tribu-
tary. This stream collects the surface waters from the tributary and extends
for one-quarter mile down the deep depression between the medial and right
lateral moraines and disappears in a system of crevasses that cut this portion of
the glacier.

Two short, but rather deep, drainage channels occur upon the western side
of the glacier, lying upon the inner side of the left lateral moraine. One of
these has incised the ice to a depth of 18 to 20 feet. Since these streams have
probably occupied their present sites for many years it is rather remarkable that
they have not completely cut through the ice to the rocky bed beneath. From '
the fact that they have not done so we are forced to conclude, either that the
rate of cutting is surprisingly slow, or else that the glacier thickens in such a
way that the bottoms of the streams are elevated with reference to the base
of the glacier. It is possible that the longitudinal compression to which the
glacier is subjected in its lower half may be sufficient to secure this result.

c. Marginal drainage. Upon the eastern side of the lower Victoria, along
the base of Mt. Aberdeen and Castle Crags, there is no visible marginal drainage
at present, but water may be heard trickling amongst the rocky de"bris. Just
at the head of this depression, however, where the tributary joins the main
Victoria, there is evidence of earlier drainage here, in the form of an abandoned
lake bed, with a length of 500 to 600 feet. A small gravel delta was formed
at the head, while the rest of the bed is filled with silt. A still smaller lakelet
existed for a short time at the nose of Mt. Lefroy, in the depression between
this side of the tributary and the Victoria. Upon the western side of the glacier
there occurs a similar marginal lakelet, between the glacier and Mt. Whyte,
fed by a mountain stream and a discharge stream from the glacier itself (plate ix,
figure 2), which cascades over the lateral moraine from a dozen different places.
The lakelet has an elevation of 6,554 feet, is largely filled with fine silt, and

28 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

contains a number of low islands, supporting some vegetation. The outlet
stream is marginal for about 500 feet and then enters the side of the glacier.

d. General drainage brook. The subglacial and englacial streams are all
united into a single stream which emerges from the base of the ice at an elevation
of 6,127 feet and back about 1,000 feet from the real nose, upon the west side.
This stream cascades over the coarse blocks of the terminal moraine, receives
a small tributary from Mt. Whyte, and rushes headlong to the lake, one mile
distant, dropping some 450 feet. In comparatively recent time the dis-
charge was through a tunnel which is being rapidly destroyed by melting,
plate vin, figure 4, shows the entrance to this tunnel as it appeared in 1904 and
plate ix, figure 4, furnishes an interior view looking out and down the valley.
At this time the opening was 1 2 feet high by 7 feet broad and the tunnel could
be entered for a distance of 160 feet, when it appeared to have been clogged by
ground-morainic material. The opening narrowed rapidly toward the inner
end (plate ix, figure 3) and the severe melting of 1905 showed that it connected
with an englacial passage leading towards the main moulin. This difference
in the size indicates that when the ceiling of the portion of the tunnel shown in
plate ix, figure 4, was being fluted by torrential waters, the bed of the stream
was at a considerably higher level than that there shown and that the stream
worked its way down from an englacial to a subglacial position. The amount
of water discharged through this tunnel was probably no greater than that
at" the present time through the present exit. The floor of the tunnel, at its
entrance in 1904, had an elevation of 6,192 feet, or 65 feet above the present
place of discharge.

During periods of minimum melting, and always in the early morning,
the amount of water discharged from the glacier is relatively small. Late
in the afternoon and evening of a day of rapid melting it gushes forth with
great power and volume (plate x, figure i). It was impracticable to
measure the amount of discharge at this exit, but measurements were made
at the delta where the stream enters the lake. So little additional water was
being received at the time from the adjacent mountain slopes that the results
secured represented approximately the flow from the glacier itself. An accurate
cross-section of the stream was secured by taking the level of the bed for each
foot, establishing a gauge, and determining the velocity from surface floats.
A calculation of the flow was made, after a week of minimum melting, by averag-
ing the flow at 9:00 A.M. and at 6:00 P.M. During this period there were 0.671
inch of rainfall near the nose of the glacier. A . similar determination was
made after a week of maximum melting, with 0.030 inch of rainfall. These
results gave 73 and 93 cubic feet per second for the average flow. At the
time of the minimum flow the water at the exit from the glacier was found to
possess 0.230 oz. of sediment to the cubic foot and 0.506 oz. during the time of
maximum discharge. This is enough to make the water decidedly turbid.
The total amount of sediment carried out daily during the maximum discharge
period was estimated to be about six tons, and one-third this amount for the

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 2Q

minimum period of flow. During the spring and fall the flow is very greatly
reduced and must be very scant, or nothing, in the winter. Mr. Robert Camp-
bell informs me that he has seen water in the stream, however, beneath the
snow and ice of winter.

e. Water temperatures. The temperature of the water at the main exit
was found to vary from 32.0° F. to 32.4,° at various times of the day. Near the
site of the camp, just outside of the older of the two great block moraines, and
some 2,000 feet from the exit, a series of observations was made upon the temper-
ature of the water in the west branch of the drainage brook. The observations
were made between July 2 and 27, 1904, and in the early morning, near midday,
and in the evening, but not at any stated hours. Most of them were taken
between 7 and 8 A.M.; 12 and 2 P.M., and 8 and 9 P.M. Of 56 observations,
those, for the morning averaged 35.4°, for midday 41.4°, and for the evening
35.2°. A small amount of drainage was received from Mt. Whyte, which must
have materially affected the temperatures. Upon July 18, 1904, simultaneous
observations were made at the glacier, camp, and at the delta, one mile below.
The maximum temperature for the day was 51.7° F. and the minimum 34.8°.
The results were as follows:

Time. Glacier. Camp. Delta.

9:00 A.M. 32.4° 35.2° 35.8°

6:po P.M. 32.4° 34-o° 36.0°

By the time the water has moved across Lake Louise to the foot, its tem-
perature has been raised some 6° to 12°, the temperature ranging from 42° to
48°, during the summer months. The temperature of the Bow River at Laggan,
into which Lake Louise is drained, was found to be 54.9° F., Aug. 15, 1904.

5. FORWARD MOVEMENT.

o. Measurements. Long before the attention of scientists was directed
to glaciers as suitable subjects for investigation, the Swiss peasants had dis-
covered that they possessed a forward, down- valley movement and that they
slowly transported boulders and other objects left upon their surface. In 1841
the Bishop of Annency, M. Rendu, published his remarkable work, Theorie
des Glaciers de la Savoie (edited by George Forbes, 1874, translated by
Alfred Wills), in which he shows a surprising insight into the laws of
glacial motion. "Between the Mer de Glace and a river," he writes, "there is
a resemblance so complete that it is impossible to find in the latter a circum-
stance which does not exist in the fo'rmer" (page 85). Since this was written,
in 1839, glacial investigation has done much to justify, if not to actually verify
the generalization. Streams of ice and streams of water have many character-
istics in common, as well as important differences. All are agreed that the
cause of the movement is, in both cases, the force of gravity, acting upon the
mass itself, or some other mass in contact with it. As to the nature of the ice
movement, those whose opinions should carry the greatest weight are not yet

30 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

in agreement. So far as the present investigation is concerned the phenomena
observed can be satisfactorily explained only upon the theory that under certain
circumstances, and within certain limits, ice is capable of behaving as a plastic
body, that is, capable of yielding continuously to stress, without rupture.
In the discussion of this property of ice in the closing chapter of this report
it is pointed out that the "plasticity" of ice, a crystalline substance, must be
thought of as essentially different from that manifested by such amorphous
substances as wax or asphaltum.

That the Victoria Glacier is flowing valleyward is capable of direct demon-
stration. From range lines across the glacier, the Messrs. Vaux marked the
position of a certain large boulder July 26, 1899. This was on the portion of the
glacier opposite the tributary and back some 6,800 feet from the nose. By July
24, 1900, this boulder was found to have moved forward 147 feet, while a second
"near the terminal moraine" was found to have moved 115 feet.1 So far as
may be inferred from the phenomenon of the "dirt bands," to be later described,
the former boulder was not in the locus of maximum surface movement, but
some 360 feet to the west. In order to gather more definite data concerning
the movement of the lower Victoria a line of 18 steel plates was set by means of a
transit, the plates being placed as nearly as convenient at an average distance of

w fl*

— *<

5 t-B- **
o

ioo feet. The line of plates was back 3,600 feet from the nose of the glacier (see
map) and was marked by setting the instrument over an established point upon
a large block on the shoulder of Mt. Whyte and sighting across to the sharp
edge of an easily recognizable cavity in the face of Mt. Aberdeen. The plates were
of the style successfully used by the Vaux brothers upon the Illecillewaet Glacier
and were 6X6XJ inches, with a 9-inch piece of f inch gas pipe screwed into the
center. They were given two coats of brilliant red paint, were numbered, and
had the actual reference line marked in white, after the short piece of pipe had
been driven into the ice. An arrow was marked upon each for purposes of
orienting, in case they should become turned. In general, where the melting
was greatest, it was found that the pipe did not sink into the ice and retain
its vertical position, but allowed the plate to drop to one side. The end of

1 Proceedings of the Academy of Natural Sciences of Philadelphia, Mch., 1901, p. 213. "Observations
Made in 1900 on Glaciers in British Columbia."

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the pipe, however, marked the spot upon the ice over which it had stood, so
that its original position could be readily restored (see cross-section). It so
happened that the next ten days in the valley were cool and cloudy, with small
amounts of rain almost daily. This period was followed by ten days, and more, of
bright warm weather, with considerable surface melting. After each of these periods
the instrument was set in its former position and the original line again established,
but upon ice which had moved down into this position. Measurements were made
from this new line to the various plates and the amount of forward movement thus
determined. The full data gathered from observations upon this set of plates
are shown in table v. The relative vertical position of the plates is shown in
the cross-section, as well as their relation to the surface features of the glacier.

TABLE V.

OBSERVATIONS UPON THE SERIES OF STEEL PLATES, SET ACROSS THE VICTORIA

GLACIER, JULY 9, 1904.

(Total Distance across Glacier along line of plates 2,167 feet.)

J.J

t-f

1

>,

1"

0 >.

£g

« j«

1"

a
I

•g

s

"^i

1*

_g

'5

^J

IB

M 0\
V. -

£3

|

d

«.s°

O 'O

g-a

s

rtSJ .

£

.9-3

"d ^

So
s *3

O ^"

•2 >. w

rt -
.2 S o

|!

§5

rt J?

«fo

"8 .
||

«.§'

R emarks.

"•£

*j •*->

g«T ^

^"V "*

tS «

££

'+3 S **"

«*o c "

'*J ^Cfl

o> c

%

— w

1

i!

SIS'

si?

Q.S

U'S

SIS

l?5

<Sa

S-3S

<l

31

W "

Feet.

Feet.

Feet.

Feet.

Inches.

Inches.

Inches.

Inches.

Feet.

Inches.

Feet.

Feet.

i.

63-4

6585.60

6585.47

0.13

0.068

-0.787

1.181

O.02O

O 21

0.006

0.18

62

Crest of left lateral

moraine.

2.

163.2

6578.20

6576.11

2.09

1.116

0.394

o.ooo

0.197

3.00

0.084

2.56

zoo

Motion irregular.

3.

262.8

6577.74

6576.02

1.72

0.936

O.OOO

o.ooo

o.ooo

137

N

Plate lost in 1905.

4-

342.7

6575.26

6573.04

2.22

1. 174

2.953

2.953

o.«9S

15-70

0.445

13-54

160

Motion for warm pe-

riod retarded.

5-

443-3

6582.64

6579.89

2.75

1.386

7.874

10.630

0.925

26.10

0.744

22.63

200

Western slope of low
crest.

6.

549-9

6595-27

6592.70

2-57

1.231

10-433

17.126

1.378

33.80

0.959

29.17

245

Western slope of low

crest.

7-

642.1

6601.22

6599.24

1.98

0.866

11.811

22.835

1.732

47.60

1.350

41.06

275

Crest of low divide.

8.

74L4

6595.36

6593-06

2.30

1.009

12.402

27.362

1.988

56.50

1.603

48.76

292

Eastern slope of low

crest.

9-

839.8

6589.06

6586.26

2.80

1.250

18.504

26.378

2.244

51.50

1.460

44-41

302

Eastern slope of low

crest.

10.

938.3

6585.02

6582.15

2.87

1.292

18.110

25-984

2.205

72.10

2.045

62.20

306

Eastern slope of low

crest.

ii.

1037.6

6579-59

6576.65

2.94

1-273

19.882

33.661

2.677

76.30

2.165

65-85

306

Maximum motion for

year.

12.

1222. 0

6574.49

6570.86

3.63

1-724

18.701

22.638

2.067

74.20

2.105

64.03

287

Lowest plate.

13-

1320.5

6580.91

6577.11

3.80

1-735

18.701

36.024

2.736

74.8o

2.122

64.54

276

Maximum summer ab-

lation.

14.

I4I9.7

6593.45

6590.15

3.30

1.491

16.929

33-071

2.500

71.00

2.OI4

61.26

262

Gradual slope to me-

dial.

15-

I5I9.I

6613.69

6611.45

2.24

0.959

10.827

31-693

2.126

67.40

I.9I2

58.16

252

Gradual slope to me-

dial.

16.

1618.0

6629.03

6627.25

1.78

0.689

14.370

3I.S99

2.283

63-30

1.795

54.60

236

Gradual slope to me-

dial.

Bldr.

1704.0

6647.90

6646.46

1.44

0.535

11.024

29.528

2.028

58.50

i. 660

50.49

220

Boulder on medial

moraine.

17.

1791.6

6653.14

6652.03

i. II

0.374

11.614

25.787

1.870

55-50

1.574

47-88

1 88

Crest of medial

moraine.

18.

2083.5

6687.12

6687.15

-0.03

-9.646

4.134

0.17

0.005

0.15

80

Crest of right lateral

moraine.

In column 7 is given the total forward movement for the cool period, July 9 to
19, from which the average daily movement can be seen at a glance by shifting

32 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the decimal point one place to the left. The greatest movement was shown by
plate ii, almost 20 inches, and this was the plate nearest the centre. From
this point the motion diminished very gradually toward the west, but less rapidly
eastward, toward the crest of the medial moraine, and then fell off very suddenly.
The plates in the crests of the two lateral moraines had moved up-stream, sup-
posedly from the settling of the debris due to ice melting beneath. Plate 3
seemed to be located upon a stagnant portion of ice, showing no movement for
either the cool or the warm period, and then could not be found in 1905, appar-
ently because of the side cutting of a surface stream. For the warm period,
July 20 to 29, the results are given in the next column for ready comparison.
The motion is seen to be greater, almost twice as great for plate 14 and almost
three times as great for plate 15. Both plates upon the lateral moraines had
moved ahead, although that upon the right lateral had not yet regained its
original position. Plate 2 showed no movement whatever for the .warm period,
and number 4 showed no increase of movement. In column 9 there is given the
average daily motion for these 20 days of midsummer, which ma)- be regarded as
typical of the season. For this period the greatest forward movement was shown
by plate 13, having a daily average of 2.736 inches. It is located some 287 feet
east of the centre of the glacier, and, as previously pointed out, showed the
maximum ablation. Number 3 showed no movement whatever. On September
5, 1905, the plates were again located, all being readily found except plate 3,
and their distances from the original line determined. These results are listed
in column 10, from which have been calculated the average daily motion and the
motion for a year. In all cases there was a down-stream movement indicated,
although very slight for the two lateral moraines. The greatest movement was
shown by plate 1 1 , the one nearest the centre of the glacier, amounting to a total
of 76.3 feet for the entire period of 423 days, or a daily average of 2.165 inches.
This represents a yearly motion of about 66 feet. The central position of the
plate of maximum movement for the year was to be expected from the very
straight course of this part of the glacier. Its daily motion for the year is about
8 1 per cent, of its midsummer motion, which means that for the greater part
of the year the movement is fairly uniform. The table suggests that during the
season of maximum motion there are cross-currents set up in the ice, and, with
reference to the body of the ice itself, not the bed, even back currents. During
the year, however, the impulse is steadily and regularly forward. Studies upon
the dirt bands of Forbes, to be later discussed, indicate that as we approach the
steeper ice slope opposite Lefroy, the motion is more rapid than that given in the
table.

b. Frontal changes. The rate of melting about the nose and side of the glacier,
in connection with the rate of forward movement of the ice, determines the
behavior of the front. When these two factors are balanced the glacier appears
to halt, and, if carrying debris, begins to build a terminal moraine. If either
the rate ot forward movement, or the rate of recession from melting, is in
excess then the glacial extremity advances, or retreats, entirely regardless of the

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 33

fact that the ice of the glacier is continually moving forward. Owing to the rock
veneer which completely covers the nose of Victoria, the amount of melting
even during midsummer is very small. The last episode here was one of advance,
the glacier having extended itself, some decades ago, into a forest of spruce and
fir and checked its own advance by mounting a heavy moraine of rock fragments
which it was incapable of pushing aside (plate v, figure i). The cut stumps and
broken trunks which lie about the nose, some of them entirely out of reach of
the present glacier, appear to have been produced by an avalanche from between
Mt. Aberdeen and Castle Crags, which encircled the nose when it stood some-
what farther back than at present. Trees now growing in the path of this
avalanche are 28.3 inches in circumference and by calculation should be 130
years old. In order to determine how the nose was behaving, three accurate
measurements were made with a steel tape between definite points upon coarse
blocks of the old moraine and upon others that seemed rather firmly embedded
in the frontal slope. Between July 9 and September 13, 1904, in all 66 days,
each of the latter blocks had settled back approximately an inch, presumably
owing to the wastage of the ice beneath from melting. Confirmatory evidence
that such melting was in progress was furnished by a small dear stream of water at
32°, which escaped through the rocks just west of the nose. Between September
13, 1904, and September 2, 1905, when measurements were again made, this
small recession was partially made up, but the blocks still lacked .36 in. to .72
in. of regaining their former position. With the front so delicately poised it is
evident that a very small additional impulse from behind would inaugurate an
advance.

At a point 2 ,000 feet up from the real nose, at about the middle of the oblique
ice front already noted, there lies a large red quartzite boulder, which was used
by the Messrs. Vaux as a reference block. This is the largest of the three blocks
in the middle foreground of plate x, figure 3, as it appeared in August, 1903.
This boulder was observed protruding from the ice, a little over half-way up the
face, in the midsummer of 1898, by Prof. Charles E. Fay. In this position it was
photographed by him, and also a week later, when it had fallen. Plate x, figure
2, shows the boulder in position in the ice. In 1899, July 26, this boulder was
found by the Messrs Vaux to be 20 feet from the ice front. How much of this
20 feet was due to recession and how much to the rolling or bounding of the block
in falling, cannot now be determined. On July 24, 1900, the boulder was found
to be 26 feet from the ice, indicating a recession of 6 feet for the year. August
23, 1903, the block was found by the writer to be 76 feet from the ice foot, giving
an average recession since 1899 of 14 feet. The following July the block was
marked with bright red paint, so that it could be readily located by others: "A.
Ice foot 74.5 ft. 7/23/'o4. Sr." The elevation of a line upon the face, deter-
mined by spirit level from Lake Louise as a base, was recorded as 6,264 feet1
above sea-level. When compared with the distance noted above for the previous

1 This elevation was based upon 5,675 feet for the height of Lake Louise above sea-level. The corrected
elevation as now given by the Canadian Topographic Survey is 5,670 feet, or five feet lower.

34 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

year this would indicate an advance, whereas the glacier was actually in retreat.
The necessity of taking the measurements as nearly as possible at the same cor-
responding time in the season becomes evident. In 1904 measurements were also
made August 4 and September 13, giving distances of 76 and 86.45 feet respect-
ively. If we assume a uniform recession between these last two dates we have
a daily amount of 0.26 feet and for the interval between August 4 and August 23,
at which date the measurement was made in 1903, an additional recession of 4.94
feet. This amount, added to that of August 4, gives 80.94 feet, and the approxi-
mate recession from August 23, 1903, to August 23, 1904, was 5 feet. In 1905 the
measurements were made September 2, and gave a distance of 106.8 feet from
the reference block to the ice foot. This means a recession of about 25 feet for
the year 1904-5. For the series of six years 1899 to 1905 the total amount of
recession observed at this point was 86.8 feet, or an average of 14.5 feet annually.
The following summary for this boulder may be given:

1 898. Fell from ice early in August.

1899. 20 feet from ice foot.
1899-1900. Ice receded 6 feet.
1900-1903. Average recession of 19 feet.
1903-1904. Ice receded 5 feet.
1904-1905. Ice receded 25 feet.

1899 to 1905. Average recession 14.5 feet.

About 375 feet nearer the nose a second block was selected for reference and
upon July 23, 1904, marked " B. To ice 38.5 ft. 7/23/'o4. Sr." Between this
date and August 4, 1904, the recession amounted to 3.9 feet and up to September
13 equalled 11.5 feet. The distance from the block to the ice foot was again
measured September 2, 1905, and amounted to 63.2 feet, indicating a recession
between September 13, 1904, and the latter date of 24.7 feet. Calculated, as
above, for the year September 2, 1904, to September 2, 1905, the recession was
approximately 15 feet. Since the exposure of the ice face opposite blocks A and
B is so nearly uniform, we may assume safely that the rate of melting upon the
oblique face is substantially the same at the two points of reference. The
diminished recession of the ice at B would then indicate that the forward move-
ment of the layers here must be greater than at A. From data already cited
it is seen that the forward movement at the nose is insignificant and it appears
that the main current of ice, as it approaches the nose, is deflected to the west-
ward and that the oblique ice wall is in reality part of the front.

c. Shearing. The steeply inclined ice front, having a slope of 46°, near
reference block B,. shows a succession of ice strata, more or less well defined, which
dip back into the glacier at a rather steep angle. At the mouth of the abandoned
drainage tunnel (plate vm, figure 4, and plate xn, figure 3), these strata in 1904
had a dip of 26°, which is below the actual dip. Between these strata there is dust,
sand, a little fine gravel, and, occasionally, a cobble-stone, but the amount of for-
eign matter is small and inconspicuous. A few consecutive days' visits to this

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 35

part of the glacier, in early July, showed that a differential movement between
adjacent strata seemed to be taking place (plate x, figure 4) , the upper layers being
pushed beyond the lower. There was not enough foreign matter in the layers to
explain the phenomenon by differential melting. In the case of the South Point
and other Greenland glaciers Prof. T. C. Chamberlin observed a jutting of the
upper stratum which was apparently due to the more rapid melting of the under
layer, owing to its heavier load of dark -colored debris and consequent more rapid
absorption of the sun's heat. Upon the same glacier, however, he found very
conclusive evidence that the upper stratum may be pushed bodily over the lower,
giving rise to a shearing action between the adjacent strata.1 In August, 1903,
Prof. I. C. Russell found a similar phenomenon on one of the small glaciers
visited on the Three Sisters, Oregon. In this case he thought the evidence con-
clusive that the jutting of the upper stratum was due to differential melting.2
In order to ascertain whether or not this was a similar case of shearing, a place
was selected 50 to 52 feet above the base of the ice and heavy spikes driven into
the ice until their heads were flush with the surface. Three were placed in the
base of the upper stratum, about three feet thick, and three corresponding ones
in the upper part of the subjacent layer, which had a thickness of about two feet.
July 2 1 the upper layer projected beyond the lower 19.7 inches at the place selected
for observation. Two days later it was evident that the melting was greater
upon the upper layer, in spite of which it now projected 24.4 inches beyond the
lower. The spikes were now visited regularly for 10 days, July 25 to August 3,
the amount of melting measured, as well as the amount of projection of the two
layers, and the spikes reset. These measurements were necessarily rough, but
they showed each day that the melting was greater upon the upper stratum,
the average amount for each spike being 1.76 inches, while that for the lower
stratum was 1.53 inches, or nearly J inch less. Some sand and fine gravel,
washed down from above, daily accumulated in the lee of the projecting upper
layer and gave the appearance of a concentration of dirt in the upper part of the
lower stratum. When this dirt was small in amount it was observed that melting
was accelerated; when greater in amount, that the melting was retarded. The
upper stratum continued to gain slowly, but irregularly; reached a maximum
of 26.6 inches and closed at 25.6 inches, or about 6 inches more than at the begin-
ning of the observations. The results are tabulated below for inspection. Time
did not permit the verification of the results at other points where the same
thing appeared to be taking place, but there seemed to be no question tha t the
upper layer was moving bodily over the lower. This movement represents a
shearing of the body of the glacier, the shearing-plane lying between the adjacent
strata, but not a shearing of the ice itself. Knowing how readily iron absorbs
heat it may be supposed that six-inch spikes might induce melting sufficiently
to render their use unsatisfactory. Lying in the ice horizontally there was con-
siderable melting about the outer half of the spike, allowing it to sag and slide

'Journal of Geology, vol. in, 1895, p. 676.

* " Glacier Cornices," Journal of Geology, vol. xi, 1903, p. 783.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

forward, but they did not seem to penetrate the ice any by such action. But
even if such an effect was produced to an appreciable extent, it would have
been more pronounced upon the upper row of spikes, which received more sun's
heat owing to their more exposed position, and the actual melting upon the upper
layer would have been still greater than is indicated in the table. Although
the purpose of the experiment was to ascertain the differential melting, rather

TABLE VI.
SHEARING OBSERVATIONS, VICTORIA ICE FRONT.

Date.

Upper stratum.

Lower stratum.

Projection of
upper layer
beyond lower.

Temperature.

Remarks.

July 23

Melting about spikes.

Melting about spikes.

Max. Min.

July 25
26
" 27
28
" '9
" 3°
" 31
August i

" 2

3
:o days' obs.

2.75 in. 2.00 in. 2.50 in.

1.50 in. 2.25 in. 2.25 in.

25.2 in.

6o.7°F. 35.0° F

Cloudy.

2.00 2.06 2.25

1. 25 '• 25 3. 25

25.0

55.2° 40.0°

Warm.

2.37 2.56 2.37

1-25 3-37 2.25

26.6

68.4° 38.8°

Warm and bright.

3.25 2.56 2.50

2.75 »-35 2.38

35.6

69-70 45.4°

Very warm.

1.62 1.56 1.38

I. IQ 1.50 1.87

24.6

J4.o" 45.5°

Cloudy.

0.87 0.69 1.25

0.94 0.75 i .06

25.0

74.0° 3«.3°

Cool and cloudy.

i. 75 i.«9 i. oo

l.oo 1. 06 1.44

?<:.4

50.4° 38.0°

Mostly cloudy.

1.50 1.63 2.00

1.63 1.38 1.50

26.2

Bright.

2.31 2.00 2.38

1.50 1.50 1.75

25.0

Bright.

2.OO 2.12 I.5O

1.38 1.62 1.56

»5.6

Bright and warm.

Average melting 1.76 in.

Average melting 1.53 in.

Increase 5.9 in.
since July 21.

than the actual, it is interesting to compare the maximum average daily effect
here observed with the maximum melting observed upon the surface of the
glacier, where least protected by debris. See table v, column 6, plate 13, page 31.
d. Crevasses. The general forward movement of the ice, and its inability to
adjust itself to inequalities in its bed, give rise to systems of cracks, or crevasses.
These show that the limit of tensional strain, without rupture, has been exceeded
in this part of the ice. They occur in all parts of the glacier from the bergschrund
to the very nose, and when insecurely covered with snow, they form the greatest
menace to glacial exploration. The inexperienced cannot be too strongly
cautioned against the danger arising from these concealed traps, against which the
judgment of best trained Swiss guides is sometimes pitted in vain. In passing
from one portion of its bed to a sufficiently steeper slope, as that opposite the nose
of Mt. Lefroy, v-shaped cracks in the ice occur, extending directly across the
glacier. They penetrate to considerable depths into the ice, as measured in
feet, but their depth, compared with the total thickness of the ice, is probably
small, unless the change in the inclination of the bed is very abrupt, when they
may reach the bed of the glacier. When these transverse crevasses have an east-
west trend, the sun's rays strike the northern lip of the crevasse more strongly
than the southern and, in the course of the season, it becomes more rounded.
In passing down the slope the crevasse walls come together and the crevasse is
healed, except for the slight depression caused by the greater melting upon the

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 37

northern crevasse wall. It is into this depression, which becomes convex down-
stream owing to the more rapid central movement, that fine de~bris may collect
and give rise to the "dirt bands" of Forbes, to be presently described. As
the glacier rounds Lefroy and enters a broader portion of its valley, it has a
chance to spread laterally, and longtitudinal and somewhat radiating crevasses
are opened which may intersect those having the transverse position. If the
glacier is again contracted these crevasses will also be closed, and if any depression
is left, it will slope down-stream and not have a tendency to collect debris.

The more rapid movement of the middle portion of the glacier, when compared
with the sides, which are retarded by the friction of the valley walls, induces
tensional strains between the central and marginal masses. In consequence,
along the sides, there is opened up a characteristic system of marginal crevasses
at right angles to the resultant strain. These extend inward and upward,
making, theoretically, with the sides angles of about 45°. The difference between
the central and marginal flow must reach a certain value, and be sufficiently
abrupt, otherwise the ice seems capable of yielding without rupture. In this
way we may account for the absence of marginal crevasses over the lower west
side of the Victoria. The very sudden change in movement, shown in table v,
between the margin and the ice of the near-by medial moraine, plates 18 and 17,
is evidently responsible for the series of marginal crevasses that are seen between
the line of plates and the tributary (see plate in). From their absence upon this
side, farther down, we infer that the ice beneath the medial moraine becomes
more sluggish as the main flow is deflected westward. Opposite Mt. Lefroy
conditions are favorable for their formation and they are well represented upon
either side. Opposite the tributary they do not occur, as the marginal ice is
sufficiently yielding. Upon the tributary itself these crevasses are well repre-
sented, except over the collecting area for the Lefroy. After their formation
their inner ends may be swung around until they assume a transverse, or even
reversed, position, as seen upon the Aberdeen side of the Lefroy. Here we find
one series, averaging N. 51° E. and making angles of about 66° with the margin
but ranging from 5 2° to 86°; and a second series, many of them nearly closed, and
apparently older than the preceding, having an average direction of N. 95° E. and
making with the side angles of about 111°.

The size of many crevasses in the spring and their contents of fresh snow show
that they may persist through a series of seasons. Sometimes they become
partially filled with water which may melt out cavities in their walls and give
rise to the most exquisite ice grottoes, a peep into which is worth miles of travel.
The closing of crevasses sometimes confines pools of water, often under hydro-
static, or ice pressure, and as the surface of the ice is lowered by melting, the water
suddenly bursts forth with geyser-like action. The compression of air enclosed
in cavities, or brought in by surface streams, often gives rise to a bubbling at the
surface and a faint hissing, or chirping sound — the "sighing " of the glacier.

38 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

CHAPTER III.
VICTORIA GLACIER (Continued).

i. GLACIAL STRUCTURE.

a. Stratification. In this chapter there is set off for description a number of
features, especially well shown upon the Victoria and its tributary the Lefroy,
but which were more or less well represented upon the other glaciers also and are
characteristic of glaciers in general. Among the first of these is the stratification
the origin of which in the neVe" has been given on page 22. It is conceivable that a
stratification in the basal layers might arise exceptionally through the operation
of differential stresses in the body of an unstratified glacier. As pointed out by
Chamberlin in the case of the massive Greenland glaciers shearing-planes may thus
arise leading to a concentration of debris. The lower stratum over which the
shearing takes place may be protected from the shearing thrust, may be more
heavily charged with debris, or may be more rigid because of its temperature
and water content.1 In the case of the Canadian glaciers studied it seems
probable that the strata are depositional, in very large part, at least. Conditions
most favorable for the formation of shearing-planes would seem to be found in the
case of the Illecillewaet Glacier, owing to the body of ice and its rapid descent
from its reservoir. The depositional stratification is almost completely obliterated
by the ice cascade and none other has arisen to take its place.

The stratification of the Victoria continues throughout the glacier's extent,
and is seen at the oblique front, in the drainage tunnels and channels, in the
moulins, and upon the walls of the crevasses. The line of demarcation between
adjacent strata is usually only a soiled streak, but sometimes there is sand, gravel,
and an occasional cobble-stone. The strata vary in thickness from 1 2 inches to
10 or 12 feet, as seen upon the Lefroy. This thickness would indicate that 9 to no
feet of loose snow had taken part in their formation. The average thickness of
the Victoria strata is not too great to suppose that they may represent the
accumulated and compacted snow fall of the year. Those of unusual thickness
are to be ascribed to avalanches. About the mouth of the abandoned drainage tun-
nel in 1904 the stratification of the ice was well displayed (plate vm, figure 4, and
plate xii, figure 3) as previously referred to. Three strata here averaged 2 6 inches,
the full thickness of the lower one not being seen. The uppermost layer was
wedge-shaped and thickened from 13 inches to 81 inches. The strata all dipped
back into the body of the glacier at an average angle of 26°, as measured upon the
tunnel walls, but this was less than the actual angle when measured at right
angles to the strike of the layers. The irregularities shown in the strata here,
as well as in the oblique ice face, are probably due to the partial nourishment of
the glacier by means of avalanches of snow and ice. Upon the regenerated Lefroy
Glacier the strata are massive, 6 to 12 feet in thickness, having been produced
entirely from the avalanches from Mt. Lefroy. These strata all dip towards the

' See Geology, vol. I, Chamberlin and Salisbury, p. 303.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XI.

FIG. I. — Line of contact between two "dirt zones," Lefroy Glacier. These zones repre-
sent outcropping edges of depositional strata.

Mitn

FIG. 2. — 'Dirt zones" upon Lefroy Glacier, frequently confused with "dirt bands" of Forbes. Compare figure 2,

plate XVI.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 39

region of accumulation, directly beneath the front of the hanging glacier. In
the lower part of the glacier this dip averages 22°, ranging from 12° to 26°,
while farther up-stream the dip is more gentle, only 5° to 10°, as well seen
in the crevasse walls. The Mitre Glacier, near the junction of its two feeding
streams, is crevassed and faulted and displays a very regular stratification
(plate vn, figure 2).

b. Dirt zones. Upon a moderately steep slope, such as is found upon the
lower Lefroy, the outcropping edges of the strata, somewhat differently charged
with debris, give rise to broad contrasting zones which pass evenly and symmetri-
cally around the slope. As generally seen these bands are convex in the direction
of flow, but irregularities in the surface slope of the ice, or in the angles at which
the strata come to the surf ace, may make them concave down-stream for portions,
at least, of their course (plate xi, figure 2). The upper edge of one zone upon the
Lefroy contrasts very strongly with the adjacent layer, as shown in plate iv,
figure 2. It was the abnormal position of this line, first seen from the Devil's
Thumb, that furnished the clue needed to decipher the relation of the Lefroy to the
Mitre Glacier. A nearer view of this zone line, and two adjacent ones, is shown
in plate vin, figure i, and a still nearer view in plate xi, figure i. Because of the
irregularity and small size of the strata, as well as the debris covering, the phe-
nomenon is not well seen upon the Victoria. At the place where it should show
the best it is, furthermore, obscured by the dirt bands of Forbes, with which
the zones are often confused. These two features are so different in origin and
significance, yet often so similar in appearance, that they should be sharply
separated in the field and in descriptions of glaciers. Plate iv, figure 2 shows
the dirt zones, upon the Lefroy, at the left, and the dirt bands, upon the Victoria,
in the middle foreground.

c. Granular stricture. A lump of ice from the body of a stratum, which has
not yet begun to show any signs of melting, is compact, firm, brittle, without cleav-
age, and beautifully blue by transmitted light. It appears quite homogeneous,
except for the presence of air spaces, which may be sparingly and irregularly
scattered through the ice, or they may be arranged in seams, to be presently
described. Under the polariscope, in thin slices, the ice is seen to be crystalline
in structure and made up of closely pressed polyhedrons, ranging in size from
hazel nuts to goose eggs. These polyhedrons are the so-called glacial granules,
that may be traced back to the neVe", growing smaller and smaller, upon an
average, as we recede from the nose. They fit tightly together, interlocking
perfectly, have curved rather than plane faces, and show no spaces nor signs of
any cementing material between the individual granules. There seemed to be a
correspondence between the size of the glacier and the size of the granules seen
about the nose, the largest granules being observed in the Illecillewaet and Yoho
glaciers, in the case of the latter ranging from 0.2 inch to 2.75 inches and averaging
about one inch. From the fact that such granules occur in no other form of ice,
that they may be traced back to the ne've', becoming smaller and smaller and
more numerous, the inference is reasonable that, in some way, these granules

40 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

must be derived from those pellets which constitute the typical neve". The ques-
tion as to how the granules are developed at once arises, but cannot be yet an-
swered with certainty. (For a fuller discussion of this subject see page 127). That
the larger are not produced by the simple freezing together of a certain number
of the smaller pellets is shown by the fact that each mature granule is crystallo-
graphically homogeneous. Those who have written most recently upon the sub-
ject hold the view that the granules are permitted to grow by a process of partial
melting and refreezing, the larger thus appropriating to themselves the water
derived from the melting of the smaller. Mugge holds that this melting takes
place at the outer limits of the individual granules because of the constant read-
justment of pressures within the body of the glacier,1 and in this change of the
granules he sees the cause of glacial motion. Chamberlin believes that a similar
change occurs because of differential stresses upon the granules undergoing constant
adjustment, assisted by whatever heat energy may be conducted into the glacier
' from above.2 Drygalski recently argues in favor of a melting of the granule
by pressure both internally and at its outer surfaces, by which some granules
may be completely liquified and subsequently refrozen.3 Upon this action
he bases his theory of glacial motion and the orientation of the granules
about the nose, as brought out in his Greenland report in 1897 cited below.

Experiments of Hagenbach-Bischoff in 1 883 showed that when two ice crystals,
having differently directed axes, are pressed together they unite without melting
into a single crystal, "the larger eating up the smaller." The union differs from
the regelation of Tyndall in that there is a rearrangement of the molecules by
which the resultant crystal is crystallographically and optically homogeneous. To
distinguish it from the method of granular growth due to melting and refreezing
it is spoken of as a "dry union." This principle applied to the glacier would lead
to a continual reduction in the number of granules and a corresponding increase in
their size, as pointed out by Hagenbach-Bischoff, Heim, and Emden. It will be
shown later (page 128) that this theory of granular growth seems to the writer
to best explain the remarkably perfect preservation of the often delicate laminae
and blue bands seen about the nose and sides of the glacier. Combined with
the special type of plasticity exhibited by ice crystals this method of perfect
dry welding may explain the absence of noticeable distortion of the ice granules,
which, as urged by Chamberlin, should be observed in the direction of flow if the
glacier moves because of its viscosity.

In order to determine whether or not there was any tendency towards the
orientation of the granules in the basal layers about the nose, thin slabs of ice

'"Weitere Versuche iiber die Translationsfahigkeit des Eises, nebst Bemerkungen fiber die Bedeutung
der Structur des gronlandischen Inlandeises," Neues Jahrbuch fur Min., Geol., und Pal., 1900, Bd. n,
S. 87 zu 98.

* "Recent Glacial Studies in Greenland," Presidential Address before the Geological Society of America,
Bull. Geol. Soc., vol. 6, 1895, p. 211; "A Contribution to the Theory of Glacial Motion," Decennial Publica-
tions of the University of Chicago, vol. ix, 1904, pp. 10 and n; Geology, by Chamberlin and Salisbury
vol. i, 1904, pp. 299 to 306.

a " Ueber die Structur des gronlandischen Inlandeises und ihre Bedeutung fur die Theorie der Gletscher-
bewegung," Neues Jahrbuch fur Min., Geol., und Pal., 1900, Bd. I., S. 71 zu 86.

e 3

"1

o

I

>, JjO

-

S I

S-a

a£

•SS

o
I

= fe

^
a

1 5

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 41

were sawed out in various directions and melted down to thin slices by rubbing them
over the face of a warm saw-blade. Examining these sections with the polariscope,
it was found that in the case of those cut horizontally from the glacier, from
J to ^ of the granules remained dark when revolved. In the case of sections
cut vertically, either across or lengthwise of the glacier, only an occasional granule
was found to show this phenomenon. From this it appears that there is a ten-
dency towards the orientation of the granules near the lower portions of the
Victoria, Yoho, and Illecillewaet glaciers, a considerable percentage of the
granules having their main optic axes in a vertical position. The same phe-
nomenon was observed by Drygalski in the case of the Greenland glaciers.1

d. Capillary structure. When glacial ice is subjected to a moderate melting
temperature for a sufficient length of time there is developed a network of capillary
tubes, at the junctions of three or more granules. These tubes are approximately
circular in cross-section and from 0.008 inch to 0.04 inch in diameter. Their
walls reflect the light strongly and give the appearance of silver threads, more or
less perfectly outlining the granules. From the ease with which liquids course
through the tubes one infers that they are free from or contain but little air.
From beneath the margin of the Yoho Glacier it was possible to get some of
them upon the camera-plate, although, many of them being out of focus, they
all appear disconnected (plate xn, figure i). By making a strong solution of po-
tassium permanganate and placing it in a basin hollowed in the ice, the capillaries
were in a few minutes beautifully infiltrated, the red solution contrasting strongly
with the rich blue ice (plate xn, figure 2). Upon the faces of crevasse walls, and
in the drainage tunnels, where the sides are smoothed by melting, these tubes may
be seen in longitudinal section, forming a pattern by which the irregular
granules are outlined. These are the tubes which Agassiz and Forbes found in
the Alpine glaciers, but which Huxley and Tyndall did not discover. Agassiz
was in error in supposing the entire body of the glacier to be permeated with such
a system of capillary tubes and Huxley in denying that any part of it was.

e. Melting features. As melting proceeds the capillaries become larger ; ir-
regular, "crinkly" spaces are opened between the faces of adjoining granules, and
the delicate network is gradually obliterated, as shown in portions of plate xn,
figure 2 . With this increased reflecting surface the ice loses its deep blue color,
becomes whiter, and when the granules are small it assumes somewhat the
appearance of ne>e. A slight pressure now, or a sharp blow, will cause the ice
to crumble into its component granules. These granules are shown, but rather
indistinctly, in plate xn, figure 3. While still in position, as well as after they
have fallen apart, the granules are seen to be covered completely with delicate
parallel ridges and rows of fine points winding over the surface and having no
definite direction with reference to the crystal. The ridges and rows of points
are about 0.04 inch distant, but show some variation, and form a complicated
pattern that is different for each granule, suggesting more strongly than any-
thing else the ridges seen upon the inside of one's finger-tips. This phenomenon

< Grdnland-Expedition der Gesellschaft fur Erdkvnde zu Berlin, 1891-93, Bd. i, 1897, S. 494.

42 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

was noted by Drygalski in the granules of the Greenland glaciers and described
briefly upon page 488 of his report cited. It had been previously observed
and described by Emden in his paper Uber das Gletscherkorn, p. 22, figure
5, and designated as melting water curves. While neighboring granules were
in position, no correspondence could be made out between the ridges and
furrows of adjoining faces. An attempt was made to take impressions of the
markings but no suitable material was at hand. The wall preparation "alabas-
tine" reproduced perfectly the finger markings, but refused to work with a wet
ice surface. The "stripes of Forel" are delicate ridges, passing around the
granules at right angles to the main optic axis, and evidently connected with
the intimate crystalline structure of the crystal. They mark the edges of the
very fine plates of which each ice crystal is composed, placed together with their
flat faces perpendicular to the main optic axis. The ridges here described are
entirely different and do not suggest to the writer any possible explanation.
They are certainly due to the manner of surface melting but it is far from appar-
ent what could give rise to such a pattern. In the prisms of lake ice Emden
found both the melting curves and Forel's striping present, with an intermediate
type of ribbing, and concluded that all three were due to one and the same cause
and independent of the structure of the crystal (p. 24).

Granules that have been well acted upon by the sun show a system of very
flat, circular disks, all with their planes parallel and at right angles to the main
optic axis. These were first observed and figured by Agassiz in his Systeme
Glaciaire, 1847 (plate vi, figures 7 and 10) and described also in his Geological
Sketches, vol. i, p. 275. They were believed by him to be air bubbles, flattened
by pressure, although observed to lie differently in adjoining granules. These
are now known as "Tyndall's melting figures," described in his Glaciers of the
Alps, Ed. 1896, pp. 353 to 361. They represent "vacuous space," left in the
ice by the contraction of the water when changed from its solid to its liquid
condition, the melting planes coinciding with the crystalline plates, of which
the granule appears to be composed. They are thus serviceable in enabling one
to determine the direction of the main axis of each granule, but there were not
enough of them seen at one time about the nose of the glaciers to settle the
question of the orientation of the granules.

/. Blue bands. Many observations were made upon the blue bands, of
which the strata are generally composed, with the hope of shedding some light
upon their position and direction in the ice and their relations to the strata. In
general, they were found well developed about the nose and along the sides of
the glaciers, well up toward the ne"ve" region. The lower Victoria has too much
debris covering to enable them to be well seen at the surface, but in the tunnels
and moulins and along the walls of the surface streams they are to be seen in a
good state of development. At the mouth of the tunnel they were found to
average 0.59 inch to 0.75 inch and to dip back in to the glacier at an average angle
of 9°, while the average slope of the strata was 26°. This unconformity of the
laminae and strata is well shown in plate xn, figure 3, although the laminae

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 43

become indistinct in proportion as the granules separate. In the moulins,
opposite the oblique ice face the average inclination up-stream was found to be
30°. Under the medial moraine, near the nose of Mt. Lefroy, the bands were
longitudinal, vertical near the centre but radiating, fan-like, upon either side;
the outer ones inclining as much as 45°. Although for over 65 years the subject
of study, we are not much nearer an explanation of this common glacial feature
than when first observed in 1814 by Brewster. The idea of Forbes that they
represent ice-filled crevasses, or shearing-planes, has been generally abandoned.
The early view of Agassiz, that they represent the original lamination of the neve
snow, successively compacted by rain or melting, and then frozen (Geological
Sketches, p. 247), has been revived by Reid1 and Hess.2 Crammer accepts
this same view of the origin of these bands, and argues further that they repre-
sent shearing-planes along which the motion of the glacier proceeds.3 In his prize
essay, Uber das Gletscherkorn, p. 37, Emden advances the theory that these blue
bands were formed by the overflow from glacial brooks, infiltrated and frozen.
The view of Tyndall, that these blue bands result from pressure and, when
formed, are at right angles to it, had received very general acceptance. In
the former view the lamination is to be regarded as an organic part of the glacier;
in the latter, the banding is of secondary origin, and might not be present at all,
under certain circumstances. Tyndall's theory is set forth clearly in his Glaciers
of the Alps, chapter 31, and is summarized thus: "The ice of the glacier must
undoubtedly be liquified to some extent by the tremendous pressure to which it
is here subjected. Surfaces of discontinuity will in all probability be formed,
which facilitate the escape of the imprisoned air. The small quantity of water
produced will be partly imbibed by the adjacent porous ice, and will be refrozen
when relieved from the pressure. This action, associated with that ascribed to
pressure in the last section, appears to me to furnish a complete physical expla-
nation of the laminated structure of glacier-ice."

The Lefroy Glacier, being a regenerated and at the same time a parasitic one,
moving in a different direction from its host, furnishes an opportunity for testing
our two theories. In plunging 2,000 feet into the valley all traces of the original
stratification and lamination of the ne" ve" must be destroyed . Since the avalanches
of snow and ice occur only, or mainly, during a few months of the year, it may
be safely granted that layers of this material will be spread out, more or less
unevenly, about the base of the cliff, alternating probably with layers of snow
which falls directly into the valley, or is in part drifted there. The result of this
action will be to restore the stratification seen in the hanging glacier at the crest
of the precipice. It cannot be assumed, however, that anything like the
original lamination of the ice can be reproduced. Possibly around the margin
of the area covered by the avalanches, there might be built up a succession of

1 " The Relation of the Blue Veins of Glaciers to the Stratification," Comptes Rendus IX. Congres Geol.
Internal, de Vienne, 1903, pp. 703 to 706.

2 Die Gletscher, 1904, p. 175.

' Eis- und Gletsckerstudien. Neues Jahrbuch fur Min., Geol., und Pal., xvin. Beilage-Band, 1904,
pp. 105 and 106.

44 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

fine layers, but such a deposit would be of limited thickness since it would very
soon be pushed outward and beyond the reach of the snow dust. The bulk of
the avalanched ice would come down in great heaps, which could show neither
original nor acquired lamination. Granting that some of the avalanched snow
and ice would become finely stratified, we would expect it to alternate with much
more that was not, and with frequent layers, produced by the direct snowfall into
the valley, showing the typical neVe lamination. Furthermore, the position of
the Lefroy upon the Mitre is such that these laminae along the sides of the former,
as well as over the surface, should run across the valley and should be entirely
conformable with the strata.

Upon the other hand if the banding is of secondary origin and the result of
pressure against the valley walls, it should be entirely similar in adjacent strata
of the same character, exactly as found in ordinary glaciers not formed as is the
Lefroy, should be found near the sides of the valley and parallel with them,
and should show an utter disregard for the position of the Lefroy strata. In
ascending the Lefroy to apply our test we find a beautifully perfect and typical
banding upon the Aberdeen side, well shown upon the crevasse walls, under the
lateral moraine. The inclination of the blue bands is very steep, ranging from
72° to 90° and averaging 83°, as they dip downwards and into the body of the
glacier. These bands are continuous across the gently inclined strata and cut
them at a high angle. Plate xn, figure 4, shows the perfectly developed bands,
the margin of the glacier lying to the right, but does not give the desired view
of the strata. Toward the centre of the Lefroy these bands become obscure
at the surface, or disappear entirely, but are found again upon the Lefroy side,
between the collecting, region and the nose of Mt. Lefroy. So far as this feature
is concerned the Mitre and Lefroy seem to be a unit and the evidence is all in
favor of the pressure theory.

Near the nose of the Illecillewaet Glacier the blue band structure is very
perfectly shown about the sides, as seen in plate xin, figure i. There is no
lateral pressure upon either side and the bands conform with the valley floor.
Furthermore, they would be conformable with the strata, providing the latter
were present, but these have been destroyed, presumably at the ice cascade
farther up the slope. It may be maintained that at such a cascade it is only the
superficial layers that are . disrupted and that their fragments are destroyed
by melting, while the basal layers are preserved intact. This is undoubtedly
true, at times, but in the case of the Illecillewaet, the stratification, well seen
above the cascade, has been destroyed to the very base and it is difficult to
believe that the much more delicate lamination could possibly have escaped
destruction at the same time. Beneath this same glacier boulders are seen flut-
ing the under surface, as the ice is pressed against them and melted; this is shown
in figures i and 2, plate xxxiv. If the banding were simply the original ne"ve"
stratification the edges would be cut off squarely. Upon examining the ice
which has been pressed against a boulder there may be seen a set of bands curv-
ing about the stone, as though they had been there produced.

[ITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XIII.

i.— llliic hands giving rise to "dirt stripes," near nose of Illecillewaet Glacier. Bands
would here he conformable with strata if latter were present.

FIG. 2.— Contorted blue bands, Yoho Glacier. Supposed to indicate differential ic

flowage.

-Ice dyke filled with two tiers of hori/ontal ice prisms meeting
towards center.

FIG. 4.— Crevasse in Victoria Glacier, showing how superficial debris may
attain an englacial or subglacial position.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. , 45

Wherever observed, the phenomenon of blue bands suggested the structure
in rocks known as schistosity, rather than stratification, the bands thinning out
and overlapping. It is possible that they may still be due to pressure and yet
the ice may not have become liquid, as Tyndall supposed in order to account
for the scarcity of air bubbles in the blue bands, when compared with the whitish
vesicular ice in which they are embedded. There may be serious doubts as to
whether the pressure has been sufficient to produce liquefaction in all such cases
where the bands occur, and there is no reason for thinking that the crystalline
condition of the ice would be essentially different in refreezing. We may account
for the irregular and often contorted banding (plate xm, figure 2) by assuming
that differential movements have occurred in the ice mass since the bands were
formed. A double set might be induced without the complete obliteration of the
first. It is quite possible that, in the case of a glacier of the simplest supposable
type, having a very even bed and without the restraint of rocky walls or lateral
moraines, the original lamination of the neVe" would be preserved to the nose
and give rise to a certain type of "blue band." It would seem that such a type,
however, could be distinguished from the more common variety and that it
would lose in distinctness towards the nose. The writer had come to the con-
clusion that the diverse views held by investigators concerning the origin of
these bands were due to the fact that two very similar structures had been
studied under the same name, when his attention was attracted to the following
paragraphs written by Agassiz when glacial study was still in its infancy :

"Undoubtedly, in both these instances, we have two kinds of blue bands,
namely: those formed primitively in a horizontal position, indicating seams of
stratification, and those which have arisen subsequently in connection with the
movement of the whole mass. . . . With these facts before us, it seems to
me plain that the primitive blue bands arise with the stratification of the snow
in the very first formation of the glacier, while the secondary blue bands are
formed subsequently, in consequence of the onward progress of the glacier and
the pressure to which it is subjected. The secondary blue bands intersect the
planes of stratification at every possible angle, and may therefore seem identical
with the stratification in some places, while in others they cut it at right angles."
Geological Sketches, vol. i, pp. 260 and 261.

In this report the writer uses the term lamina by which to refer to these
' ' primitive blue bands ' ' arising in the neVe, and blue bands for the similar, but
essentially different, structure resulting, apparently, from pressure, or from some
other possible agency.

g. Ice dykes. These were well developed upon the lower Lefroy in the early
part of the summer, but became somewhat obscured as the season advanced.
They were found sparingly upon the Wenkchemna, but were not observed upon
the other glaciers studied. They consisted of gashes in the body of the ice,
apparently former crevasses, from two to fifteen inches across, which were filled
with columnar ice crystals. The columns varied in diameter from J to i inch
and stood at right angles to the crevasse walls, having thus an approximately

46 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

horizontal position. Very commonly the inner ends of the columns met and
interlocked at the centre, but sometimes they were simply attached by their
bases to the crevasse walls and left a space at the centre. Plate xin, figure 3,
will give some idea of the appearance of these dykes, although the individual
ice crystals could not be made to show in a general view. As a rule the columns
were straight, but sometimes curved and geniculated. The dykes were some-
times many feet in length, occasionally cutting across the walls of crevasses,
presumably younger in age. In certain cases similar columns were found filling
elliptical cavities in the ice, the crystals meeting at the centre. Such structures
as these were observed by Agassiz upon the Aar Glacier and described by him
in 1847 under the name of "glace d'eau." 1 Although their origin was not under-
stood he clearly saw that they resulted from the freezing of water in cavities in the
ice. The ellipsoidal cavities with their radially arranged columns were figured
upon plate vi of his atlas and described under the name " etoile de glacier,"
or " Gletscherstern" (p. 187). These structures probably arise from the freezing
of water-filled crevasses, moulins, and smaller cavities, the cooling surfaces being
the walls of the cavity, instead of the atmosphere. When a lake surface freezes
similar columns of ice are formed, with their main axes at right angles to the
cooling surface, and, hence, ordinarily vertical. In the case of these dykes the
columns also take a position at right angles to the surface of refrigeration,
but these surfaces now being vertical the columns assume a horizontal position.2
If the freezing is complete, the columns meet at the centre, the growth of the
columns proceeding at about the same rate, inward from the sides. Should the
water be drained off before the freezing is complete a space will be left at
the centre. They are probably formed in the early part of the season, while the
body of the glacier still retains some of its winter's temperature and after the
melting has proceeded far enough to supply the necessary water. After being
once formed they would persist through many seasons, although their upper
surfaces might be obscured by various agencies. Somewhat similar dykes were
sparingly observed upon the western side of the Lefroy but filled with granular
ice, instead of the ice columns. Obviously these have had an entirely different
history. The most plausible explanation is that they represent crevasses
which were filled with the granular ice avalanched from the hanging glacier
upon Mt. Lefroy.

2. SURFACE FEATURES.

a.. Superficial debris. The narrow valley through which flows the upper
third of the Victoria Glacier, permits the avalanches of snow and ice to distribute
rock debris over the entire surface. The most of this material' is derived from
the Mt. Victoria side, from which the avalanches may shoot completely across

t NouveUes Etudes et Experiences sur les Glaciers Acluels, 1847, Premiere Partie, p. 185, plate vi,
figures 14, 15, et 16.

2 While this report is going through the press the author has been enabled to study the valuable paper
of Crammer referred to upon page 43- Under the head of Leisten he describes similar structures (p. 104)
and ascribes to them the origin herejjiven.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XIV.

FIG. I. — Stony till, left lateral moraine, Victoria Glacier. Manufactured beneath
banging glacier upon Mt. Victoria and carried down with avalanches.

Whyte.

Devil's Thumb.

Bow Valley.

Lake Louise.

I''IG- 2.— Sharply crested left lateral moraine, Victoria Glacier. Moraine consists of a core of ice over which is spread
a relatively thin covering of clay, sand, and rock fragments.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 47

the valley. This being the region of accumulation, rather than melting, the
rock debris is almost completely enveloped in snow and remains temporarily
covered (plate v, figure 2). As the neve" is pushed beyond the snow-line upon
the glacier, surface melting begins and the rock fragments begin to make their
appearance at the surface. As this action continues the rock rubbish is con-
centrated more and more, forming an almost complete veneering over the lower
third of the glacier, completely obscuring the ice except where it has been incised
by the drainage streams. The most of this material is sharp and angular, con-
sisting of irregular fragments .of quartzite, sandstone, limestone, dolomite, and
quartz and argillaceous schists; in the main of Cambrian age. The ice lying im-
mediately to the west of the medial moraine has come from the Lefroy side of the
valley and being less well covered with de"bris has experienced more surface
melting. This depression thus formed, shown in the cross-section along the line of
plates (page 30), determined the position of the main drainage stream, previously
described. The effect of this debris, in general, is to retard surface ablation
and recession about the nose, so that the glacier attains a lower altitude than
would otherwise be possible for it under the present climatic conditions. So
far as we may judge from the ice front, the walls of the tunnels, crevasses, mou-
lins, and drainage streams, the Victoria is not carrying much englacial material.
A portion of this is in the position originally deposited in the neVe" and a portion
has worked down from the surface by means of the crevasses, as shown in plate
xm, figure 4.

b. Lateral moraines. Along the margins of the upper Victoria and Lefroy
conditions are especially favorable for the reception of rock detritus, both from
the action of the ice avalanches and from the various weathering agencies that
are operating upon the overtowering cliffs. Material derived from the cliff walls
will ordinarily be sharp and angular, but may rarely show a single glaciated
face, produced when in its original position during an earlier stage of glaciation.
Most of this has been pried loose by the water in the seams and joints expanding
in the process of freezing. The material carried by the hanging glaciers is almost,
if not entirely, subglacial and has been subjected to severe abrading action
between the ice and its rocky bed. Boulders, cobbles, and pebbles have had their
corners and edges partially rounded, have had their faces bruised, gouged,. and
irregularly scratched, and are embedded in glacial sand and clay, of a bluish gray
color. This ground-morainic material, mixed indiscriminately with that from
the cliffs, is heaped up along the neVe" margins, embedded in snow and ice.
Moved slowly along, very slowly compared with the central portions of the ne"ve",
the quantity is augmented and by the time the snow-line is reached there is
formed a thick band of this debris covering the margin of the ice. Protected
from the action of sun and rain more effectually than the general surface
of the glacier, in spite of its debris covering, the ice beneath melts less rapidly
and the marginal material is gradually elevated, with reference to the general
surface. About the sides of this marginal ice ridge the debris slides and rolls
down, allowing the less well protected ice above to melt into a sharp crested

48 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

ice ridge, with a veneering of rock rubbish, the whole looking like a great rail-
road embankment, as seen in plate xiv, figure 2. The ordinary visitor is
scarcely prepared to admit the existence of the ice core, which constitutes, in
reality, the main bulk of the ridge (see plate XL, figure i, from the Asulkan
Glacier). In this way are formed the lateral moraines. Should the glacier com-
pletely disappear from the valley by melting it is obvious that the lateral
moraine would be gently set down along the side of the valley, forming a ridge,
but of insignificant proportions compared with its apparent bulk upon the
glacier.

Upon the western margin of the Victoria, the glacier's left, opposite the en-
trance of the tributary, there occurs a considerable mass of angular debris, contrib-
uted from the Mt. Victoria side of the valley. Most of it is arranged in three or
four somewhat poorly defined ridges, parallel with the margin of the glacier. A
sudden contraction occurs here in the breadth of the glacier (see plate in), and
there is continued a prominent, sharp-crested ridge for one-quarter mile, marking
the margin of the glacier and losing gradually in height (plate xiv, figure 2 ) . This
portion of the left lateral consists almost entirely of ground-morainic material
derived from the hanging glacier upon Mt. Victoria (plate xiv, figure i). Soaked
with water after heavy rains, mud flows occur, upon the surface of which cobbles
and small boulders are slowly moved down the marginal slopes, thus reducing the
covering of the ice core and permitting further melting. Along the base of Mt.
Whyte there are found two small moranic ridges, consisting mostly of angular
material, from which the ice has withdrawn rather recently. They appear to be
the continuation of the two outer ridges which farther up-stream rest upon the
ice itself.

The right lateral of the lower Victoria is derived entirely from the right lateral
of the double tributary, already described. It consists at first of two high, very
sharply crested ridges, mainly of ground moraine, which can be traced around
into the great accumulation dumped at the base of Mt. Aberdeen by the parasitic
Lefroy Glacier (plate vin, figure i ; plate xv, figure i). The angular material has
been derived mainly from Mt. Aberdeen, while the ground moraine comes from
the hanging glacier of Lefroy, as previously described. The inner of the two
morainic ridges is being destroyed by sliding and mud flows into the depression
between it and the near-by medial moraine. In places it has become so sharp
that only with the greatest difficulty can one maintain a foothold upon its crest.
About 2,000 feet back from the nose, an outer third ridge makes its appearance
(plate xv, figure 2), and together the three pass around and over the nose,
separating into minor ridges and mingling with those of the medial and frontal
moraines (plate iv, figure i). The lower portion of this moraine has the appear-
ance of composure and comparative stability, giving support to moss, ferns,
alpine plants, shrubs, and evergreens. One Lyall's larch was noted 8 feet high
and 2 inches in diameter at the base.

Since the upper Victoria receives relatively little material from Lefroy, the
right lateral above the tributary is rather meagre, and inconspicuous. As

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XV.

Aberdeen.

Mitre.

Lefroy.

Kit;. I.— Gruuiid-morainic material manufactured beneath hanging glacier upon Ml. Lefroy

and carried across Mitre Glacier by parasitic Lefroy Glacier. Beginning of

ridges seen below in figure 2.

Lefroy.

Victoria.

if ic. 2.— Right lateral and medial moraines of Victoria Glacier. Longitudinal ridges in lateral are well shown, the

work of the parasitic Lefroy Glacier.

OF THE

f UNIVERSITY

OF

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 49

previously pointed out, a small amount of ground moraine escapes from being
carried across the valley and moves down in the left lateral of Lefroy. In
addition to this there is a large detrital cone, with its base resting upon the
ice, and slowly dealing out morainic material as the ice moves down the valley
(plate iv, figure 2). The covering of the general surface of the lower Victoria with
rock debris prevents a great amount of differential melting, so that the lateral and
medial moraines attain no great height.

c. Medial moraine. Owing to the stream-like nature of the flow, the left
lateral of the Lefroy and the right lateral of the upper Victoria unite at the nose
of Mt. Lefroy into a single medial moraine. This is at first a poorly defined
ridge, but it becomes higher and broader as it moves across the valley from
which emerges the tributary and serves as a divide for the two main drainage
systems (plate iv, figure i). Owing to the small volume of ice delivered to the
Victoria by the double tributary, the medial moraine lies close to the right lateral,
being separated at first by a deep depression, shown in plate xv, figure 2, which
gradually disappears below as the two moraines merge. The western slope of the
medial becomes long and gradual in the lower part. The entire length of the
moraine is about 7,500 feet. Toward the nose it broadens as shown upon the map
and in plate iv, figure i and becomes poorly defined, implying a sluggish condition
of the ice upon which it rests. Its crevassed condition in the neighborhood of the
line of plates was described upon page 37. Owing to the source of the material
above noted the moraine contains a certain amount of ground-morainic material,
but the bulk of it is angular and consists of quartzites, sandstone, schists,
dolomite, and limestone. Some of the blocks show algae, tracks, lingulas, and
bryozoan-like stems. It has practically all been derived from Mt. Lefroy.

d. Terminal moraine. Although the front of the ice at the nose is in a con-
dition of halt, the ice is practically stagnant and no frontal moraine has yet been
formed (plate v, figure i). Along the oblique ice front the retreat has been gradual
enough to distribute the superficial and englacial rock debris somewhat uniformly
over the valley floor and there has thus been formed no prominent ridge, as shown
in plate vm, figure 4. The apparent heaps seen at the right, alongside the face, still
contain a core of ice, which will eventually melt and allow the rock to settle upon
the valley floor. A small ridge, from 100 to 125 feet back from the ice, indicates
a somewhat recent short period of halt, perhaps but one or two decades ago.
It is quite probable that this halt was contemporaneous with that of the Illecille-
waet, which closed in 1887. Between the oblique front and the nose conditions
have been favorable for the formation of a somewhat poorly defined terminal
moraine, i. e., the front has been in a condition of halt while the ice was moving
forward and dumping its load of angular debris. Two of the ridges that pass
across the glacier, just back from the nose, extend off the ice upon the terminal
moraine, without interruption, testifying still further to the sluggish condition
of the ice about the nose. The medial moraine has introduced some ground-
morainic material into the mass which has furnished a foothold for vegetation.
Spruce and larch are climbing up the slope, the largest of the former showing

50 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

70 rings of growth and of the latter 77 rings. It is over this morainic heap that
the drainage brook from the glacier cascades. The swift stream and its load of
hard angular sediment have a perceptible rounding effect upon the corners,
edges, and faces of even the hardest quartzites. This effect was most strikingly
shown in quartzite boulders lying in the bed of a glacial stream coming from
the Asulkan ridge.

e. Dirt bands. Under this term there was described by Forbes, in 1843,
(Travels through the Alps of Savoy, p. 162), a superficial feature of certain
glaciers which is of much interest and. possibly, of much importance. It is
found in those glaciers which change their slope sufficiently to give rise to a
distinct system of transverse crevasses, not necessarily to a cascade or ice-fall.
The phenomenon was not understood by Forbes himself and, by various writers
since,1 has been confused with the dirt zones, described upon page 39, which are
the outcropping edges of variously marked strata. It is to the keen observation
and shrewd interpretation of Tyndall that we are indebted for the true explana-
tion 2 of the feature. The Victoria and the Lefroy glaciers furnish an excellent
opportunity for the study of dirt bands under very simple conditions, as
well as the dirt zones for comparison. The two types of structure may be gotten
upon the same photographic plate and are well shown in plate iv, figure 2. In
very simple form the dirt bands may be seen cutting across the dirt zones upon
the lower Lefroy, owing to the abnormal position of the latter. Under ordinary
conditions the two would be more or less conformable and possibly difficult
to separate.

The typical dirt bands are of such a nature that they can be seen most strik-
ingly at a distance of a half-mile or more from the glacier and at a considerable
elevation above it. When once seen, however, it is possible to locate them in a
very general way while upon the surface of the glacier itself. In the summer of
1904, from the summit of the Devil's Thumb, which overlooks the Victoria Glacier
from a height of 8,000 feet, there could be counted 23 soiled streaks passing across
the glacier. Beginning near the crest of the ice slope opposite the nose of Mt.
Lefroy, the bands were narrow, straight, and extended nearly across the glacier.
They showed so dimly that there was uncertainty in regard to the count, until
they had been gone over a number of times. Upon the face of the slope they
became" more distinct, curved so as to be convex down-stream, and correspondingly
shortened. A few of them could be traced around into the transverse crevasses
which had not been completely closed. Beyond the foot of the ice slope the
bands became still better defined, especially upon the southern, or up-stream
margin, narrower, more closely .placed, and changed their shape from arcs of
circumferences to hyperbolas. Towards the lower end of the series the bands
became much shorter, the arms extending into and blending with the surface
debris, and their apices appeared to mark the locus of maximum surface velocity

' Agassiz, Geological Sketches, vol. i, pp. 244 and 254; Russell, Glaciers of North America, p. 43; Hess,
Die Gletscher, p. 169.

* Glaciers of the Alps, pt. n, chapter ,26.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XVI.

FIG. I.— Formation of Forbes's "dirt bands," Deville Glacier, Selkirk*. From summit of Mt. Fox (10,572 feet),
looking eastward. Photographed, 1902, by Arthur O. Wheeler.

FIG. 2. — Forbes's "dirt bands," Victoria Glacier. Photographed from the Lefroy Glacier, July, 11)04. Often confused

with "dirt zones." Compare figure 2, plate xi.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of the ice. Finally the bands mingled with the superficial rock covering of the
glacier and were lost. Standing upon an individual band the dark color seems
to be imparted by the fine dust and sand and not by the coarser debris. In
September, 1905, owing to the excessive melting of the summer, the bands stood
out with unusual clearness, so that they were photographed from the side of the
Lefroy Glacier, as shown in plate xvi, figure 2. By signaling to an assistant,
the well defined up-stream margins of 19 of the bands were located by erecting
small cairns of rock, and their distances apart, in the line of their apices, were
later measured (see map, plate in). The results were as follows, beginning near
the foot of the ice slope. The average interval between the bands is 97 feet.

Band No. i -----

v»-i
" . ---"~

No.
" No.

feet
feet

" _---i26 feet
No. 4~--VL"_V

-.---124 feet
No. <;----:""

~~~---~-*l3 feet

" No -T-::''''

" No. 8 ----..-_•_"_"

1 No.
" No. 6

feet

i oo feet

_ feet

No. g-~:~

•-_-_-_-_— 75 feet
" No 10 ------

Band No. 10 — __

~

1 No. n-=r;

~

No. I2-.r^-_"~

No. 13-^;^"

.. M 7-

No. 14-c-^VL

No. 15---'.
1 No. i6--V_1

1 No. i7'~:~"

~

" No. 18--=:;:

_
No. 19

IV--— 100 feet
"__V--— -86 feet
V-V----88 feet
l_-r-— -57 feet
"rir-- 81 feet
"_~V--^- 66 feet
.II-- -83 feet
."_"--- 83 feet
'.'.'---" 45 feet

For reasons to be given later the writer believes that the intervals between
these bands mark the annual progress of the ice down the slope, as conjectured
by Tyndall, and offers the following explanation of the phenomenon. As the ice
of the glacier is pushed over the crest of the ridge in its bed, which is responsible
for its change in surface slope, there is formed successively a series of transverse
crevasses, as explained upon page 36 of this report. The distance between
these crevasses will be determined mainly by the thickness of the ice and the
change in its angle of slope. Since the glacier is moving forward in winter as
well as summer, although at a less rate, these crevasses must originate at all
seasons of the year. Those which have been formed in the late fall, or winter,
upon passing down the slope will be perfectly healed, since their lips have ex-
perienced practically no melting from the sun's action. The opposite crevasse
walls come slowly together, refreeze, and leave no visible scar in the ice. Those
crevasses, however, which have formed in the late spring and summer have their
lips much rounded by the sun's rays. If the glacier is moving northward as in
the case of the Victoria and Lefroy, the northern, or down-stream lip of the
crevasse will receive the maximum effect, the southern comparatively little.
Should the glacier be moving southward, the northern lip of the crevasse would
still be the one most strongly acted upon by the sun, but in this case it would be
the up-stream side. Glaciers flowing east or west, and having their transverse
crevasses in an approximately north-south position, would have their crevasse

52 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

walls affected more evenly, unless surrounding mountain cliffs interfered. In
the healing of such crevasses there would be left a depression, representing the
sun's action upon the lips of the crevasse, not simply for one season but through
a series, and in this depression the wind-blown dust would collect and the fine
de"bris would be washed by rain and melting ice from the adjacent portions of the
glacier, rendering it lighter by. contrast. Owing to the more rapid central
movement of the ice the bands, at first nearly straight, will begin to curve down-
stream and become more and more sharply bent, their apices marking the locus
of maximum surface motion. Between them will lie swellings, or ridges, having
the same general form of the depressions, from which much of the finer dirt has
been removed. These ridges and intervening depressions may be very inconspicu-
ous, as upon the Victoria, or they may become very prominent, as shown upon the
Deville Glacier in the Selkirks, forming what Forbes termed ' ' wrinkles ' ' (plate x vi ,
figure i). They mark that portion of the ice which passed the crest of the
slope in the late fall and winter and appear as ridges, partly because of the severe
compression to which the ice is subjected and mainly because the adjacent ice
has been lowered by melting. Owing to the more rapid movement of the ice
down the slope the bands will be farther apart and less well defined, than after the
more gentle slope below has been reached and the ice is subjected to longitudinal
compression. Upon this more gentle slope they have a better chance to catch
and retain the fine debris. Since the sun's action was more powerful at the
center of the crevasse, the depression is greater at the apex of the band and
persists after that of the extremities has been finally lost by surface melting.
In consequence the 'bands become shorter and shorter and lastly disappear,
when ablation has reduced the surface to a general slope and the fine debris is
redistributed. Very often it must happen that instead of a single crevasse
being formed during the season of melting there would be formed a series of
them. Upon a steep slope of the Asulkan they seem to be formed in pairs as
shown in plate xvii, figure i, in which it is seen that the ridge of ice separating
two adjacent crevasses is acted upon from either side and lowered, assisting in
the formation of the depression. The crevasses that are forming the depres-
sions, preparatory to the reception of the dirt, may be traced around to the
almost healed crevasses at the left, while between them are seen traces of crevasses
that have healed with practically no marginal melting. These are presumably
those which opened and closed soon enough to escape the rounding action.
If the surface slope is too great the depression produced in the ice may not be
sufficient to retain enough dust to bring out the series distinctly, as is the case
with the Asulkan just noted. Study figure i, plate xxix, from the Yoho Glacier.
That the method of formation of these dirt bands is essentially as outlined
above admits of no doubt. The question as to whether they are produced
annually, or at irregular intervals, needs to be investigated. The average inter-
val of those bands originally described by Forbes upon the Mer-de-Glace was
711 feet. Opposite his station D the interval was 667 feet (Travels through the
Alps of Savoy, p, 165). In a postscript to his volume, p. 420, he gives the move-

o

J3

c
o

rC
O

O 3

a «<
U

=

t/3
I

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 53

ment of one of the lateral points opposite station D as 483 feet for the year
and suggests that "the movement of the center is probably, at least, two-fifths
greater, corresponding closely with the intervals of the /dirt bands' of the
glacier." Although the size of the intervals in this series differs plainly to the
eye, still Forbes states that the difference for any one interval is probably not
a tenth of the mean. From the same point of view as that used by Forbes in
1842, Tyndall counted upon the Mer-de-Glace, 17 years later, exactly the same
number of bands and remarked: 'The entire series of bands which I observed
with the exception of one or two, must have been the successors of those
observed by Professor Forbes; and my finding the same number after an interval
of so many years proves that the bands must be due to some regularly recurrent
cause." In Chapter xxxn of his Glaciers of the Alps, Tyndall has described his
"white ice-seams," the "bandes lactees" of the French and "weissen Blatter"
of the Germans. These are due, in part, to the filling of transverse crevasses,
left open during the fall, with snow and then its later compression into a white
vesicular ice. Since, in general, these crevasses would be those which had been
acted upon by the summer sun, they would be the counterpart of the dirt bands
under discussion. Seve" found the average interval for these white seams upon
the Boium Glacier, in Norway, to be 218 feet and that this represented also the
average forward annual movement.1 So far as the Victoria Glacier is concerned
we have not sufficient data at hand to settle the- question of the annual character
of the dirt bands. At the line of plates, about a third of a mile below, the
maximum annual movement of the ice was found to be 65.85 feet. The average
annual interval for the lower half of the series is 76.56 feet, which is about what
would be expected in the way of annual ice movement, when compared with the
above. We should also expect the movement to increase as we approached
the crest of the ice slope. So that the actual and relative spacing of the bands
very strongly suggests their annual character. If due to some "regularly
recurrent cause," as Tyndall suggests, this cause must recur with the seasons.
We are, however, not entirely without evidence that the intervals between
the dirt bands indicate approximately the annual movement of the ice. As
pointed out upon page 30, the Messrs. Vaux marked the location of a large
boulder upon this portion of the glacier July 26, 1899. From range lines, one
year later, they determined that the boulder had moved forward 147 feet. In
September, 1905, this boulder was found opposite the gth band of the series
given upon page 51. In 1899 it should have lain opposite the 3rd band and,
if the motion there had been the same as it was in 1904-5, it should have moved
in 1899-1900 the distance of 126 feet. The previous year it should have moved
174 feet. My field notes say that the second and third bands were indistinct,
so that there is strong probability that the three intervals between one and four
may not have been properly distributed. The average for the three is 153 feet,
which agrees very well with the actual observed motion of the boulder. If the
dirt band intervals are an approximate indication of the annual ice movement

1 Quoted from Helm's Gletscherkunde, p. 140.

54 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the Vaux boulder had moved downward in 1905 from its original position some
676 feet, or at an average rate of about 113 feet per annum.

/. Dirt stripes. Somewhat closely related to the dirt bands just described,
so far as their method of formation is concerned, are the fine streaks of dirt seen
along the margins of most glaciers, sufficiently free from surface debris. They
may be found, however, anywhere upon the glacier that the blue bands are
well developed, reach the surface at a fairly steep angle and are being subjected
to surface melting. The blue bands, being composed of relatively firm, compact
ice, are more resistant of the sun's action, than the vesicular ice in which
they are embedded and project as delicate ridges, separated by narrow
furrows. Into these furrows the wind-blown dust settles and is washed
from the adjoining ridges, forming narrow, parallel dirt streaks, or stripes.
When well developed, as upon the Lefroy, the glacier has the appearance of
having been swept with a coarse wire broom; the strokes having all been long,
regular and parallel. The dirt stripes mark the position of the vesicular bands
in the ice and the lighter streaks between the position of the blue bands them-
selves. In this way the banding is clearly shown at the surface, whereas, other-
wise, it might be obscure. Views of these stripes have already been shown in plate
xn, figure 4 and plate xm, figures 1,2, Sometimes they run down the face of a
crevasse wall (plate xn, figure 4), as though they might be something more
than a superficial feature, but a little chipping of the ice shows plainly that they
are not. After they have once been formed the dirt stripes will absorb the
sun's heat and still further emphasize the small furrows. Running, in general,
lengthwise of the glacier these furrows become the sites of minute rills which
have a tendency to clear away the fine dirt, as fast is it collects. For this
reason, as well as because of the nature of the banding -itself, the individual
stripes are not continuous for any considerable distance. They are sometimes
so closely placed that 10 stripes may be counted within the distance of an inch,
but are usually considerably coarser.

g. Dust and pebble wells. Where small pebbles, or patches of fine dirt,
often black from the presence of organic matter,1 are thinly distributed over
the surface of the ice, heat is absorbed and the ice immediately beneath is melted
more rapidly than the surrounding ice. Cavities are thus formed with vertical
walls, which for a time retain the water. They sink into the ice for a few inches,
until protected from the direct rays of the sun by their own walls, when further
melting would be delayed until the general surface was lowered sufficiently
to allow the sun to again reach the foreign matter at the bottom. Such wells
are shown in plate xvn, figure 2. Although their depth at any one time is seldom
greater than a finger's length, still in the course of the season their total length
would be 9 to 10 feet upon the Victoria and Lefroy. A thin film of water often
freezes at night over the surface and then thaws out promptly when again
exposed to the sun. After thus freezing the water is at times drawn into the

'A sample collected from the Illecillewaet in 1903 contained 14 per cent, of organic matter, enough so
that when set away moist in a warm room it soon became offensive.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 55

glacier, by means of the capillaries developed between the granules, leaving
the well free from water, but with its ice cover. Where pebbles, or small
dirt patches, are abundant, as shown in the last figure, the ice between the
adjoining wells is melted more rapidly by the sun than it would ordinarily be,
forms minute pinnacles and appears whitish and spongy. In this way the lower-
ing of the general surface of the glacier by ablation is accelerated. By keeping
itself thus at the bottom of a small well the dirt of these small patches is pre-
vented from being blown away, or washed away, and thus it is possible that the
same well may persist through, not only a season, but a succession of seasons.
Should the well, however, collect additional dirt, beyond a certain limit, this
excess would then protect the bottom of the well from further melting, the
adjoining ice would soon be lowered below the bottom of the well and the well
would be literally turned wrong-side-out. Where one has a few days to spare
about the same glacier an interesting experiment would be to sift dirt into a
group of typical wells, filling them to varying depths, and observing the result.
Such an experiment may easily be performed upon a snow bank of sufficient
depth, when it is being strongly acted upon by the spring sun. It would prepare
the way for a clear understanding of the next three surface features to be
described.

h. Debris cones. When the amount of dirt, sand, gravel, or rock debris,
is sufficient to protect the surface of the ice from melting, or to even partially
protect it, over a limited area, the surrounding ice surface will be lowered
more rapidly than that beneath the protecting material and the debris will
begin to be elevated, with reference to the neighboring surface. The loose
debris will slide, or roll down about the side, exposing the edges and corners to
the melting action of the sun, allowing still more sliding of the debris and still
further melting. The ice core will finally assume the form of a ridge, cone, or
mound, with its thin veneering of foreign matter, as in the case of the lateral and
medial moraines already described. The companion figures 3 and 4, plate xvn,
show the structure of a small gravel cone, only 15 to 16 inches in height; figure 3,
as it was found upon the ice, figure 4, after the gravel upon one side had been
washed off to show the ice core. It is seen what a thin covering will suffice to
bring about the result. Depending upon the nature of the covering they are
known as dirt, sand and gravel cones, and boulder mounds, and they may vary
in height from a few inches to many feet. In plate xix, figure i is shown
a mound upon the Wenkchemna Glacier, estimated to be 80 feet high. This
pile of rock rubbish was either dumped in a heap by an avalanche, or collected
in the bottom of a lakelet, as described by Russell for the Malaspina in Alaska.1
Cones of all types, varying in height, from a few inches to 12 or 15 feet are
to be found upon the Victoria in the region of maximum melting. They may
persist from one season to another, but there is a limit to the height to which
any particular cone may attain. As the height of the cone grows the lateral
surface is increased, over which the debris must be spread in order to suffi-

1 Glaciers of North America, p. 115.

56 GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

ciently protect the. ice. When this covering becomes too thin, or when it is
blown off, or washed down the steep slopes by heavy rains, the ice core becomes
exposed and rapid melting ensues, resulting in the destruction of the feature.
In case the surface covering is distributed about the base and the pure ice core
exposed, the melting does not cease when the general level is reached but con-
tinues more rapidly than the surrounding ice to which has been transferred the
debris. Instead of a cone, we may now get a basin-shaped depression, which is
gradually extended laterally by melting, and into this depression the original
material may again slide and be collected at the centre until there is sufficient
to prevent further melting. An interesting and instructive experiment, in
connection with that suggested upon the dirt wells, is to wash down the gravel,
sand or dirt, from a collection of small cones, mark the location, and watch the
changes from day to day.

i. Glacial tables. In the case of a single rock fragment, of sufficient size,
resting upon the ice over which surface melting occurs, protection is afforded
the ice immediately beneath. As the result of the more rapid melting of the
surrounding ice the rock is relatively elevated upon a pedestal of ice and there
results what is termed a "glacial table"; as seen in plate xvm, figure i. As the
rock is elevated a short and narrow ridge of ice lying to the north of the pedestal
(observe the shadow in the figure) is protected from the noonday sun, so that
viewed from the east or west the pedestal is unsymmetrical. This lack of
symmetry is further emphasized by the undercutting action of the rays of the
noonday sun upon the southern side. Some observations were made with a
view of discovering the lower limit of the rock fragments that were capable
of furnishing the protection necessary to form tables. The following were
found forming low tables, or starting to form them. It is obvious that the
color and nature of the rock would both have their influence in determining
the effect upon the ice.

Dark gray limestone, 12x12x4.5 inches.

13 x QX 3 "

Light " " ii x 6x3.5 "

Reddish quartzite, 10x8x2.5
Rusted limestone, 9 x 4.5 x i
Dark limestone, 8x4x2.5 to 3 inches.

In the case of the last specimen the thicker end was found to be protecting,
while the thinner was inducing melting. Owing to the undercutting action
of the sun's rays blocks of this size can form only low tables. Larger blocks
may rise to a height of three to five or six feet upon the Victoria, the latter
heights being unusual. They may persist from one season to another but there
is a limit to the height which any particular table may attain, determined mainly
by the size and shape of the rock. As the rays undercut, mainly upon the
southern side, the block begins to lean to the south and finally topples off in
that direction (plate xvm, figure 2). The remnant of the pedestal is removed by

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE — SHERZER.

PLATE XVIII.

FIG. I. — Glacial table, Victoria Glacier, looking southwest. Observe undercutting action of
rays upon south side and shadow cast by the rock upon north side, with
resultant ridge of ice.

IMG. 2. — Dethroned glacial table, Victoria Glacier, looking northeast. The boulder fallen
to the south by undercutting action of sun's rays.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE — SHERZER.

PLATE XIX.

FIG. I. — Boulder mound, Wenkchemna Glacier, illustrating the protective effect of rocky debris. Estimated to be

eighty feet in height.

Whyte. Devil's Thumb.

Bow Valley.

Fairview.

FIG. -2. — Surface lakelet, Victoria Glacier, resulting from lack of debris protection. Enlargement towards right is still
in progress by melting, but has practically ceased towards the left, owing to the rocky cover.

IIVERSITY 1

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 57

melting, the block settles into its position of equilibrium and the making of a
glacial table begins anew. In exceptional cases the undercutting of the pedestal
may be done by a surface stream. In the case of 25 tables selected at random,
it was found that the longer of the horizontal axes of the pedestals had an average
magnetic bearing of N. 36° W., or 11° W. of true north. With a larger, or a
different, series, the average would probably be more nearly the true north.

/. Surface lakelets. Upon the middle portion of the Victoria, western side,
where the ice is presumably quite stagnant, there occurs a series of surface
lakelets, the crater-like basins of which have been hollowed in the ice. The
largest of this series is somewhat elliptical in form, 200 feet long by 100 feet
broad (plate xix, figure 2), and filled with deep blue water in which miniature
ice-bergs may be seen floating about. The southern and eastern banks of the
lakelet are from 12 to 20 feet high and under cut, apparently by the melting
action of the lake water. The northern and western banks have been acted
upon more strongly by the sun, causing them to recede and the debris to slide
down until the margins of the lake are filled and the ice banks veneered sufficiently
to retard melting (plate xix, figure 2). These banks are as steep as the debris
can stand and from 25 to 30 feet in height. From the still steeper ice walls
the gravel and small boulders are splashing into the water with a sound suggestive
of considerable depth. The lake has no visible outlet and persists from season
to season. Several similar lakelets, but smaller, occur in the same vicinity,
some having their sides completely veneered with rock debris, which has checked
melting and allowed the lakelet to become almost dry.

These lakelets may have originated in marginal crevasses and been enlarged
and shaped by melting, or they may have originated by surface melting over
certain limited areas less well protected by debris covering. In the preceding
discussion of debris cones it was shown how miniature basins might originate.
In a stagnant portion of the glacier it is possible that the basins of such lakelets
might arise in a similar manner from such a mound as that figured from the
Wenkchemna Glacier (plate xix, figure i). The rock debris rolling or sliding to
the base would leave the cone sufficiently bare to permit rapid melting to a depth
at which the marginal d6bris would begin to slide back again. The accumula-
tion of the de"bris in the basin would check further melting at the center, while
the surrounding ice has lost, in proportion, a part of its d6bris. As observed by
Russell upon the Malaspina, the surrounding ice would be lowered until the basin
disappeared and what had been the centre of the basin would become the crest
of a boulder mound. If conditions remained favorable, i.e., sufficient thickness
of stagnant ice and continuous surface ablation, the sides of the mound would
become more and more steep, as it gained in height and the time would come
when the bulk of the debris would slide or roll to the base, the ice core would
be removed to the general level and a new basin would be started. For the
larger lakelets the complete cycle would probably have to be reckoned in decades
and centuries.

k. Rock reflection (?). A final surface feature remains to be described, al-

$8 GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

t

though a satisfactory explanation has not been found. Observations were begun
upon the Victoria in 1904 before the winter snow had disappeared from the lower
half of the glacier, which is ordinarily bare in the summer. It was noted that as
the surface boulders protruded through the snow a larger percentage of them
showed melted areas upon their northern sides, the shape and size of which sus-
tained a certain relation to the breadth, height, shape, and possibly position of the
boulders themselves. Over the melted area the snow was removed, in whole
or part, to the soiled surface of the glacier, so that the feature shows with
much clearness in the photographs secured. The phenomenon was seen
upon the Lefroy, as well as the Victoria, and sparingly upon the Wenkchemna
in midsummer, near the neVe" lines. The same thing was also seen upon the
snow of an avalanche which had descended from Mt. Whyte and carried along
some small rock fragments, which were scattered over the surface. The block
shown in plate xxiv, figure 4 is a gray quartzite standing 10 inches high and is 29
inches broad. The melted area has the same length as the rock and has a corre-
spondence in outline. The farther right hand corner of the rock is somewhat
lower than the general surface and the corresponding corner of the melted area
is seen to be rounded and incompletely melted. Boulders showing the phenom-
enon were not hard to find upon the Victoria, but were very abundant. The
north-south axes of ten of the areas, selected at random, gave an average
magnetic reading of N. 25.5° W., with less range than was shown in the case of
the glacial tables. The magnetic decimation of the region, as obtained by the
Canadian Topographic Survey, is N. 25^' E., so that these areas are oriented
with reference to the noonday sun, and might have been used for determining
approximately the meridian and the magnetic declination. The natural infer-
ence is that the phenomenon is due to the reflection of heat from the surface
of the boulders, this action being at a maximum when the sun is upon the meridian.

3. FORMER ACTIVITY.

a. Terminal moraines. Between the present poorly defined terminal moraine,
described upon page 49, and Lake Louise there occurs a series of ancient
terminal moraines which furnish evidence of the glacier's former extent and
greater activity. The first two of these moraines are remarkable in that they
consist of massive blocks of quartzite and sandrock, tumultuously heaped
together and without the usual filling of gravel, sand, and clay. The position of
these is shown upon the map, plate in. The spaces between the great blocks
enable man, or animals, to creep in between and under them and they form
an ideal home for the marmots. For moraines of a somewhat similar appear-
ance, although probably different history, in the Mount Ktaadn region Prof.
R. S. Tarr has used the expressive term "bear-den moraine." ' The. inner of
these two moraines extends obliquely across the valley from the present nose,
being partially overridden by the glacier and along the side of the valley nearly
parallel with the oblique front. It consists of what were originally massive

1 Glaciation of Mount Ktaadn, Maine. Bulletin Geol. Soc. of Amer., vol. xi, 1900, plate 37.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 59

|

blocks of quartzite, sandstone, and schist tumultuously heaped into a ridge 30 to 40
feet high. The sandstone and schist have undergone considerable disintegration,
in place, forming more or less soil which supports a growth of moss, shrubs,
spruce, and fir. The rings of growth, in the spruce and fir of the Lake Louise
Valley, were found to average 0.884 rnm., and, in the case of the averages for
individual trees, to range from 0.5 1 mm. to 1.26 mm. As the tree matures the new
rings are excessively thin; owing to the reduction in the relative amount of leaf
surface, the scant precipitation, the short growing season, and the lack of direct
sun light in a deep valley with a north-south trend. The largest tree found upon
this moraine gave a circumference of 221 cm., at a distance of 50 cm. from the
base, and should be approximately 400 years old. The material of this moraine
is arranged in two main heaps, between which the glacial brook passes, that
upon the eastern side having come from Mt. Lefroy and that upon the west
from Victoria. After the formation of the moraine the glacier retreated up
the valley to a point greater than that occupied by the present nose.

From 300 to 800 feet farther down the valley there occurs the second of this
type of moraines, the blocks consisting very largely of quartzite, lichen-covered
and moss-grown, but not disintegrated. As in the case of the preceding moraine
the blocks are disposed in two heaps, upon either side of the glacial brook,
the bulk of it lying to the west, where it forms an oblique ridge 700 to 800 feet
long, with a maximum breadth of 300 feet and a height of 70 to 80 feet. Made
up of such coarse blocks and with no filling of sand, gravel, or clay the whole
presents a very imposing mass and impresses one with the possible importance of
glaciers as geological agents. The largest block seen had split in falling, measured
31x25x15 feet, and was estimated to weigh 970 tons. The blocks are generally
sharp and angular and have not been subjected to stream or ice action. They were
carried either upon the ice or within it and show almost no signs of ice abrasion.
An occasional single face is glaciated but in such a way as to show that this was
done when, in its original position, it formed the face of the cliff. Owing to the
lack of soil the growth of shrubs and trees is scant. The largest spruce seen
upon the moraine itself was estimated to be 450 years old, while another just
beyond the outer edge was estimated at 580 years. Upon the side of Mt. Whyte
just at the line of plates, there is a large collection of similar blocks, and appar-
ently of equal age, which appear to have become stranded here while the others
were undergoing transportation. It should be noted that the present Victoria
is entirely incapable of making such a moraine now, no matter how prolonged
the halt. The present terminal moraine, and the oldest of the series to be
described, are essentially alike but very different from these great block, or
bear-den moraines. The cliffs which contributed the bulk of the material,
so far as we may judge from their favorable situation and the location of the
blocks, have a north-northwest trend and the blocks fell from them to the
eastward.

Down the valley, a distance of one-quarter of a mile, there occurs a double
detrital cone, derived from the opposite mountain slopes of Mt. Whyte and Castle

6o GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

Crags. The slide from the latter slope is still in process of formation and consists
of freshly broken angular fragments. That from Mt. Whyte is older and covered
with spruce, fir, and willow. This material overlies, and almost conceals, a
triple moraine, composed of boulders, gravel, and clay. The inner and higher
ridge of the three curves regularly across the valley and in one place about 60
feet of it have been cut away by the drainage brook. It here has a breadth of
100 feet, a height of about 20 feet, and consists largely of a yellowish, stony
till. Between it and the outer bear-den moraine the drainage stream has filled
in with sand and gravel, forming a nearly level flat. Lower down the slope
and 400 feet distant, there is another morainic ridge, approximately parallel
with the first, but made up more largely of boulders while 350 feet farther
on there is a third ridge containing a higher percentage of yellowish clay.

b. Lake Louise basin. This lake is roughly elliptical, with a major axis of i^
miles and a width of \ to f of a mile, placed with its longer axis parallel with the
valley (see plate i and plate xiv, figure 2). The chief irregularity in the outline
is due to the presence of a rock slide from Mt. Fairview and to the delta deposits
at the head. The level of the lake is given by the Topographic Survey as 5,670
feet above sea level. This level, however, fluctuates in the course of the
season and from year to year, being some 15 inches higher in the spring, as a
result of the rapid melting of the snow. Four determinations upon the inflow
at the head gave an average of 80 cubic feet per second. Two measurements
upon the outflow, through a rectangular orifice at the dam, gave an average
of 88 cubic feet per second, the lake receiving a small additional flow from
Mirror Lake and Lake Agnes. In midsummer, owing to the presence of the
glacial sediment, the lake has a superb green color, but during the winter this
has a chance to settle to the bottom and in the spring the color is more of a
blue, the natural color of pure water.

From the studies of Mr. W. D. Wilcox, reported in the paper previously cited
(page 7), the basin of the lake is seen to be a U-shaped trough, with a maxi-
mum depth of 230 feet just beyond the centre. This shows that the valley
was excavated by ice and that, in all probability, the bottom of the lake is a
glacially excavated rock-basin, similar to that of Lake Agnes, just west but
at a higher level. Bedrock is found upon all sides of the lake, except about the
foot, where there is the ground-morainic dam previously described (page 8)
filling the valley to an unknown depth. At the head of the lake the valley walls
are much contracted, being but 570 feet apart. They consist of a very firm
quartzite, in the main, with a little slaty schist, dipping up the valley at an angle
of 10° to 15°. This feature of the valley must have greatly contracted the ancient
glacier passing through the gateway, caused it to thicken correspondingly
and to vigorously gouge out its bed until it had a chance to again expand laterally.
In this way we may account for the presence of the rock-basin, but should
expect the deepest part of it to be somewhat nearer the head of the lake than is
shown in Wilcox's contour map. It is not at all improbable, however, that the
deepest part of the rock -basin is really so located and that there has been a

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XX.

Scale
0 50 100 200 300 -4-00 ft.

i i

Map of delta, head of Lake Louise, by W. H. Sherzer. Plane-table survey, July, 1904. Field assistants De

Forrest Ross and Frederick Larmour.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 6 1

filling of the lake bed here with glacial sediment and debris from the Mt. Fair-
view slide, to the extent, at least, of 50 to 60 feet.

c. Lake Louise delta. After the ancient Victoria permanently retreated from
the head of the lake, there began the formation, at the head of Lake Louise, of
a gravel, sand, and silt delta, which is a minimum measure of the amount of
glacial erosion that has since been taking place in the upper part of the valley.
This delta extends into the lake 400 feet, with an average breadth of a.bout 300
feet, as shown upon the map, plate xx, that was prepared with a plane-table in
July, 1904. A shelf of fine sediment borders that portion above water and
then drops off rapidly, forming a layer of unknown thickness over the bottom
of the lake. Much of the delta is elevated but 10 to 12 inches above the
lake level and is under water during high water stages of the lake. Portions
of it are flooded during midsummer after periods of excessive melting. In this
way the delta is gradually growing in height. Low, broad levees line the
main stream. This comparatively small portion, which projects into the lake,
is simply part of a very much more extensive deposit of glacial silt, sand, and
gravel, which reaches up the valley for a distance of | to ^ of a mile and has a
breadth of 500 to 600 feet. Over this very gradual slope the glacial drainage
courses in three rapid, turbid streams, which unite into a single channel, 40 to
50 feet broad and one to two feet deep. Grasses and moss cover the lower,
flat portion of the deposit, close set willow and spruce the central part, and
isolated willows the upper gravelly region. The rock-slide from Mt. Fairview
has encroached upon the lake as well as the morainic dam at the foot of the
lake, upon which stands the chalet. Between this dam and the range of moun-
tains in the background lies the Bow River Valley, which supported a great trunk
glacier, leading eastward from the mountains (plate xiv, figure 2).

d. Lake Louise Valley. When the mountains were in process of making it
is very probable that this valley began as a structural feature either as a trough
between mountain folds, or from the opening of a joint in the rock strata.
Before the coming of the glaciers it is probable that ages of stream action, aided
by atmospheric weathering, deepened and broadened the original feature into a
V-shaped valley. With the advent of the perennial snowfields a new geological
agent entered, deepening the bed and broadening out the base so that the cross-
section of the valley, particularly the lower half, assumed the characteristic
U-shape of glacially excavated, or glacially modified valleys. An inspection of
the general views of the valley, such as plate i, shows a lower portion with steeply
inclined, nearly vertical walls, while the upper portion has more flaring walls,
the arms of a truncated v. This upper portion was glaciated to a height of about
9,000 feet above sea level, or some 3 ,000 feet above the valley floor, the walls being
smoothed and fluted and the spurs evenly truncated; see the shoulder of Mt.
Fairview at the right in plate xix, figure 2. It is not improbable that these
more gentle, higher slopes represent portions of the original pre-glacial valley,
produced mainly by the action of weather and running water, and not very
materially modified by the ice. These slopes produced until they intersect may

62 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

be assumed to represent approximately the original valley before the ice invasion.
The lower portion, with the very steep walls and broad base, probably represents
that portion of the valley which was profoundly affected by the great ice stream
that so nearly filled the valley. The destruction of the strata necessary to
secure such a result was very probably accomplished by the disrupting power of
the ice, known as "plucking."

e. Ancient till sheet. The rock debris was, in great part, delivered to the
Bow Glacier and carried beyond the mountains. The finer portions, consisting
of gravel, sand, and clay, with some cobbles and boulders, secured lodgment in
certain favorable places, particularly near the mouth of the valley, and com-
pacted by the great weight of ice formed the ground moraine, or sheet of till.
From 75 feet to 100 feet of this deposit have been exposed between the foot of
the lake and the Bow River, but the maximum thickness was very probably
much greater. It forms an irregular sheet, of unknown extent and thickness,
reaching up to and under the modern glacier, mantling the actual rock bottom
of the valley. It is entirely without stratification and the rock fragments are
largely bruised and scratched. The color, as seen in the exposed sections, is a
brownish-yellow, as the iron of the clay is being slowly oxydized. The more
deeply buried beds would, undoubtedly, show more of the bluish-gray, which
characterizes this material when it is fresh.

CHAPTER IV.

WENKCHEMNA GLACIER.

i. GENERAL CHARACTERISTICS.

NESTLING close in behind the northern base of that grand array of peaks for
which the Canadian Geographic Board has recently adopted the name Wenk-
chemna Group lies the Wenkchemna Glacier. The name is of Sioux origin, wik-
chemna signifying ten, and was given the glacier by Mr. S. E. S. Allen, in allusion to
its relation to the series of ten peaks upon which he bestowed the Indian numerals.
These peaks, the highest of which has been rechristened Mt. Deltaform, with
their high connecting ridges constitute here the great Continental Divide (plate
xxi). Between them and Mt. Temple lies a broad valley originally called "Deso-
lation Valley" by Wilcox, but now known as the Valley of the Ten Peaks.
The glacier occupies the southern half of the upper third of this valley, and
faces north, while the valley itself slopes eastward and then northeastward.
In a direct line it lies only about six miles south-southeast from the Victoria,
and may be reached by crossing the Mitre Pass, encircling the Horseshoe Glacier
at the head of Paradise Valley and entering the Valley of the Ten Peaks by the
Wastach Pass. The ordinary way of reaching the glacier, however, is from
the chalet at the foot of Lake Louise, over a good trail to Moraine Lake,
where a summer camp is maintained by the Canadian Pacific Railway. From

o

I_
3

a.
5

*G °

a o

c iu

'

- c

2 «3

o ?r

u

OF THE

UNIVERSITY

OF

X

ft,

H

I

i

- ' X

•:

w

N

a
o
Q
H
J

O

g
tf)

O
H

CO

2

H
Z
O
U

o

V3

H

i

a

t

Ul

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 63

this camp, at the foot of the lake, the trail is rough for horses, but practicable to the
front of the glacier, about two miles distant.

Quite in contrast with the Victoria, and with mountain glaciers in general,
the Wenkchemna presents' noteworthy peculiarities of form, in that it is very
broad for its length and has a remarkable amount of frontage. Its breadth
is about three miles, while its length is from one-half to one mile, the frontage
amounting to something over three miles. The area of the glacier is estimated
at about two square miles. It "lies mainly between 7,500 feet and 6,400 feet
above sea level, the easternmost nose attaining the latter elevation, or about
400 feet higher than the Victoria.

2. PIEDMONT TYPE.

The peculiarities above noted in form are dependent upon the very unusual
method of formation. Instead of there being a trunk stream, to which the
minor ice streams are tributary, the entire glacier results from the amalgamation
of twelve, more or less, independent ice streams, each with its own feeding ground,
which lie side by side. There is no propriety in speaking of these streams as
tributaries; but since they are all nourished from the same general source, the
snow which falls upon the eastern slopes of the Ten Peaks, since they coexist,
are tolerant of one another's presence, and maintain their own identity and
independent velocity from ne"ve" to nose, they may be spoken of as "commensal
streams," to borrow an adjective from the biologists. The form of the front,
the position of the medial moraines, and the neVe" areas enable us to differentiate
these streams as shown upon the map,. and less well in plate xxi. Owing to the
general slope of the valley floor, the commensal streams are deflected eastward,
their natural course being northward. However, it is quite apparent that they
interfere with one another's movements. The easternmost stream is relatively
very small, terminating back some 3,000 feet from the nose of its neighbor,
where it is forming a terminal moraine. Its neighbor to the west is narrow,
but in conjunction with streams three and four, counting from the east, it reaches
the general front and together they form a broad rounded nose. Number five
spreads out fan-like at its lower end, and in consequence six and seven, in their
lower third, are deflected rather sharply to the north. Streams seven and eight,
from Mt. Deltaform, are exhibiting the greatest amount of relative activity.
In the western part of the glacier the ice streams are turned eastward by the
tremendous accumulation of morainic blocks which they are unable to push
ahead or override. In consequence, number nine, which has only a limited
collecting area, is considerably compressed, being forced laterally against its
sturdier neighbor to the east. It is not likely that any one of these streams
could exist by itself as an independent glacier, since it would flatten out, be
more thinly clad with de"bris, waste more rapidly from surface and lateral melting,
and disappear soon after leaving the shadow of the mountains.

This form of glacier is known as the ' ' piedmont type, ' ' and, so far as the writer
is aware, only one other example, the Malaspina, of Alaska, has thus far been

64 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

described. This has a breadth of 70 miles and an average length of 20 to 25
miles, with approximately i ,500 square miles of area. It is made up of four great
commensal glaciers, with an innumerable number of smaller ones. Farther
west there lies the Bering Glacier, known to be of the same type, but not yet
visited and described. It is quite likely that this variety of glacier is more
common than has been recognized, since in addition to the Wenkchemna there
is the Horseshoe Glacier at the head of the adjoining Paradise Valley, with some
1 6 commensal streams, and the Asulkan in the Selkirks (plate xxxix, figure 2)
which represents a piedmont glacier in process of disintegration into its compo-
nent streams. If the ordinary Alpine glacier, with its tributaries, is compared
to a river, the body of a piedmont glacier should be thought of as an ice lake,
of greater or less magnitude.

3. NOURISHMENT.

Each of the component streams of the Wenkchemna may be traced back to a
more or less well defined and fairly distinct patch of neve. This may be no
more than a cone of avalanched snow, or it may be a strip of permanent snow
field filling a couloir in the mountain side, or between two adjacent peaks.
The highest point of the Divide here is Mt. Deltaform, with an elevation of 1 1,225
feet, somewhat lower than Victoria. To the west, and closely connected with
Deltaform, is Neptuak (Allen's No. 9) with an elevation of 8,767 feet. East-
ward from Deltaform there occur in order No. 7 (10,648 feet), No. 6 (10,520
feet), No. 5 (10,018 feet) and No. 4 (10,028 feet). The northern face of this array
of peaks is very abrupt, furnishing only a meager collecting area for the snow.
The snowfall is probably not materially different from that at the head of the
Lake Louise Valley, where it is estimated as about 25 feet annually. That which
clings to the steep slopes during the winter is largely avalanched upon the glacier
below in the spring and early summer. The most of that which remains is melted
and but little survives the warm season. The snow accumulates along the
northern base of the range, where it is protected from the noonday sun, allowing
it to become converted into neVe" and compacted into ice. Owing to the increased
altitude of the glacier's surface toward the west there is a greater deposit along
the base of Deltaform and Neptuak (plate xxi). This rather meager supply of
snow could support a glacier of such dimensions only because of certain favorable
conditions. Being in the lee of some 3,000 feet of nearly vertical cliff, its neVe"
field is sheltered from the noonday sun, while that portion which is exposed is
almost completely veneered with a protective covering of rock debris. The
form of glacier, with the ice streams lying side by side, reduces the lateral melting
to a minimum, while the slope of the valley floor, upon which the glacier rests,
is sufficiently gentle to allow the ice to remain in a very sluggish condition.

4. DRAINAGE.

Because of the conditions just outlined the amount of ablation is reduced
to a minimum and the surface drainage streams are correspondingly small and

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XXIII.

FIG. i. — Drainage brook from Wenkchemna Glacier, August, 11)04. Free from sediment

and having a summer temperature of 35° to 36° F.
No. 2. No. 3.

No. 5.

FIG. 2. — General view of eastern end of Wenkchemna Glacier, August, 1904. Looking southward. The very complete
covering of rock debris and irregular surface are well shown.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 65

inconspicuous. Some near the center have cut their way a few feet into the ice.
None of them reach the margin of the glacier, but find their way to the base
through crevasses, or moulins. Opposite peak No. 7 a small stream of pure
water from the valley enters the side of the glacier. The drainage from the Wenk-
chemna Lake, which collects the waters from the base of Mt. Hungabee, reaches the
glacier through the great accumulation of morainic blocks (plate xxiv, figure 3).
No surface lakelets were observed upon the glacier, although numerous depres-
sions occur, suggestive of the former sites of lakelets. The water is probably
lacking now because of the small amount of surface melting. There was practi-
cally no marginal drainage observed. At the east end, between the side of the
mountain and the glacier, there occurs a small lakelet and opposite Mt. Delta-
form, at the front, there is a very shallow lakelet, or marsh, of insignificant
proportions.

The various subglacial drainage streams are collected beneath the glacier into
a single stream which gushes from the extreme eastern compound nose and cascades
over the coarse blocks of the frontal moraine in several channels. These form a
single broad drainage brook (plate xxm, figure i), about 100 feet across, shallow
and rapid, which enters Moraine Lake one-half mile below, dropping 210 feet
in the distance. The volume could not be measured, but was estimated at about
90 cubic feet per second, being somewhat in excess of the volume of the Victoria
drainage brook. The volume did not fluctuate during the day nearly so much as
is usual for glacial brooks, suggesting that the supply is not so dependent upon
immediate melting of the ice. This is further indicated by the temperature of
the water, which remained during the middle week in August very steadily
at 35.6° P., and rarely varying more than 0.2° to 0.3°, no matter at what time
of the day taken. September 8, 1905, it was still 35.6° at 12:30 P.M. A sur-
prising feature of the brook is its remarkable freedom from glacial sediment,
the water issuing from the glacier perfectly pure. Flowing over coarse gravel
and boulders it acquires no sediment upon the way to the lake and has formed
not even the suggestion of a delta at the head of Moraine Lake (plate xxv).
This indicates that the subglacial erosion is practically nothing and has remained
so for centuries, testifying to the sluggish condition of the glacier. In flowing the
half mile to the lake the temperature of the water in August was raised from
35.6° to 36° or 37°. The lake is about a mile long, has an elevation of 6,190
feet, is apparently shallow, and filled with the purest water of an intense blue
color. Passing the length of the lake the temperature in August is raised to
about 44° P. The freedom of the water from sediment permits it to exhibit its
natural color, a simple glance at which, as far away as the color could be dis-
tinguished, enables one to safely predict the absence of sediment from the brook
emptying into the lake. Should the glacier become active and start to erode its
bed the color of the lake would change to some shade of green.

5. MORAINES.
Except for the comparatively narrow strip of ne"ve" along the base of

66 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

the cliff, well seen in plate xxr, the entire upper surface of the glacier is
veneered with angular rock debris, effectually preventing surface melting, as
already shown. This material is derived from the Wenkchemna group of peaks,
through the agency of avalanches and the ordinary processes of weathering.
With the entire breadth of the glacier spread out along the base of the cliff
all portions receive their quota, leaving no portion of the ice exposed to the
sun. The de'bris, at first, is covered with the snow, but it is concentrated by
melting until the amount is sufficient to prevent further loss of ice at the surface
when the action ceases. No ground-morainic material was observed" upon the
surface, in contrast with the Victoria, and this is -accounted for by the absence
of hanging glaciers; What might be mistaken for such upon the northern face
of Mt. Deltaform, and upon either side of peaks 4 and 5 (plate xxi), are simply
the continuous neVe" fields of the commensal streams. This method of acquiring
its load leads to a somewhat irregular distribution of the rock de'bris, resulting in
hummocks and depressions, especially towards the northeastern corner (plates xxi
and xxv). These irregularities of surface are also shown in plate xxin, figure 2.
It was from this portion of the glacier that the view for plate xix, figure i, was
taken. This irregularity of surface renders travelling across the glacier laborious
and somewhat dangerous, except near the neVe line. The almost complete
concealment of the ice by de'bris renders this glacier a poor one for the study
of ice structure and the usual surface features. The third ice stream, coming
from between peaks 5 and 6, in the vicinity of the neVe" shows stratification
and dirt zones to advantage, the strata ranging from five to ten feet in thickness.
Low glacial tables occur here and the phenomenon described upon page 58 of
this report, as possibly due to reflection of heat from the surface boulders.

The line of junction between neighboring ice streams is roughly indicated
upon the surface by ridges of rock de'bris, somewhat low and poorly defined
near the neVe", but gaining in height and distinctness in their course across
the glacier. These ridges are the lateral moraines of the individual ice streams
and they are especially well defined over the eastern third of the glacier. Upon
either side of the third stream these ridges are double for a considerable dis-
tance. Toward the western end of the glacier these moraines are neither so
well defined nor so continuous and, owing to the deflection of the streams to
the eastward, by the ancient moraine, swing around into a position almost
parallel with the frontal. By the blending of these lateral moraines upon
adjacent streams the single ridges resulting become the medial moraines of
the piedmont glacier, and the outermost laterals of the two marginal streams
become the laterals of the unified glacier. At the eastern end of the glacier the
first stream is so short that the right lateral of the second ice stream constitutes
the right lateral of the glacier as a whole. If the first stream ever extended to the
front then this lateral moraine was originally a medial. A deep depression, snow-
filled in 1904, separates it from the double de'bris cone and from the mountain
spur shown upon the map. At the western side of the glacier, owing to the
deflection eastward of the ice streams, there is no distinction to be made between

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 67

the lateral and the frontal moraine. They end in a peculiar series of short
closely pressed ridges, slightly concave outward.

Since a glacier of this kind cannot be said to have an end, the term frontal
may be more appropriately applied to the rock debris that is being dumped
along the united extremities of the individual ice streams. Were the glacier
to begin a uniform retreat from its present position, there would be left a ridge
of angular rock debris, over three miles in length, marking the shape and present
position of the front. Inside of this would be left upon the valley floor the debris
which now mantles the surface of the ice, or is contained within. Because of
the very slow advance, to be noted below, the frontal morainic material over
the eastern half is being very slowly urged forward, giving a steep and unstable
frontal slope, but not so steep that it can not be climbed at almost any point.
At only one point, nearly opposite peak No. 7, is there any ice showing and
here the de"bris cover is partially lacking. Should the ice front actually halt a
frontal moraine would form very slowly, in spite of the amount of ddbris carried,
because of the sluggish condition of the ice. Toward the western side the front
becomes less steep and high and finally merges into the neVe and snow bank
which mantles the col between Neptuak and Hungabee.

6. CREVASSES.

The glacier is remarkably free from crevasses in the lower part and about the
sides. In the case of the commensal streams, measurements would probably
show that the sides were moving forward at about the same rate as the centers,
so that there is lacking that differential movement that gives rise to marginal
crevasses. The mutual pressure from the sides is sufficient to prevent the
opening of radial crevasses along the front. The absence of prominent transverse
crevasses indicates that the bed is of even slope and the motion slow enough
to allow the ice to yield, without rupture, to most of the inequalities that do
exist. The absence of crevasses in this case is quite as instructive as their pres-
ence would be. Upon the steeper portions of the neVe" slopes there occur nu-
merous transverse breaks of the nature of bergschrunds, caused by the upper mass
clinging, for the time being, to the rocky wall while the lower portion draws
away from it. If kept under inspection these schrunds would be found to close
up, as they work their way down the slope and to open again at a higher level.

7. MOVEMENT ABOUT THE FRONT.

In a little booklet prepared for the Canadian Pacific Railway by Messrs.
George and William Vaux, and entitled Glaciers, attention was first called
to the evidence that this glacier is advancing into the adjoining forest. No
data were at hand for determining the amount of this forward movement, or
whether it is still in progress. Dead trunks of forest trees, from which the
bark and branches have fallen, are seen projecting from near the frontal slope
(plate xxiv, figure i). Some of these trees were probably killed by a forest fire

68 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

which swept through the valley 70 to 80 years ago. Other trees in similar posi-
tion, but also dead, still retain their bark and boughs, but show no signs of fire.
It is likely that these trees were killed, and more or less. displaced, by the ad-
vance of the ice front (plate xxiv, figure 2), since which time the ice has advanced
less than a dozen feet. This is still further evidence of the almost stagnant
condition of the glacier. Only at one point, near the center, were there any
trees which have been recently cut by rolling blocks from the frontal slope.
This is taking place about the nose of the stream coming from Mt. Deltaform.

In order to gather some definite data concerning the frontal movements,
a series of eight sets of reference blocks was established along the eastern half
of the front, beginning at a point just east of the drainage brook. Between
certain marked points upon boulders that had rolled forward and others firmly
embedded in the frontal slope, accurate measurements were made with a steel
tape.. From August 9 to September 12, 1904, an interval of 34 days, it was
found that there was no perceptible movement at the station east of the drainage
brook. Passing westward along the front, and up the valley, the data indicated
that there had been a wastage of the ice, causing the blocks to settle back 1.2
inches and 0.7 inch. The next two stations showed an advance of 1.9 and 1.3
inches, while the next two gave a retreat of i.o and 4.6 inches respectively.
At the upper station, where the trees had been freshly cut, the advance for
the 34 days amounted to n.8 inches. One year later, September 8, 1905,
measurements were again made between the series of blocks and at all of the
stations (the upper block at station D could not be located because of disturb-
ance) there was a small advance indicated, varying from 1.7 inches to 20.4
inches. The least movement was about the extremities of the easternmost
streams and the greatest was towards the center.

A summary of the measurements is given below.

Stations. Movement for 34 days, Aug. 9 to Movement for 361 days, Sept. 12,

Sept. 12, 1904. 1904, to Sept. 8, 1905.

A o.o inches i . 7 inches.

A -1.2

B -0.7

C 1.9

E -i .o

F -4.6

G ii. 8

2-4
4.8

12 . o

Missing.

3-2
20.4

IS- 1

These figures indicate that the component glaciers are as independent m their
movements as in their structure, and that some may be stationary, or in retreat,
while others lying alongside are advancing. The question of the frontal be-
havior of a. piedmont glacier is thus seen to be complicated in proportion to
the complexity of its structure. Measurements made at single stations can give
only very incomplete data concerning the glacier as a whole. There should be
at least one such measurement for each commensal stream.

8. FORMER ACTIVITY.
a. Bear-den moraines. Along the western front of the Wenkchemna, for a

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 69

distance of over a mile, there occurs a tremendous accumulation of huge mo-
rainic blocks of a red and brown sandstone. The blocks are much disintegrated
by the weather, and falling apart, but each roughly indicates its former size and
shape (plate xxiv, figure 3) . Near the upper end of the valley the moraine is a half
mile across, extending from the glacier to the foot of Eiffel Peak and almost
completely surrounding Wenkchemna Lake, as shown upon the map. Toward
the east the moraine becomes narrower and there may be distinguished an
older portion, partially soil-covered and forested. These latter blocks are more
completely coated with lichens and have plainly the appearance of greater age.
There is thus evidence that the moraine was formed at two different periods,
the older portion surrounding the lake and extending eastward, and that a con-
siderable interval, as expressed in years, separated the two periods. Opposite
peak No. 7 there occurs at the front an accumulation of coarse blocks, evi-
dently part of the younger of the two ancient moraines, which the present glacier
has been able to partially override. Farther west the glacier has been unequal
to the task of either pushing the blocks ahead or of overriding them, and the
streams have been deflected eastward, as previously noted. When the forma-
tion of the older of the two moraines began the glacier reached across the valley
and deposition started. The glacier had so little depth, owing to the meager
supply of ne"ve", that it was unable to heap the blocks into a great ridge. Ap-
parently the rather thin edge of ice was pressed against the moraine and there
melted by pressure, and the blocks deposited along the southern margin until a
belt a quarter mile in breadth was formed. A considerable period intervened
during which time the eastern portion of the glacier had shrunken to much
smaller proportions than it had formerly held and than it has at present. In
a manner still to be accounted for, the glacier a second time became loaded
with very coarse blocks and started to advance, possibly because of the protection
afforded the surface by this load. Encountering the former moraine, however,
it was unable to go over, or around, and it used its energy in a vain attempt
to push the obstacles ahead. The pressure exerted caused the melting of the ice
and the blocks were deposited in another broad, continuous belt, parallel with
the first. As though it had learned wisdom by the experience, the glacier now
rather calmly turns eastward and passes around the obstruction which it has
built in its own path. These two moraines, lying side by side, are of the same
type as the two described in the Lake Louise Valley, and, as far as may be judged
roughly from their general appearance, of approximately the same age. The
cliff from which the material was obtained has a west-northwest trend and the
blocks were dropped from it to the eastward.

b. Moraine Lake. This lake has had a different history from that of Lake
Louise in that it is apparently not a rock-basin and so attains no great depth.
It is, however, like Lake Louise in that it has a morainic dam across its foot,
although in the case of Moraine Lake the dam is of a different type. This
consists of a sharply defined heap of rock de"bris about 400 feet long, placed
at right angles to the main axis of the valley. The ridge increases in height

70 GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

rather gradually toward the west and attains a height of about 70 feet, ending
abruptly, as steeply as the debris will stand and with no trace of any continuance
across the valley. This is so unusual a feature for a terminal moraine that many
are disposed to consider the mass as a rock slide from the adjoining mountain
face. There are several varieties of rock represented in the heap and time did
not permit an examination of the adjoining face to see whether they might have
had such a source. The strata in the region, however, are so nearly horizontal
that, whether the feature is a moraine or a slide, the same rocks probably occur in
the adjoining mountain as farther up the valley. Standing upon the highest
crest, the ridge is seen to be double, the outer one somewhat convex down stream,
while from the western end there passes a short spur down the valley. The
writer is disposed to accept the view of Wflcox, who gave the name to the lake,
that we have here a moraine. It is, however, not of the bear-den type found
farther up the valley, but very much older than the most ancient of the two.
Its general lack of vegetation may be due to the scarcity of suitable soil, although
it does support a sparse growth of timber. The unusual features of the mass,
considered as a moraine, will be understood when the unusual nature of the glacier
that formed it is considered. This represents the position of the front of the
easternmost ice stream, of the ancient piedmont Wenkchemna during a pro-
longed period of the halt. This moraine originally abutted against a wall of ice
at the west end, the side of the adjacent ice stream, which probably extended
far down the valley and may have been engaged in making a correlative moraine.
A relatively small amount of the debris was dragged for a short distance down
stream by this neighbor, forming the spur above noted. When this ice wall
melted away finally the debris rolled down and assumed the "angle of repose."
As has been pointed out the present easternmost stream is short, compared with
its neighbor, and were it to make a sufficiently prolonged halt there might be
produced, upon a smaller scale, this identical feature.

c. Valley of Ten Peaks. The time that could be devoted to this glacier did
not permit of an examination of the valley from the lake to the Bow River, or of
the interesting Consolation Valley, which still supports a glacier at its head.
Observations of only a general nature from the elevated trail around Mt. Temple
could be made. There is evidence that the entire valley was occupied by a great
ice stream, a tributary of the trunk glacier that filled the Bow Valley, the glacier
then being of the Alpine type. The lower half of the valley was altered by the
ice into the characteristic U-shape, while the upper half retained its flaring side
walls from pre-pleistocene time. In making the bend from its east-southeast
course to the northeast, the glacier pressed hard against the western face of
Mt. Babel, while upon the opposite, or concave, side there was deposited a high
ridge of ground-morainic material which swings around in a very regular curve
from the Eiffel to Mt. Temple. From the northeastern shoulder of Mt. Temple
there extends into the Bow Valley, curving gently down stream, a spur of ground-
morainic material, identical with that described for the Lake Louise Valley upon
page 8. . This was deposited beneath the ice and along the line of junction of the

ft!

E

!Z
C

0.- THE

(( UNIVERSITY D

OF

-• r

X
X

H

H
4

J

a.

Ex

•73

«

K

2

H
X

1

o

Q

s

I

z

U!

O
H

H

-

5

H

S5
O
o

O
ir.

H

til

y

j
<S

•a
•d

o
o.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 7 1

ancient Wenkchemna and Bow glaciers. Looking down the valley towards the
Bow, Babel Mountain is on the right and Mt. Temple upon the left, rising above
the high morainic ridge mentioned. From the shoulder of Mt. Temple there ex-
tends to the center of the view the morainic ridge reaching out into the Bow
Valley, gradually losing in height and breadth. As pointed out by Wilcox, this
deposit probably mantles a rock spur which escaped destruction by the ice.
During the height of glaciation a tributary glacier moved in northwestward from
Consolation Valley and joined the Wenkchemna at a level 400 to 500 feet above
the present valley floor, forming a "hanging- valley." From the "Tower of
Babel " there curves across the mouth of the valley what appears to be a morainic
ridge, of the same nature and origin as that just described. The height of this
hanging- valley above that of the Ten Peaks is believed by some to measure the
differential erosion between the ancient Wenkchemna and that of the tributary
which occupied this valley.

CHAPTER V.

YOHO GLACIER.

i. GENERAL CHARACTERISTICS.

THIS glacier, the largest and most northerly situated of the series studied
constitutes a tongue of ice from the great Waputik snow-ice field which mantles
the Continental Divide to the north of the railway. Its nose lies in latitude 51°
34', at the head of the picturesque Yoho Valley, and is most conveniently reached
from Field, via Emerald Lake. The day's ride, over a fairly good trail, up this
ice-cut valley, with its hanging glaciers and plunging cataracts, is an experience
never to be forgotten. The return trip to Field should be made over the Burgess
Pass. During the summer the Canadian Pacific Railway maintains a camp at
Laughing Falls, some four miles from the glacier. The glacier was first made
known through the descriptions and photographs of Jean Habel,1 secured in
1897, and each summer since it has been visited by gradually increasing numbers
of tourists and students. The original name was derived from the Wapta River,
another name for the Kicking Horse, the name "wapta" itself meaning river
in the Stoney Indian language. The name Yoho since approved by the Canadian
Geographic Board is the Indian exclamation of surprise and wonderment.

As one emerges from the forest and comes suddenly face to face with the
glacier, plunging at him from above, he is greatly impressed with its size and appar-
ent powrer. Its freedom from surface debris better enables it to meet the popular
idea of what a glacier should look like, — the Victoria and Wenkchemna, having
been somewhat disappointing (plate xxvii, figure 3) in this respect. The Yoho
has the general form of a gauntlet mitten, extending in a south-southeast direc-
tion, with the thumb upon the eastern side of the valley and partly surrounding
a great rock embossment (see plates xxvi and xxvui, figure 2). Independently of

1 " The North Fork of the Wapta," Afpalachia, Vol. vm, No. 4, 1898, pp; 327-336.

72 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

its ne"ve" it is three miles in length and for its upper two-thirds nearly one mile
in breadth. In rounding the rock embossment noted it narrows to a half
mile and then tapers regularly to a blunt nose. The mean elevation of the
ne"ve", according to Wheeler, is 8,400 feet, and the nose descends to an altitude
of 5,670 feet. By noting the temperature of boiling water, Habel determined
this elevation at 5,680 feet. The nose of the Yoho is thus 330 feet lower than
that of the Victoria, and 730 feet lower than the Wenkchemna, in spite of the
lack of de"bris covering and the southerly exposure. This is due, without doubt.
to the greater precipitation and the greater size of the collecting area, by which
a much larger body of ice is amassed. The mean average slope from the neV£
to the nose is about 900 feet to the mile; the main part of the descent, however,
is in the lower half. The general inclination of the ice about the front is 20° to
25°. Upon the western side the glacier presses more or less firmly against the
valley wall, except for a short distance, where the ice is steep and not to be ascended
without cutting steps. By crossing the drainage stream, which is not a simple
proposition unless one is mounted, one may easily ascend the glacier without
the cutting of steps, the ice slope being very gentle. Skirting the crevasses and
crossing back to the west side of the glacier, the neVe" may be easily and safely
reached. Since its discovery by Habel, the glacier has maintained a great archway
of ice at its lower extremity, which spans 250 feet of space, and is estimated to be
70 feet high, from which escapes the drainage. Owing probably to its southern ex-
posure there is formed a cavern beneath the arch, as seen in plate xxvii, figure 4,
extending back into the ice 100 to 200 feet, but not forming a subglacial tunnel.
Toward the close of the summer season, the arch has become so weakened by
melting and the formation of a transverse crevasse (plate xxvn, figure 3) that
the entire structure collapses and lies a heap of azure ruins. The blocks of ice
are melted down to a size that the stream can push, roll, or float, some head being
obtained for the stream by the damming action of the ice debris. Finally it is
all removed and the making of a new archway is started.

The actual nose of the glacier lies to the east of the archway and rests upon
limestone bedrock, with only a sprinkling of ground-morainic material. Upon
either side, and for some distance beyond the nose of the glacier, there is bed-
rock exposed, which has been smoothed in places by previous ice action and
in other places roughened by plucking. Upon the western margin of the drain-
age brook there are shale strata upon edge which have been thus roughened.
Excessively thin laminae alternate of an intense red and yellow color. There are
no reliable data for estimating the thickness of the ice, but it seems to be consid-
erable. In the case of the Victoria and Wenkchemna glaciers the most conspicu-
ous geological work being done is transportation, in the case of the Yoho we
have very plainly a great engine of erosion.

2. NOURISHMENT.

The collecting area of the Yoho is triangular in outline and includes the region
between Mt. Collie (10,315 feet), Mt. Baker (10,441 feet), and Mt. Gordon (10. 336

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 73

feet), the sides of which triangle are approximately 3x4x5 miles (plate xxvin,
figure 2). The area is located upon the western slope of the Great Divide, the
crest of which extends in a curve from Mt. Baker to Mt. Gordon. The collecting
area is estimated at about 62 square miles. Upon the western slope of the Divide
it presumably receives more precipitation than falls in the Lake Louise Valley
and that of the Ten Peaks. From the meager data available at Field we calcu-
lated that the precipitation may amount there to 42 inches per annum, page 1 1 .
To the north it would be somewhat less and may be assumed to be 40 inches.
Over the neVe" area the great bulk of this would fall as snow, but that which was
precipitated as rain would be absorbed at once and rendered available for the
glacier, representing about 33 1 feet of snowfall each year. This amount over the
collecting area, the region in which the snow is manufactured into glacial ice, would
represent some 224 million cubic yards of snow, or about 24,396,000 cubic yards
of ice, available each year for the Yoho. If our assumed data are approximately
correct, this must represent the amount of ice to be disposed of annually by
melting and evaporation. Converted into water, this volume of ice would pro-
duce 22,372,000 cubic yards of water, or 604,032,000 cubic feet of the same.
Distributed over the months May to September inclusive, during which time
the melting is most active, this would give an average flow of about 46 cubic
feet per second. During midsummer the flow is probably four or five times this
amount, due largely to the fact that the actual area drained is much larger than
the single n6v6 field, and that the melting is now at a maximum. The ne"ve"
coming in from the eastern, or Mt. Gordon side, as well as that from the western
or Mt. Collie side, has already been compacted into glacial ice before reaching
the main flow from Mt. Baker. This ice is incorporated into the Yoho ne"ve"
with whatever debris it may be carrying. The absence of overhanging cliffs
about the ne"ve" area, quite in contrast with the Victoria and Wenkchemna
glaciers, prevents the neVe" snow from becoming charged with rock de"bris.
The glaciers from the slopes of Gordon and Collie, as well as the main stream
from Mt. Baker, are carrying only subglacial material, of which we have
evidence later. To this is to be ascribed the freedom of the glacier from surface
debris. This condition of the ice, combined with its southern exposure to the
sun, is unfavorable to the maintenance of a glacier at low altitudes. This is
entirely obviated, however, by the greater bulk of ice available when this
glacier is compared with the two previously described.

3. DISTRIBUTARY.

Except for the ne"ve"-covered glaciers above noted from Mts. Gordon and
Collie, the Yoho has no tributaries; but instead, what has been termed a dis-
tributary, to assist it in getting rid of its ice supply. A considerable volume of
ice is deflected around to the eastern side of the rock embossment and is there pre-
maturely melted. (See plates xxvi, xxvin, figure 2, and plate xxvii, figure 3.)
This tongue of ice forms a very pretty little glacier, one-half mile long by one-
quarter mile broad and tapering down to a blunt nose (plate xxvii, figure i).

74 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

The surface is soiled with wind-blown dirt but it carries little debris. At an
earlier stage it brought down from above the left lateral morainic material for the
Yoho, and its ice extends still well under this ancient moraine. The axis of the
glacier is curved as it is forced around the rock embossment, which it hugs closely
and is still engaged in fluting and polishing (plate xxvii, figure i). The upper
crest of the rock embossment has an elevation of 6,960 feet, while that of the
nose is 6,320 feet, or 650 feet above that of the main glacial stream. The average
slope would be at the rate of about i ,300 feet to the mile. It is evidently in retreat
although no "data are available for determining the rate. In the upper portion
of its course, it appears to descend over a steep step in its bed and is much
crevassed. These crevasses completely heal, however, or are destroyed to their
bases by melting, leaving the lower half exceptionally smooth. Over this portion
of the small glacier there are developed three very pretty drainage systems, two
of them marginal and the third central. The central drainage, which is collected
into a trunk stream (plate xxvii, figure i) from a network of small tributaries,
has cut a longitudinal channel in the ice, and continues to a point just east of the
nose. Habel's map of the Yoho Glacier shows this tongue of ice continuous
around the rock embossment, forming of it a rock island, or so-called "nunatak."
Such a position it originally held, but not less than 200 years ago, so far as we
may judge from the size of trees growing in the valley. At a still earlier stage
of glaciation the entire embossment was overridden by the ice, much of it being
disrupted, wherever the ice could get a satisfactory grip upon the strata. The
resistant portions were planed down, rounded, and fluted.

4. MORAINES.

Because of the lack of overtowering cliffs, above noted, the general surface of
the Yoho is practically free from coarse rock debris, in striking contrast with the
Victoria and Wenkchemna. For the same reason also the lateral moraines
are poorly developed and almost absent in the lower half mile. Upon the western
margin, just before reaching a broad glaciated valley, originally carrying a
tributary, the right lateral moraine begins to make its appearance and develops
across the mouth of the valley into a well defined ridge of stony till, or ground-
morainic matter. This ridge is continuous up the slope to the line of junction
of the main Yoho with the glacier from the eastern slopes of Mt. Collie, wheie
the latter is seen to be delivering this material from its under side to the surface
of the Yoho. This ground moraine has been produced between the Collie
Glacier and its bed, frozen into the ice, and urged down the slope.

The left lateral moraine begins along the southern side of the rock emboss-
ment, down in the valley, as a double ridge from which the ice has been withdrawn.
It extends around the embossment, on the west side, for a distance of some
4,500 feet, developing into a prominent, high, sharp-crested ridge at the head
and curving across to the eastward. This consists, also, almost entirely of ground-
morainic matter, which must have been derived from the basal layers of the ice,
which became stranded at the head and about the west side of the rock emboss-

4

A
a

^
o

x

ERSVTV

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 75

merit. In the valley occupied by the distributary described, and from 600 to
700 feet down from its nose, there begins a second lateral which curves around
upon the debris-covered ice and continues for two miles along the shoulder of
Mt. Gordon. This portion of Gordon has an elevation of 9,510 feet, but its
cliffs do not overhang and contribute only a moderate amount of material to the
moraine. A considerable portion of it consists of ground moraine which may
be traced to the glacier from Mt. Gordon. This sustains the same relation to the
Yoho as does the Collie Glacier upon the opposite side of the valley. Up to
heights of 30 to 40 feet patches of morainic material can be seen upon the valley
wall, left there when the surface of the glacier stood at a higher level. The rocks
in the moraine are largely limestone, light and dark, yellow and mottled; some
pieces being oolitic.

An older lateral moraine, upon the eastern side of the valley slope, may be
traced for some 2,000 feet up the valley from the nose. The ridge is some 200 to
300 feet from the margin of the ice, is but three to four feet high and inconspicu-
ous, but it has heaped promiscuously over it a mass of broken tree trunks very
evidently brought down by an avalanche and heaped against the side of the ice
when it stood here. The distance from the margin of the ice to the ridge, at one
point, was found to be 260 feet, along the slope. The wood is somewhat decayed
and gives some appearance of age. A photograph was taken when the trunks
were covered with a light fall of snow, which had melted from the surrounding
rock, rendering them much more conspicuous than they would otherwise be
with their dark surroundings. Growing in the path of the avalanche trees
were found, the largest of which gave 25, 28, and 47 rings respectively. It is
likely that the avalanche occurred between 1850 and 1860, since which time
the glacier has been retreating down the slope at the average rate of 5 to 6 feet
per annum.

With so little rock debris carried upon and within the glacier it would require
a very prolonged halt of the front in order to build up a terminal moraine of any
considerable proportions. About 200 feet from the present nose, at the end of
the bedrock upon which it rests, there swings in from the side a weakly developed
double ridge, low and inconspicuous. It may be the correlative of the lateral
moraine above noted, carrying the avalanched timber, or it may mark a still
more recent halt in the general retreat. Within recent time the glacier has
deposited very little ground moraine, the conditions not being favorable for its
lodgment. The lowermost stratum shows upon the western side of the drainage
stream, beneath the archway, and is seen to be charged with debris. Much of
this is delivered to the stream and swept away by the swift current, the remainder
being spread thinly over the valley floor.

5. CREVASSES.

Opposite the head of the rock embossment described, the glacier and its dis-
tributary plunge over a steep step in their beds, of which the embossment itself
is probably a portion which the glacier was unable to reduce to the general level

76 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of its bed. In making this rapid descent the ice is crevassed both transversely
and longitudinally. The irregular blocks of ice formed melt into sharp points,
or steeples, to which the term "seracs" is applied. The transverse crevasses
have been noted which open just behind the archway at the nose, allowing the
arch to collapse toward the close of the melting season. The absence of pressure
in front allows the arch to drop forward, faster than the ice can yield to the ten-
sional strain, and the crevasse is the result. Along both margins, for nearly the
entire three miles, the normal lateral crevasses, described for the Victoria, occur.
They extend inwards and upwards for varying distances, are irregularly spaced
and become less numerous toward the neVe", where some of them are snow-filled
and snow-covered. In passing the ice cascade the ice is too much shattered to
permit the formation and preservation of the dirt bands described upon page
5 2 of this report. However, at the crest of one of the minor slopes the phenome-
non may be seen, as shown in plate xxix, figure i, where the depressions for
three bands are plainly marked out. These mark the sites of former crevasses,
and, if rightly interpreted, the distances between them show the approximate
annual motion at this portion of the glacier.

6. ICE STRUCTURE.

In both the main glacier and the distributary the stratification of the ice
is poorly preserved, possibly because of its destruction in passing the cascade.
General views, as well as detailed ones, give almost no trace of the strata. In plate
xxvii, figure 4, one stratum, relatively much charged with debris, forms the base
of the arch, but does not appear upon the opposite side. This basal stratum
where seen is 2 to 5 feet in thickness. Its upper surface may represent a shearing-
plane, the body of the stratum being held more rigidly by its content of debris
while the superincumbent ice is forced over it.1 In places where the strata are
still preserved, the dividing planes show poorly, and it is to be noted that this
may arise because of the paucity of foreign material concentrated at the upper
surfaces of the strata. In the case of this particular glacier the size of the ne"ve"
field precludes any but the finest dust from reaching the general surface, and
with so few peaks uncovered in the region the supply of dust must be meager.
No opportunity was afforded for observing the structure of the ne"ve" itself. As
pointed out by Reid, in the paper cited upon page 43, the basal layers of a
glacier may be able to pass a cascade without suffering destruction, while the
upper strata may be destroyed and in large part melted away. This may be the
cause of the poor development of strata in the case of the Yoho, and the absence
of the dirt zones, which should show especially well over the smooth lower half
of the distributary.

In spite of the almost complete obliteration of the strata in the upper part,
the blue bands are shown in great perfection where the ice presses against the
west valley wall. The edges run parallel with the margin and the bands dip

i " The Influence of Debris on the Flow of Glaciers," I. C. Russell, Journal of Geology, vol. in, p. 833,
1895.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 77

down into the ice at angles of 42°, 48°, 53°, 54°, 56°, and 61°. Upon the walls
of the crevasses they may be seen to curve around into a position parallel with the
valley floor. At the surface the position and approximate thickness of the bands
are indicated by the dirt stripes. Differential movements of the ice, after the
formation of the bands, have given rise to curved, twisted, and contorted patterns
in numerous places towards the center (plate xni, figure 2).

The fine development of glacial granules and capillaries in the Yoho Glacier
has been already noted upon pages 39 and 41. They here attain the largest size
of any seen in the series of glaciers studied and appear to have about the same
amount of orientation near the nose.

7. DRAINAGE.

Owing to the crevassed condition of the main glacier, there is little oppor-
tunity for the development of surface drainage streams, the water soon making its
way to the bottom of the bed. In the upper portion where the crevasses are not so
numerous toward the center, there seems to be too little melting to call for much
surface drainage. The drainage upon the distributary has already been referred
to. There enters its side a strong flow of water from the hanging- valley to the
east (plate xxix, figure 2), derived from the glaciers lying between Mt. Balfour and
Mt. Gordon. Opposite the head of the rock embossment there is a short strip
of marginal drainage as shown in the map, but the stream is small and the flow
weak. Upon the opposite side of the valley two streams with a brisk flow
enter the side of the Yoho from the broad, glaciated valley noted, while a third
flows down the northern slope from the glacier upon Mt. Collie. Marginal or
surface lakelets were nowhere observed. Augmented with the flow from the
hanging valley, there rushes from beneath the nose of the distributary a torrent
of slightly turbid water, which flows for 4,000 feet over the debris-strewn floor
and enters the side of the Yoho. At the upper end of the line of avalanched
wood described, this stream has cut a gorge, 40 to 50 feet deep, across a ridge
of limestone strata. The gorge extends beneath the present margin of the ice
and, in all probability, has been cut very largely under subglacial conditions,
this part of the valley being under ice when the distributary completely encircled
the rock embossment. This stream flows for 1,600 feet beneath the ice of the
lower Yoho and contributes the bulk of the water which issues from the cavern
at the nose, the North Fork of the Kicking Horse.1 This stream, although
shallow at first, is rapid and has a breadth of 240 feet, spreading over the gravel
flat with a network of channels. About one-quarter mile from the exit the chan-
nels are collected into a single one, forming a river of very respectable size,
considering its youth. The water is somewhat turbid, but much less so than that
which ordinarily issues from the Victoria. This is because the drainage from the
glacier itself is so largely diluted with that from the adjoining valleys, derived
in considerable part from the simple melting of snow and carrying a minimum
of sediment. Owing to the volume and velocity of the stream, much of the rock

1 Marked Yoho River upon the latest maps of theXHanadianjTopographic Survey.

78 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

debris is carried and rolled down the valley. The channel beds are lined with
coarse rounded boulders, making the fording of the stream, afoot or mounted,
somewhat difficult, especially after a day of rapid melting. In the early morning
the volume and velocity are somewhat reduced.

The dilution of the Yoho drainage, with that from the adjoining valleys,
raises the temperature, as was noted in the case of the Wenkchemna drainage
brook. The last two weeks of August the temperature ranged from 33.8° F. to
35.2° and gave an average of 34.7°. Opposite the Takakkaw Falls, about five
miles from the nose, the stream has descended 770 feet and its August temperature
has been raised from three to four degrees. At Field the temperature of the
Kicking Horse, August 23, 1904, was found to be 44. 2° F. at 5:15 P. M. On
August 30, 1905, at 6:30 A. M. it was 39.6°.

8. FRONTAL CHANGES.

In August, 1 90 1, reference marks about the nose of the Yoho were indepen-
dently established by Miss Vaux and Mr. H. W. Du Bois, from which it was deter-
mined in 1904 that the glacier had receded, in the three years, a distance of in
feet (August 16, 1901 to August 18, 1904), or at the average rate of 37 feet a
year. Measured to the block of ice which had until very recently constituted
the nose, the distance was 92.1 feet, of which 23 feet was for the year 1903-4,
reducing the average to about 31 feet for the three years. Between August 18,
1904, and August 31, 1905, the retreat was found to have been but 9 feet. The
average annual retreat for the four years 1901 to 1905 has been 30 feet. At a
second station upon the western side of the drainage stream the retreat from
August 17, 1904, to August 31, 1905, was 4.6 feet. From these meager data
it seems that the Yoho is having its retreat checked.

9. FORMER ACTIVITY.

a. Moraines. Lack of time prevented any careful survey of the entire Yoho
Valley from the glacier to where the valley joins that of the main Kicking Horse.
No coarse moraines of the type described for the Victoria and Wenkchemna
were seen and their absence is easily accounted for by noting the absence of steep
cliffs about the glacier and its ne" v£ fields, plate xxvni , figure 2 . In passing up the
valley the trail crosses two steep ridges, densely covered with vegetation, but
these appear to be of the nature of mountain spurs, or rock slides, from the
western side of the valley. About i ,000 feet from the present nose there is an
interesting display of modified ground moraine, lying mainly upon the eastern
side of the stream (plate xxvni, figure i). The structures consist of low
knolls and crescentic ridges, connected with the weak lateral moraines by faint
ridges. Six series may be made out, concentrically placed and with their con-
vexities directed down stream, diminishing in height and distinctness toward
the glacier. The ridges vary in height from one to twelve feet, the longest
being in the form of a semicircumference with a radius of twenty feet. The
ridges possess the smooth, rounded outlines of drumlins, but lack their profile

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XXVIII.

FIG. i. — Knolls and ridges of grouml-morainic material in front of Yoho Glacier. Suggestive of
frontal moraines of latest ice sheets in North America and Europe.

Collie. Briker.

Gordon.

Balfo.ir.

Waputik snowfield.

r

Ki<;. 2. — General view looking up Yoho Valley, showing Wapta and Waputik .snowfields. Photographed, 1904, by
Arthur O. Wheeler, from summit of Mt. Wapta (9,960 feet). Looking north-northwest.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHF.RZER.

PLATE XXIX.

FK;. I. — Formation of Forbes's dirt bands from crevasses, Yoho Glacier, August, 1904.

G. 2. — Hanging Valley, head of Yoho, lying between Mts. Balfour and Gordon.

August, 1904.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 79

and arrangement. They are like drumlins in that they consist of ground-morainic
material, with but a thin dressing of gravel and sand. They differ essentially
from drumlins in that their longer axes are parallel with the former ice margin,
instead of at right angles to it. If composed of stratified sand and gravel they
would be kames. In short they have the structure of drumlins and the form
and position of kames. They so much resemble in miniature the knolls and
ridges formed by the last great ice sheet in its retreat from the United States
and Germany that their origin becomes of especial interest. They have evidently
been formed by the glacier, with its nose consisting of a series of lobes, plowing
into the ground moraine previously deposited upon the valley floor. A retreat
occurred and then an advance, falling a little short of the first position, by which
a series of mounds and curved ridges was pushed up. This was repeated,
at least a half dozen times, each advance being somewhat weaker and falling a
little short of the preceding. Finally the glacier advanced over the entire series,
but overrode them so lightly that instead of being destroyed they were simply
smoothed and rounded. In melting back the last time, either from the ice itself,
or from the subglacial drainage, or from both sources combined, a thin layer of
sand and gravel was deposited over the structures. Had the ice lobes been larger
the ridges would have appeared more nearly straight, as we find them in the case
of the Pleistocene deposits.

b. Plucking action. About the nose of the glacier, as has been already
noted, and upon the eastern side, toward the rock embossment, there are numer-
ous illustrations of the bodily disruption of the rock strata, to which the term
plucking is applied. Conditions are most favorable for this action when the
strata are thin bedded and jointed, when the strike of the strata is transverse
to the glacier and the dip is down stream. Under these conditions the rock layers
are ripped off and the bed lowered with relative rapidity, the rock fragments
being pressed into the ice and moved forward to assist in further work of a like
nature. In this way portions of the bed are much roughened by ice action,
instead of being smoothed. The edges and corners of the strata which were able,
at the last, to resist the action of the ice will be found to be rounded more or less.
A mountain spur, lying between the nose of the glacier and Mt. Balfour was
overridden by the ice and experienced this action upon an extensive scale. The
mountain is made up of curved, concentric strata, the upper layers being of a
dark limestone. Upon the southern side these layers dip to the southwest at an
angle of about 30°. In passing over the peak from north to south many feet of
strata have been removed, those able to resist the action forming a succession of
steps upon the steeply inclined slope. One only of these steps is shown in plate
xxvii, figure 2, behind the hard crest of which the loose fragments have collected,
while upon either side they have been swept clean by the ice. This furnishes an
illustration of what is known as a "knob and tail" phenomenon. If the com-
bined height of the successive steps were ascertained we should have a figure
representing the minimum amount of this plucking action upon the southern

8o GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

slope, if not over the crest of the peak. It is well to note that this was done
with a relatively thin sheet of ice, while in the valley bed, with some 3,000 feet
additional of ice thickness, the result, under equally favorable position of the
strata, would have been correspondingly greater. It seems very probable that
the peculiar form of the peak Trolltinder (9,414 feet), just south of Balfour
(plate xxvni, figure 2), is due to similar plucking action.

c. Yoho Valley. There is abundant evidence that the entire valley, from
the Kicking Horse at Field, was occupied by an immense ice stream, seventeen
miles in length, which served as a tributary to the ancient Kicking Horse Glacier,
of Pleistocene time. It, in turn, received short tributaries from the adjoining
valleys and mountain slopes. The valley was filled with ice to a depth of 1,500
to 3,000 feet above the valley floor, by which the lower portion was transformed
into the characteristic U-shape, seen best from below. When viewed from a height
as in plate xx vin, figure 2, the more flaring walls of the upper portion become the
more conspicuous. The valley seems to have had the same general history as
that given for the Lake Louise district, page 61. Being a longitudinal valley
of the Rocky System, it was originally a trough between mountain folds, or a
great crevasse, which collecting the drainage of the region was cut into a V-
shaped valley by the joint action of running water and the weather. With the
coming of the glaciers, the valley was occupied by ice and the lower one-third
to one-half deepened and broadened, while the upper portion, as high as the ice
could operate, was simply smoothed and subdued. Spurs were cut off and
faces exposed to the action of the ice were grooved and fluted, polished or
scratched.

A series of typical hanging -valleys occur along the Yoho beginning with that
of the distributary, the floor of which is not yet uncovered. This ice stream has
not been able to lower its bed as rapidly as has the main Yoho, and when melted
back to the head of the rock embossment there will be exposed a side floor at a
higher level than the main floor. However hanging- valleys, in general, may
arise, this one seems certainly due to the differential effect of the two streams upon
their respective beds. To the right of the distributary there extends a hanging-
valley to the northeastward between Mts. Gordon and Balfour, still occupied
by two glaciers, which appear to have built conjointly a double frontal moraine
(plate xxix, figure 2) . This valley has a double floor, of which time did not permit
an examination. From the photographs taken there appears to be a lake, occupy-
ing a rock-basin upon the lower level. About 3^ miles down from the nose of the
Yoho the Twin Falls drop into the valley from the floor of a hanging-valley
coming in from the west. The falls are 310 feet in height, but their crest (6,500
feet) is 1,050 feet above the floor of the main valley opposite. Five miles down
from the glacier are seen the Takakkaw Falls, in the center of plate xxvm, figure
2 , the crest of which is i , 200 feet above the valley floor. The valley floor from the
glacier to these falls descends about 770 feet, or at the average rate of 154 feet
to the mile. These figures, based upon data supplied by Wheeler, indicate that
the main valley has been lowered from 1,000 to 1,200 feet more than the tributary

.> o

uJ o

8.

,•,«

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 8 I

valleys. With the effect in mind of relatively thin ice sheets upon the neighbor-
ing peaks, the writer is quite prepared to admit the sufficiency of glaciers to pro-
duce hanging- valleys, when the ice is deep, concentrated, and operates for along
period over stratified formations.

CHAPTER VI.

ILLECILLEWAET GLACIER,
i. GENERAL CHARACTERISTICS.

PASSING from the Rockies westward to the Selkirks,we find much evidence
of the increasing precipitation; one of the first to which our attention is unpleas-
antly called is the tantalizing number of snow-sheds which obstruct our view.
The mountains are much more completely forested than we found them in the
Rockies and nearly everywhere the valley slopes are scarred with avalanche
tracks. From extensive snow-fields (plate xxxii, figure i) hundreds of tongues of
ice descend to much lower altitudes than is possible in the Rockies, with their
slighter snowfall. The largest of these ice tongues to be seen from the railway
is the so-called "Great Glacier," or Illecillewaet,1 the glacier that gives rise to the
"rushing water." Owing to the ease with which it may be reached from the
station it has been visited by more people than any other glacier in the two
Americas, although, so far as known, it was not seen by the eye of white man
until the year 1883. In that year it was discovered by Major Rogers, who
was in search of the railway pass which now bears his name. It was originally
named Agassiz Glacier by Ernest Ingersoll,2 but this name has since been
transferred to one of the commensal streams of the great Malaspina, in Alaska.

The glacier lies just to the south of Mt. Sir Donald (10,808 feet), between it
and Glacier Crest, and as a great tongue of ice spills over the rim of the extensive
collecting basin enclosed bet \veen Mt. Sir Donald, Mt. Macoun (9, 988 feet), Mt. Fox
(10,572 feet), and Mt. Lookout (8,2 19 feet). See maps, plate xxx and xxxni. The
glacier flows to the northwest, is but i^ miles in length, and in this distance tapers
from a mile in breadth to a sharply pointed nose. The axis of the glacier is slightly
curved, with its convexity turned toward the southwest. Lying in a broad valley
with this exposure, and with no covering of debris, the glacier receives the full
effect of the noonday and afternoon sun. In spite of this the nose attains the
altitude of 4,800 feet, or 870 feet lower than the Yoho. Since the collecting
areas are very similar in size, this difference must be due mainly to the differences
in the amount of snowfall received by the two regions. The latitude of the nose
of the Illecillewaet is 51° 15', being nearly a third of a degree farther south
than the Yoho. From the neV6 line, with an elevation of about 7,500 feet,
the glacier descends 2,700 feet to the nose, or at the rate of about 2,000 feet to the

1 This name is pronounced as though it were spelled Illy- silly-wet, with the stress upon the middle
syllable.

1 " The Rocky Mountains as Seen from the Canadian Pacific Railway." Science, vol. vii., 1886, p. 243.

82 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

mile. The greater part of this drop is in the upper half, the glacier descending
from the rim of the basin in a steep cascade, by which the ice is shattered and
its original structure destroyed (plate xxxn, figure 2). In its short length the
glacier receives no tributaries, but instead has a series of short distributary noses
perched high up along the eastern line of cliffs leading to Perley Rock (plate xxx) ,
with elevations ranging from 6 ,45 o feet to 7 ,000 feet. No data exist for estimates
upon the thickness of the glacier, but the greatest thickness of ice probably
occurs below the crest of the cascade and may amount to several hundred feet.

The marginal ice is very steep upon the eastern side and for a quarter of a
mile back from the nose upon the western side, so that it can not be easily
ascended. Toward the nose the general inclination is about 30° to 35°, dimin-
ishing to 20° and less, so that one may mount the glacier from the nose for a short
distance. The ne"ve" may be reached by a rather rough climb, either around to
the east by Perley Rock, or by ascending to the depression between the glacier
and the steep left lateral moraine, keeping a sharp lookout at first for rolling
rock from the eastern face of the moraine. About the main nose, upon the
eastern side, there occur a number of minor noses, shown in plate xxx and plate
xxxii, figure 2, and also a sharply defined, trough-like depression in the surface
of the ice, from 200 to 300 feet across and tapering up stream for a considera-
ble distance. This depression appears in all the photographs taken since 1887
(plates xxxvi, xxxvn, figure 2), since which date about 400 feet of the floor
immediately beneath it have been uncovered without disclosing any cause for
the depression. In plate xxxii, figure 2, it appears to be continued up the glacier
to the cascade and may have its origin there in some obstruction of the bed by
which the ice is diverted to either side and left thinner in the lee. Just to the
left of the nose there has been uncovered, since 1898, a mass of bedrock for a
distance of 400 feet, its more rapid radiation of heat accelerating the melting of
the ice resting upon it. The rock consists of a brownish, schistose conglomerate,
furnishing an interesting display of glacial features to be described in another
section of the chapter (page 95). This is the only bedrock observed in the
floor of the valley from the glacier to the station.

2. NOURISHMENT.

Meteorological data at the station of Glacier House are, unfortunately, very
meager. The average precipitation for the five years available amounts to 56.68
inches, of which 43.7 inches (36 feet and 5 inches), or about 77 per cent, of the
whole, fell as snow.1 Over the neVe" region practically this entire amount
would be available for glacier formation and, as snow, would represent about
47 feet. The retangular snow-field extending from Mt. Sir Donald to Mts. McCoun
and Fox is about five miles long, by two miles broad, and hence contains about
ten square miles of collecting area (plate xxxm). Of this area about two-thirds,
or six to seven square miles, drains northward and feeds the Illecilliwaet,
while the remainder moves southward and nourishes the Geikie Glacier, which

' The Selkirk Range, A. O. Wheeler, vol. i, p. 414-

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XXXI.

1. <>• Position of plates, July 31 St., I

Position ofp/atfs. Sept 5th., I89S
Average slope of surface at this point Z£°o'

GLACI ER

,

evergreens, 3
average 29 years

Alder bushes • n '<?":•
'

POSITIONS OK MARKED ROCRS.

Large polished boulder on top of moraine, (Aug. '90.)
fi. — Triangular blcck. (Vaux, 1898.)
C.~ Round boulder. (Vaux, 1898.)
A— Double rock. (Vaux, 1898.)
E. — Large rock in border moraine. (Vaux, 1887.)
/'"- — Large angular block. (Not marked.)
('. — Pyramidal block. (Not marked.)
//. Tongue of ice.
/. — Large bottom rock. (Not marked. I

.— Large bottom rock. (Not marked.)
A". — Large botiom rock. (Not marked.)

- — Large bottom rock. (Not marked.)

.— Small rounded boulder. (C. O. E. 1895.)

. — Large bottom rock. (Not marked.)

.— Large boulder. (Not marked.)
-". — Two boulders. (Oct. '95.)

. — Long flat boulder. (Ice 4th Aug. end of snout, '19.)
*. — Polished boulder of green quartzite. (Not marked.)
•V. — Long, flat boulder. (16 ft. from nearest ice, '90.)

'-— Rocks tarred by W. S. Green, Aug. 13, 1.
V. — Egg-shaped boulder. (Striped with paint.)
/ '. — -Marked rock.
W. — Large rock, point of view of test pictures.

The points on border of ice marked by a star were determined trig-
onometrically.

,

V "

Scale
0 100 200 300 400 500 feet

Tongue and Moraines of the Illecillewaet Glacier, August, 1899. Published through the courtesy of George and William S. Vaux, Jr.

Proceedings of the Philadelphia Academy of Natural Sciences, 1899.

From the

OF TH:
UNIVERSITY

OF

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 83

flows westward between the Asulkan and Dawson ranges. The collecting
basins for the Illecillewaet and Yoho glaciers are almost equal in size, but the
estimated precipitation over the former is 41 per cent, greater than over the lat-
ter, giving a correspondingly greater volume of ice to be disposed of. This ena-
bles the Illecillewaet to attain a much lower altitude, as previously pointed out.
The mean elevation of the surface of the neVe" lies between 8,000 and 8,500
feet above sea level; ranging from 7,500 to 9,500 feet, not essentially different
from the Yoho neve. In the central portion the nevd is much crevassed, from
which one may infer that the thickness of the snow and ice is not great in the ba-
sin. The surface is covered with parallel ridges and furrows, probably result-
ing from the rippling action of the wind while the snow was being deposited.
These ridges when frozen render the walking somewhat difficult and treacherous.

3. MORAINES.

a. Surface debris. Because of the wide extent of the neVe" field and the
absence of precipitous cliffs about the margins, there is very little opportunity for
the ne"ve" to acquire any rock debris. There is no evidence that any considerable
quantity is gathered from the bed and carried subglacially, or englacially. It
may be that the basal layers in the neVe" region are sluggish, or even stagnant,
and that only the upper layers are being pressed out over the rim of the basin.
The result of this lack of debris in the neVe" is that the general surface of the
glacier, as in the case of the Yoho, is unblemished with rock fragments, but is
somewhat soiled from wind-blown dust concentrated over its surface by melting.
Some dust wells occur sparingly and a few poor examples of glacial tables.

b. Left lateral moraine. Along the western margin of the glacier fed by
the ordinary atmospheric agencies of rock decay operating upon the cliffs of
Glacial Crest and Mt. Lookout, there has been built up a high, sharp-
crested, left lateral moraine. The angular rock fragments are supplied
mainly by two prominent debris cones, which have formed upon the eastern
face of Glacier Crest and which rest with their bases upon the margin of the ice,
the forward movement of which has distorted the cones down stream, plate xxxn.
In the summer rocks may be seen coming down these slopes, one block starting
others and these still others, until a regular cannonading is in progress. As one
rock collides against another with terrific force, small clouds of dust arise and
we have simulated not only the roar but the smoke of 'battle. John Muir has
given us a graphic description of the streaks of fire to be seen when these ava-
lanches occur at night. In the way previously described for the Victoria (page
47) this material arranges itself in the form of a sharp-crested ridge perhaps
100 feet above the margin of the ice and 150 feet above the valley floor. The
melting of the ice core upon the inner slope makes it steep and unsteady, while
the outer has settled into a condition of more stable equilibrium and is being
slowly covered with vegetation. Upon the inner slope occasional slides of the
rock veneering occur, by which the ice core is temporarily exposed. The materials
from the upper part and sides of the scar produced roll and slide down, collecting

84 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

at the bottom and growing slowly upward until the ice is again completely
covered. In this way the material is being more thinly spread over the morainic
ice core and its melting accelerated. The great height which the moraine has
been able to attain in its lower part is due to the fact that the glacier is unpro-
tected by debris over its general surface, and the differential melting is so much
greater than it would be if the surface of the ice were protected as in the case of
the Victoria.

The material of which this moraine is composed is principally quartzite and
a silvery (sericitic ?) schist, with a binding of glacial sand and clay. Occasionally
boulders are scratched, and are very generally bruised, with their edges and
corners rounded from rough treatment they have received upon the debris cones
and the moraine. About one-third of the distance along this moraine there occur
two elongated depressions, exactly in the crest of the moraine. The larger and
better defined of the two is 125 by 50 feet and 6 to 8 feet in depth (plate xxxn,
figure 2). They are seen from a distance most plainly when the sun's rays
strike somewhat obliquely, permitting the sides to cast shadows into the bottoms
of the depressions. Attention was first called in print to these depressions by
the Vaux Brothers in 1900, l showing very plainly as they do in their early
photographs. Before these depressions were visited it was thought by the writer
that they might represent the sites of drained lakelets, similar to those found
upon the Victoria, but the explanation suggested by the above investigators
seems to be the correct one. The moraine appears to have been formed of two
ridges, laterally welded together, and these depressions appear to be spaces
left where the two ridges did not quite meet. Owing to the sliding of the de"bris
upon the inner slope of the moraine the basins are being obliterated and will
eventually completely disappear.

Including the debris cones this lower portion of the left lateral is some 4,000
feet in length, rises to less and less height above the general ice level, and gives
out for a short distance, where the ice abuts directly against the quartzite cliffs
of Glacier Crest. Up to a height of 40 to 50 feet patches of morainic matter have
lodged upon the rock shelves. One-quarter mile beyond, a crested ridge again
makes its appearance composed of materials derived from Mt. Lookout, just to
the southeast. The moraine is largely made up of quartzite and schistose boul-
ders bound together with sand and clay and supporting a sparse growth of
mosses and Alpine plants. The ridge rises to a height of 35 to 40 feet above
the ice, curves around to the eastward, becomes reduced in height, and disap-
pears under the neVe", which is strewn with rock fragments derived from the
cliffs of Mt. Lookout.

c. Terminal moraines. From the lower extremity of the left lateral there
curve around into the valley two lower ridges, of the nature of terminal moraines.
The inner and younger of these forms an inconspicuous ridge, from 6 to 10 feet
high, and passes into the terminal moraine at which the glacier was found to be

i " The Great Glacier of the Illicilliwaet." George and William S. Vaux, Jr., Appalachia, vol. ix,
1900, p. 164.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

Sir Donald. Iltecillewaet Glacier. Asulkan Glacier.

PLATE XXXII.

Mt. Bonney and glaciers.

FIG. i. — General view of peaks and snowtields from Rogers Peak (10,536 feet),
Selkirks, looking southeastward. Photographed in 1902, by Arthur O. Wheeler.

Illecillewaet neve.

Asulkan neve.

K;. 2. — General view of Illecillewaet Glacier from summit of Mt. Eagle, elevation 9,353 feet, distance two to three

miles. Photographed in 1903 by C. F. Johnson.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 85

standing in 1887. From their photographs taken in this year the Messrs. Vaux
have established the position of the ice front with reference to a very large
boulder resting upon this moraine. This terminal ridge swings around to the
north and connects with the right lateral, which is of greater age and, in the lower
part at least, has lost its ice core. See plates xxx and xxxi.

d. Right lateral moraine. The ice has withdrawn from this moraine a dis-
tance of 400 to 500 feet, leaving a somewhat subdued boulder slope and a low
ridge. This becomes higher and steeper as we approach the quartzite cliff which
intercepts it about one-half mile back from the nose. Here the moraine is
double, an older one lying just outside and parallel with it. Forest trees have
taken possession of the crest and outer slope. The rocks in the right lateral
are similar to those in the left and are found to be in the same condition of
being rounded and bruised. Upon the rocky ledges, which carry the distribu-
tary noses referred to, there is spread out more or less morainic material, some
of which has been assorted by running water. These ledges of quartzite have
been much glaciated, plucked and extend up toward Perley Rock (7,898 feet).
For about a quarter of a mile there extends an upper double moraine to the
southeastward, where it disappears under the snow. The material consists of
rounded boulders of quartzite and chloritic schist, with a filling of glacial sand.
An inspection of the map shows that there is a correspondence in the arrange-
ment and position of the lateral moraines; there being in both cases, a higher and
a lower portion, separated by quartzite ledges, carrying only a sprinkling of
morainic material. Since the cascade in the glacial stream lies between these
exposures of quartzite it is probable that the ledges are continuous beneath
the glacier; that they have proven too hard for the glacier to remove, and so it
is compelled to cascade over them.

e. Boulder pavement. Between the terminal moraine and the present nose
of the glacier there has been 'uncovered since 1887 a broad boulder belt, about
500 feet across. This consists of ground-morainic material in large part, with
the rock fragments which were carried englacially, or supraglacially, and de-
posited as the ice front receded. These boulders have been overridden by the
ice so lightly that they have not been disturbed, and yet a number of them were
glaciated while in their present position, forming what is known as a "boulder
pavement." About the present margin of the ice, boulders are being continually
uncovered which are being subjected to the same action. The ice presses against
the up-stream face of the boulder, and, either because of the warmth of the stone,
or more probably because the melting point of the ice is reduced by the pressure, or
because of both these agencies, an inverted trough, or fluting, is produced upon
the under surface of the ice, having the form of the rock. In plate xxxiv we have
these flutings shown in different stages of formation; in the last case (figure 2)
the stone was estimated to lie 70 feet back from the ice margin and was under
probably 50 feet of ice. Photographs taken some years ago of the "ice grotto"
show that it was a feature of this kind produced by an unusually large rock.
If the pressure were sufficient, the ice would settle in promptly upon the lee side

86 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of the boulder and it would be glaciated not only upon the stoss and upper sur-
face, but upon the lee side as well. A certain relation must exist between the extent
of lee-side glaciation and the thickness of the ice, which, if known, would give
some data for estimating the maximum thickness of certain Pleistocene sheets.

4. CREVASSES.

The crevassed condition of much of the n6v6, especially that in the main
direction of flow, has been noted. Faultings and dislocations occur, disclosing
the stratified nature of the ne"ve and subjacent ice. This crevassing is due,
apparently, to irregularities in the bed, rather than to differential motion, and
indicates that the ice here attains no great thickness. About the margin of the
neVe" field there occur, here and there, breaks where the ne"ve" has withdrawn
from a portion still clinging to the rocky wall. These are the bergschrunds
described upon the Victoria and Wenkchemna glaciers (pages 22 and 67). On a
line between Perley Rock and the western end of Mt. Lookout, there opens up a
series of transverse crevasses as the ice begins its descent into the valley and its
velocity is accelerated (plate xxxn, figure 2). The ice is unable to yield to the ten-
sional strain and forms long V-shaped gashes at right angles to the stress. These
become convex down stream because of the more rapid central movement of the
ice. Conditions are here favorable for the formation of Forbes' dirt bands (page
50), but the ice soon plunges over the quartzite ledges, is shattered in every
direction and all structure lost. Upon the crest of the cascade a network of
crevasses opens, dividing the ice into irregular angular blocks. These become
melted upon all sides and assume the form of pinnacles and steeples, — seracs,—
displaying beautifully the stratified structure of the ice (plate xxxv, figure 2).
Reaching the bottom of the cascade these blocks are jumbled together, many of
them completely melted and the remainder frozen together into a great ice con-
glomerate (plate xxxv, figure i)'. As pointed out upon page 44, it is quite
conceivable that some of the basal layers might be able to descend the slope
without having their structure destroyed.

The rapid central movement of the ice, due to the high average slope of the
bed, gives rise to a very complete system of marginal crevasses, extending in-
ward and upward, and showing conspicuously when the glacier is seen from a
height. In the lower third the ice does not feel the restraint of the valley walls,
spreads laterally because of its own weight, and there are opened longtitudinal and
radial crevasses, some of them extending to the margin of the ice. As the
surface is continually lowered by melting, only the bottoms of some of the
shallower crevasses remain, and these appear simply as short gashes in the other-
wise smooth surface.

5. ICE STRUCTURE.

There is evidence upon the glacier's left, back a short distance from the nose,
that the stratification in the basal portion of the glacier is not completely de-
stroyed in passing the cascade. Traces of the stratification may be seen dipping

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XXXIII.

I? t I:'

£&r %«_

/

s

,

Tfe*

Ma]> of the Selkirk snowfields and glaciers, by Arthur O. Wheeler. Reproduced through courtesy of Canadian Topographic Survey,
Department of the Interior. Approximate scale I inch = \]4 miles. Contour interval 100 feet. Reference datum mean sea-level.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHER2KR. PLATE XXXIV.

FIG. I. — Beginning of subglacial fluting by pressure-melting, Illecillewaet Glacier,

August, 1905.

FIG. 2. — Subglacial fluting by pressure-melting, Illecillewaet Glacier, August, 1903. Photo-
graphed beneath the glacier and at an estimated distance of 70 feet from the
rock responsible for the fluting.

Fir,. 3. — Roches moutonne'es near nose of lllecillewaet Glacier. Motion of ice from

right to left.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 87

inward towards the center, this portion of the ice possibly having received less
severe treatment than that nearer the center of the channel. In the case of the
Yoho the question arose (page 76) as to whether the stratification was obscure
because of its destruction by a similar cascade, or because of its original weak
development. In the case of the Illecillewaet there is sufficient bare peak and
rocky crest exposed to supply the broad ne"ve" field with successive layers of wind-
transported dust and a very perfect stratification results from the concentration of
this dirt at the surface of successive deposits (plate xxxv, figure 2). The almost
complete lack of stratification about the nose, where it should be well displayed,
along with the dirt zones, must in this case be ascribed to the cascade. The dust,
originally concentrated between the strata, is brought to the lower margin of the
ice, where it collects and drips as black mud (plate xxxiv, figure i) over the val-
ley floor. The color is due to the presence of organic matter, of which there is
enough present to render the material offensive, when set away damp in a warm
room. Four determinations of the organic matter present in material collected
in 1903 gave 16.75, IJ-25> ro.68, and 17.23 percent., or an average of 13.98 per
cent.

It seems impossible that the coarse stratification of the ice could be so com-
pletely destroyed and the finer lamination preserved so perfectly and continuously
as we should have to suppose if we referred the blue bands to the original lami-
nation of the neVe". As pointed out upon page 44 and as shown in plate xm, figure
i , the blue bands, with the superficial dirt stripes, are very clearly shown about the
nose of this glacier, from 15 to 36 being counted within the distance of a foot.
They are approximately parallel with the valley floor and would probably con-
form with the strata, providing the latter were present. They dip inward, in
general, about the nose at angles of 3° to 8°, but in places are inclined outward
by this amount. As soon as the ice begins to experience pressure from the
moraines, or the valley walls, the blue bands become more and more steeply
inclined, beginning with angles of 8° to 16° and increasing up the valley to 70°
to 75°. The relation of bands in the ice to the stones which are fluting the
under surface (plate xxxiv,) has been discussed upon page 44.

The glacial granules, with the melting phenomenon described in connection
with the Victoria Glacier, are well shown about the nose. In size they stand
next to those of the Yoho and range from the size of hickory nuts to that of hen's
eggs. They are limited largely to the blue bands, or the white seams that
intervene, but in cases are seen to cut across from one to the other. As the
granules assume distinctness the blue bands become more and more obscured.
Between the granules there is developed, under suitable melting conditions,
a very perfect and beautiful network of capillaries described upon page 41.

6. DRAINAGE.

a. Surface and marginal drainage. Upon the nights of September 7-8 and
8-9, 1904, the minimum temperature of the ice was measured by inserting a ther-
mometer to the depth of twelve inches in the face of a crevasse near the nose.

88 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

The two readings were 32.o°Fand 31.9°, indicating that this portion of the glacier
was very near to its melting temperature. Before liquefication, however, can
occur a large amount of heat must be rendered latent, when water is produced with
the same temperature as that possessed by the ice just before melting. The
heat necessary for this conversion is supplied in the main directly from the sun
but in part by that of the atmosphere, rain, friction and pressure of the ice
against the valley floor and sides, and whatever heat may be reflected, radiated,
or conducted, from the same source. Almost without exception the surface
drainage was found to have a temperature of 32°, varying but a very small
fraction of a degree from this. The surface ablation is rapid during the summer,
owing to the exposure of the ice to the sun and the absence of protective debris.

As would be expected from the greatly crevassed condition of the ice, no
surface streams of any size can develop, either over the general surface or along
the margins. Small streams flow directly to the ice margins and cut channels,
in a few cases, to the depth of a foot. This water may be absorbed at once by
the loose materials covering this portion of the bed, or it may collect and give
rise to scant marginal or subglacial drainage streams. These flows soon cease
when the sun drops behind Glacier Crest and throws the glacier into shadow.
In 1904, for a distance of some 500 feet, a small drainage stream was found
between the left lateral and the adjoining ice slope. Since the ice of the glacier
is here continuous with the morainic ice core, this stream was a surface rather
than a marginal one. From the lower end of this moraine there occurred also
two small flows of water, apparently coming from englacial channels in the
moraine. After a six hours' rain, September 8, 1904, during which 0.58 of an
inch of rain fell, the water of these streams was rendered muddy, while the
turbidity of the main glacial flow was not perceptibly affected.

b. Terminal drainage. Upon the eastern side of the glacier, drainage
streams leave the ice margin in the neighborhood of the elevated distributary
noses, as shown upon the map. The ice here rests upon bedrock and the streams
have sought the lowest depression, there being three such which are draining this
portion of the glacier. The westernmost of these breaks into a network of streams
upon emerging from the ice, and cascading over the rocky ledges again enters
the side of the glacier, to reappear at the nose. The two other streams have
cut gorges 50 to 60 feet deep across the hard strata, showing that they must
have been at work for considerable periods, probably as subglacial streams.
However, the high velocity of the water and the sharp glacial sediment enable
it to work very effectively, even upon quartzite. These two cascade over the
ledges and unite some 1,600 feet down into a single stream, which receives
tributaries from the slopes of Sir Donald. Just outside of the right lateral
moraine the stream divides into numerous branches, which reunite upon the
gentle gravelly slope and form what is known as a "braided stream." East-
ward of the upper, right lateral moraine, between it and Perley Rock, there is a de-
serted gorge, similar to those now being formed, partially filled with gravel and
morainic matter. The drainage has been deflected westward to a lower level.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XXXV.

FIG. i. — Regelation of ice blocks at foot of ice cascade, Illecillewaet Glacier, September, 1904.

Kit;. 2. — Stratification in upper part of Illecillewaet Glacier.
Copyrighted, 1902, by the Detroit Photographic Co.

GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS. °9

In 1904 there issued from the main nose of the glacier five drainage brooks,
the main one lying to the west of the nose and receiving the drainage from the
left lateral moraine. This stream follows the inner curve of the terminal moraine
and cuts across it at a point opposite the nose. The other four streams form a
network over the boulder pavement, unite into a single stream, which makes its
way across the terminal moraine and joins the main flow. About 1,200 feet
from the nose of the glacier (1904) this drainage brook from the lower and west
side of the glacier unites with the strong flow from the eastern side of the valley
and together they form the Illecillewaet, or "Rushing Water," a tributary of the
Columbia. The average flow from the glacier is not much different from that
of the Victoria and Wenkchemna, but considerably less than that from the
Yoho, which collects the drainage from a larger territory. Based upon the
estimated size of the neV6 area and the average annual precipitation, the average
flow for the months May to September, inclusive, should be about 65 cubic feet
per second. Owing to the somewhat larger drainage area and the more rapid
midsummer melting the flow seems greater than this in July and August. The
water, as it issues from beneath the nose, is only slightly turbid compared with
the Victoria, indicating a small amount of subglacial erosion. The color becomes
slightly green in the combined streams and still more so after it has received the
Asulkan drainage farther down the valley. With the loss of sediment, which is
gradually stranded here and there, the water assumes a bluish tinge, except
where lashed into foam.

c. Temperatures. During the last week in August and first week in Sep-
tember, both in 1904 and 1905, some 30 observations were made on the tempera-
ture of the drainage at the nose. The average temperature, taken at various
times of the day, was found to be 33.1° F., with a range from 32.4° to 33.8°.
That from the eastern side of the valley, taken just under the bridge on the
trail, gave an average of 39.9°. One-third of a mile down the valley, at the lower
bridge across the stream, the average temperature of the combined streams
was 38.3°, ranging from 34.9° to 40.6°. Where the Illecillewaet passes beneath
the railway, having received the Asulkan brook, four observations upon the
temperature gave an average of 39.7 °.

7. FORWARD MOVEMENT.

As early as 1888 observations were made by Rev. W. S. Green to determine
the forward movement of the glacier. On August 13, he set three poles in the
ice by boring holes with an auger, the distance from the nose not being given.1
These were visited upon the 25th of the same month and were found to have
fallen, owing to surface melting. The holes were, however, found and the
poles reset for measurement. The distances moved in the twelve days are
recorded as follows: " No. i pole, near moraine, 7 feet; No. 2, further out, 10 feet;
center of glacier, 20 feet." For the middle of the glacier this gives an average
daily motion of 20 inches. About the margin of the ice Green, at the same time,

' Among the Selkirk Glaciers, 1890, p. 218.

9o

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

tarred a number of boulders in closest proximity to the margin, which rocks
could still be indentified in 1905. (See rocks marked T, plate xxxi.)

In 1 899 George and William Vaux set a line of 8 steel plates across the glacier,
some 1,400 to 1,500 feet back from the nose, their line lying somewhat obliquely
to the main axis of the glacier (see map) . The average surface slope was given
as 22°, and the distance across the glacier along the line of plates was 1,720 feet.
A base line was laid out along the higher portion of the right lateral moraine,
229.5 feet in length, and the plates located by triangulation, July 31, 1899. Bear-
ings were taken upon the plates August 1 1/189 9, and Septembers, 1899, the lat-
ter work being done by Messrs. H. B. Muckleston and C. E. Cartwright, of
the Canadian Pacific Railway. One year later (August 6, 1900) the plates were
again located by the Messrs. Vaux and their forward movement for the 372
days determined. The following table is based upon their data published in
Appalachia, vol. ix, 1900, p. 160, and the Proceedings of the Academy of Natural
Sciences of Philadelphia, March, 1901, p. 215.

TABLE VII,

OBSERVATIONS UPON THE LINE OF STEEL PLATES SET ACROSS THE
ILLECILLEWAET GLACIER, JULY 31, 1899.

(Total Distance across Glacier along Line of Plates 1,720 feet.)

Plate.

Distance
from
East
edge.

July 31 to Aug. ii, 1899.

Aug. ii to Sept. s, 1899.

July 31 to
Sept. s, 1899.

July 31, 1899 to Aug. 6,
1900.

Total motion,
ii days.

Average
daily motion.

Total motion,
25 days.

Average
daily motion.

Average
midsummer
motion, 36 ds.

Total motion,
372 days.

Average
daily motion.

Feet.

Feet.

Inches.

Feet.

Inches.

Inches.

Feet.

Inches.

I

263

3-54

3.86

2.62

1.26

2.56

88.6

2.86

2

500

3-33

3.64 ; 8.67

4. 16

3-9°

124.0

4.00

3
4
S

605

75°
845

6.25

6.21

S-96

6.82

6-77
6. so

8-75

4.20

5-51
6.77
6.06

139-7
181.0
188.0

4-51
5-84
6.07

11.71

5.62

6

980

6-37

6.96

r3-79

6.62

6.79

197.0

6.36

7
8

1040
1310

5 .00

5 -5°

5-45
6 .00

M.33

6.88

6.16
6.00

158. S
170.0

S-n
S-48

An inspection of the above table shows that the maximum movement, for
both the summer and the entire year, lies well to the west of the axis of the
glacier. The greatest average daily movement was made by plate 6, which
lies 120 feet to the west of the center, while plates 7 and 8 show only slightly less
movement. This is in harmony with what is known concerning the flow of
glaciers on a curve, the maximum movement taking place not at the center,
as in the case of the very straight Victoria, but lying between the center and the
convex side. The average daily movement of plate 6 for the year is 94 per cent
of its summer motion, as compared with 81 per cent for the corresponding plate
upon the Victoria. For some reason, not easily explained from the data at hand

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. QI

the mean summer motion of the two most easterly plates was less than their
yearly average. It is to be noted that the maximum summer movement of 6.96
inches (July 31 to August n), 1899, is but about one-third of the maximum
movement observed by Green in 1888 (August 13 to 25). The only way to
reconcile the two results is to suppose that Green's measurements were
made farther up the slope towards the cascade, where the movement is un-
doubtedly much greater than towards the nose. Messrs. Vaux placed a
ninth plate upon the nose of the glacier and had it under observation from
August i to August 20, 1899. The average daily horizontal motion for the
first two intervals between measurements was 5.9 inches and 5.0 inches. A
crevasse then formed, detaching the block carrying the plate, and the subse-
quent apparent motion was 2.8 inches and 2.7 inches daily.

8. FRONTAL CHANGES.

a. Recession data. Owing to the easy accessibility of the glacier and its at-
tractiveness to the ordinary visitor, we have more data from which to determine
the frontal behavior of the Illecillewaet than any of the other Canadian glaciers.
As has been noted, from the photograph taken in 1887 by the Messrs. Vaux the
position of the ice at that time, with reference to a large boulder, was determined
and in 1898 marked conspicuously. In 1888 the margin of the ice was marked by
Green and the glacier was photographed by Notman & Son, of Montreal (plate
xxxvi, figure i). Reference blocks were marked in 1890 and 1895 by interested
visitors. A visit was paid to the glacier September 3, 1897, by Albrecht Penck,
of Vienna, and a sketch made of the tongue of the glacier and its relation to the
lower moraines. This was published in the Zeitschrift des Deutschen und
Osterreichischen Alpenvereins, Jahrgang 1898, Band xxix, s. 55, under the title
" Der Illecillewaetgletscher im Selkirkgebirge. ' ' The height of a number of points
was determined by an aneroid and four reference blocks established and located
upon the map. These blocks were left to be marked by a railroad employee,
but were apparently neglected and in 1904 could not be identified with absolute
certainty, owing to the changes in the ice margin. Based upon the railway
elevation at the station, Penck determined the elevation of the nose in 1897 as
4,793 feet (1,461 .meters). The foot-bridge, just beyond the modern terminal
moraine he gives an elevation of 640 feet above that of the station, or above sea
level 4,760 feet, 1 and this he uses as his datum for elevations about the glacier.
The nose of the glacier at this time lay 33 feet above the floor of the bridge, and
the crest of the adjoining lateral just opposite was 131 feet above the valley
floor at the nose.

In the year 1 898 a number of reference blocks and range lines were established
by the Messrs. Vaux and have since done excellent service in measuring the
frontal movements. In August, 1899, they made a very detailed survey of the
nose and adjoining region and prepared a large scale map which is of the greatest

1 The correction of the railroad levels reduces this elevation by 27 feet, giving the bridge-floor 4.733
feet.

Q2 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

value in the determination of changes in progress l (see plate xxxi). They have
very carefully located upon their map the reference blocks established by them-
selves and others, so that they may be readily found upon the ground. To their
untiring zeal and devotion to the cause we are very largely indebted for our
knowledge of the behavior of the Illecillewaet front since the year 1887. From
their reference blocks .the writer took measurements in 1902, 1903, 1904, and 1905,
and in 1904 established four other reference stations about the side from which
to determine the marginal changes.

There is reason for thinking that the glacier in 1887 was just completing a
rather prolonged halt at the younger of the frontal moraines described. That
it had not recently extended much beyond is proven by the size of the alder
bushes growing about the outer slope. That it had made a rather prolonged
halt at this line, either at this stage or a previous one, is shown by the size of
the moraine, which, with the small amount of debris carried by the glacier
would require a considerable time in building. From the early photographs it
is seen that the glacier was much bulkier and broader at this stage and the slopes
about the nose much steeper, enabling the glacier to maintain well its position at
the moraine (plate xxxvi, figure i). During the year 1887 to 1888 it had begun
to withdraw from the moraine, as shown clearly in the Notman view just referred
to and as indicated by the rocks blotched with tar by Green. The retreat began
somewhat gradually and attained its maximum between 1890 and 1900, averaging
for these ten years about 5 3 feet per annum. The average for the opening lustrum
of the century is 19.6 feet, the retreat being reduced until it amounted to but
two feet for the year 1904-5. For the 18 years from 1887 to 1905, the horizontal
retreat from the Vaux reference block was 597.5 feet, or at an average yearly
rate of 33.2 feet. It should be noted, however, that this measurement is not
in a line with the main axis of the glacier. The available data concerning this
glacier are given in summarized form below. The measurements were taken
variously, most of them between the middle of August and the middle of Sep-
tember, so that the retreat assigned to some years, may belong in part to the
preceding, or the following year.

Recession Data of the Nose of the Illecillewaet Glacier.

1887-1888. 10 to 15 feet.

1888-1890. Average rate about 23 feet

1890-1898. Average rate of 56 feet.

1898-1899. 16 feet.

1899-1900. 64 feet.

1900-1901. 15 feet.

1901-1902. 48 feet.

1902-1903. 22 feet.

1903-1904. ii feet.

1904-1905. 2 feet.

1905-1906. 84 feet.

' " The Great Glacier of the Illicilliwaet," George and William S. Vaux, Jr., Appalachia, vol. ix, Mch.,
1900, p. "156.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XXXVI.

FIG. I. — Illecillewaet Glacier in 1888. Photographed by Notman & Son, Montreal.

FIG. 2. — Illecillewaet Glacier in 1905, from approximately the same view-point as figure i.

GLACIERS OF THE CANADIAN ROCKIES AND SEUC1RKS. 93

Upon the face of the bedrock exposed near the nose a mark was established
September 16, 1903, immediately beneath the nearly vertical side of the ice,
the height of which was estimated as 60 feet. August 24, 1905, it was found
that the ice had withdrawn laterally 2.4 feet from the face. Passing around from
the nose eastward, three stations were established along the margin of the ice.
A large boulder was found just emerging from the ice, the first week in September,
1904, and marked "Face emerging, Sep., "04." Upon the 24th of August, 1905,
it was found that the ice had retreated here 14 feet. Farther along a medium-
sized boulder had been marked in 1903, "15 ft. to ice. ix-i6-c>3." By Sep-
tember i, 1904, a retreat of 12.5 feet had occurred here, while at the upper
station the boulder "27 ft. to ice. ix-i6-c>3," measured September 3, 1904, 27
feet, and August 25, 1905, 27.1 feet. These data indicate that the margins of
the ice have been receding as we approach the nose, more rapidly upon the eastern
side, but that farther up along the margin there has been no change for the last
two years and, very probably, for a considerably longer time The two views on
plate xxxvi, taken from almost identically the same view-point, the former in
1888 and the latter in 1905, furnish a good opportunity for noting the changes
produced in the glacier in the 17 years. It seems almost possible to recognize
the individual trees standing to the right of the center, but the lower half of the
glacier is unrecognizable. A stadia and trignometric survey of the Illecillewaet
and Asulkan glacier tongues was made in 1906 by the Messrs. Vaux and a
report made to the Philadelphia Academy of Sciences. Some additional data
concerning the movements of the steel plates upon the Illecillewaet were collected
and appear upon plate xxx of this report.

b. Ice waves. In comparing their photographs made in 1898 and 1899 from
a certain large boulder, just west of the trail, Messrs. Vaux noted an apparent
thickening in the ice just beneath the neVe" line. By drawing a delicate line be-
tween corresponding points in figures i and 2, plate xxxvi, that may be recog-
nized in the upper ne"vd region, it is seen that the ice margin along the sky-line
stands slightly higher in the 1905 view. The difference is slight, however, and can
represent but a few feet. When the Notman view of 1888 is compared with a
second, which was made in 1897, and here reproduced in plate xxxvn, figure 2,
the heaping of the ice line beneath the neVe" line is still more plainly seen. There
is thus evidence that a wave, or impulse, derived from an increased precipitation
over the ne"ve" region, travels the length of the glacier and gives rise to a halt,
or an advance, of the front; followed by a depression which permits of a retreat.
Such a depression appears to have been at the edge of the ne"ve" line in 1887 or 8,
while the glacier about the lower extremity was experiencing the effect of a pre-
vious impulse. The retreat of the glacier was greatest between 1890 and 1900,
and if we assume that it culminated at about the middle of the decade, it required
about 8 or 9 years for this trough of the wave to reach the nose, or at the average
rate of 800 to 850 feet per annum. Since in 1905 the appearance of the sky-line
along the ne"ve" corresponds so nearly with that seen in 1888, we may assume that
the crest of the wave was in this position at a date only a little later than the

94 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

mean of the two. This would bring it to about 1898 or 9, when it was especially
noted by the Messrs. Vaux. The gradual reduction in the rate of retreat observed
during the past three seasons would indicate that this impulse is making itself
felt about the nose and that either a halt, or an advance, is about to be in-
augurated. If we are correct in the inference that the neVe" line was marked by
a trough, or low stage in the height of the ice, about 1887 or 8 and that a return
to this condition has been reached about 1905 or 6 with a crest, or high-stage
condition of the ice between, an interesting relation is at once established with
Bruckner's climatic cycle (page 16). The time between the appearance of these
troughs for the passage of one-half of the wave is 18 years, and we may venture
to predict that the present relative depression will be followed by the passage
of another crest requiring about the same number of years. The nose has been
in retreat from 1888 to 1906, some 18 years, and we should expect another period
of halt or advance to soon set in. Such a condition was to be anticipated from
the marked reduction in the rate of retreat from 1902 to 1905, but the very
remarkable recession of 84 feet determined by the Messrs. . Vaux for the year
1905-1906, leaves the matter in doubt.

The relation of the glacial movements to the precipitation cycles becomes
a matter of much interest and here, as above, with our meager data, we can only
point out possibilities, which will either stand or fall, when the next half -century's
observations have been collected. From our available meteorological records
there was a deficiency of precipitation over the mountains from 1885 to 1896;
how much before 1885 this condition existed we have no means of knowing.
Since 1897 there seems to have been an excess over the normal amount, but it
was at this time that the crest of the wave made its appearance at the ne"v£ line.
Then instead of continuing to increase, as we might expect, it gave way to a
trough. The inference is, and it is only an inference, that this wave represents
the gush of ice from the collecting basin due to the excess deposited during the
phase of the cycle which antedated 1885, and probably to be correlated with the
excess in the Rockies, as recorded so strikingly in the evidence of higher lake levels,
described by Dawson (page 17). The approaching trough shown about the ne" v£ line
in 1905 must then be ascribed to diminished precipitation received over the collect-
ing region from 1 885 to 1 896, being then 9 years delayed from the close of the phase
which gave rise to it. The crest of the wave from the basin was delayed some
17 to 1 8 years from the close of the preceding phase. In a paper read before
the International Geographic Congress, at its Washington meeting (Proceedings,
1 904, p. 48 7) , Doctor Reid gave a discussion of ' ' reservoir lag, ' ' in which he demon-
strates mathematically that the thickening of ice in the collecting basin does not
keep pace with the variation in precipitation, but lags behind it. In the case of
large glaciers this lag amounts to about one-fourth of the period of the variation,
and the ice in the basin should attain its maximum thickness, only about the time
that the annual supply has settled back to the normal amount and is ready
to diminish. After the maximum ice thickness has been attained toward the
center, time is still required for the impulse to reach its maximum at the crest of

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 95

the basin, the amount of which will differ with the local conditions. In this way
we may account for the delay in the arrival of the crest at the neVe" line in the
years 1897-9.

9. FORMER ACTIVITY.

a. Rock scorings. The former work of the glacier is shown in great
beauty and variety upon the mass of bedrock now being gradually uncovered
near the nose. The hard rock features of Pleistocene glaciation are all here for
study by those interested, many of them indicating the direction of ice move-
ment and hence of practical value in the field.1 Excellent examples of the so-
called roches moutonnees occur, groups of which in the distance often resemble
crouching sheep (plate xxxiv, figure 3). In the specimen figured, the ice moved
from right to left across this projection of bedrock, the up-stream, or stoss side being
rounded and smoothed, while the down-stream, or lee side, was affected slightly,
or not at all. Portions of the rock were polished, as the ice was rubbed vigorously
across, and where the ice held rock fragments against it, systems of approximately
parallel scratches were produced, some so fine that they must have been made by
sand grains. At the last stage of the disappearance of the ice from this particular
roche moutonne'e, a small clump of rock fragments was gently dropped upon the
upper surface in insecure position. An inspection of this and the adjoining rock
in the figure, shows a system of parallel joints, dipping down-stream at a steep
angle. From the lee side of the central roche moutonne'e it is apparent that an
entire block was pried loose by the ice and that a little more vigorous action at the
joint, just beginning to open, would have removed bodily nearly the entire block.
This action is known as "plucking," already described in connection with the
Yoho (page 79), by which the work done in a few days may exceed the erosion
of years. Places may be seen upon the surface where a rock engaged in produc-
ing a shallow groove has made a succession of jumps and given rise to a series
of short parallel curves, more or less closely placed, with their concavities
directed down-stream. These are the "chatter-marks," the production of
which may be illustrated by pushing a dry finger over a polished surface. In
other cases rocks embedded in the under side of the ice have been suddenly
brought into action, producing a crescentic gouge, with its convexity directed
in the direction of flow.2 The bedrock here being a schistose conglomerate
with rather coarse, hard masses embedded in a softer matrix, there have been
produced the "knob and tail," or "knob and trail" phenomena, so useful often
in determining the direction of ice flow in the case of Pleistocene glaciers. In
one case examined there appeared a dark colored knob of harder material,
which the ice was unable to cut away as rapidly as the surrounding schist.
The projecting knob had partially protected the softer material in its lee,

1 A most valuable paper by Chamberlin upon the effect of ice upon rock will be found in the Seventh
Annual Report of the Director of the U. S. Geol. Surv., 1888, page 155.

2 See paper by Gilbert read before the Cordilleran Section of the Geological Society of America at its
1905 winter meeting. " Crescentic Gouges on Glaciated Surfaces," Bulletin Geol. Soc. of Amer., vol. 17,
PP- 303-3 U.

96 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

forming an elongated tail, or trail, extending from the knob in the direction
of ice motion. In some cases a small quartz vein cuts across the surface in such
a way as to protect in its lee a strip of the softer rock. In front of the knobs
there is cut out, as a rule, a frontal groove lying at the base and curving around
laterally into two others, one upon either side, forming the lateral grooves. In
places where the ice acted with greater vigor, owing to the concentration of its
action, or where differences existed in the structure, or hardness, of the rock
there were cut out basins and U-shaped troughs, representing, in miniature, lake
basins and glaciated valleys. One basin, with perfectly smoothed sides and
bottom, had a length of 15 feet, a breadth of 6 feet, and a depth of 6 to 8 inches
below its lower rim. The greatest depth was located one-third of its length
from the upper end, indicating where the gouging action had been greatest. One
of the troughs was 1 1 to 1 2 feet across and 4 to 5 feet in depth.

b. Bear-den moraines. Some 800 to 900 feet below the terminal moraine of
1887, or about 1,400 feet from the nose of the ice in 1904, there occurs a moraine
of the same general type as that described under this head in connection with
the Victoria. This consists of very massive blocks of quartzite, arranged in
a north to south ridge across the valley, having a breadth of about 400 feet and
a height above the general valley floor of 20 to 40 feet. The largest block
observed was measured by Messrs. Moseley and Todd and its dimensions, above
ground, were found to be about 107.5 by 28 by n feet, from which it was esti-
mated to weigh about 2,000 tons. A portion of this ridge is seen in plate xxxvi,
figure i , taken from one of the blocks of the moraine itself, looking toward the glacier
up the valley. The blocks are blackened with lichens, more or less moss-covered,
and carry enough soil to support considerable vegetation of a larger size. A spruce
growing upon the moraine had been cut and with a circumference of 128 centi-
meters gave 243 rings of growth. A hemlock, also upon the moraine, with a
circumference of 320 centimeters (50 centimeters from the base), was calculated
to be 447 years of age. This estimate was based upon the average breadth of the
annual rings of growth measured in the Illecillewaet and adjoining Asulkan
valleys. This average breadth was found to be 1.140 millimeters, as compared
with 0.884 millimeter in the Lake Louise Valley.

From the outer edge of this moraine, 1,500 feet down the valley measured
along the stream, there begins another similar but larger moraine of the same
type. Starting from the spur of Glacier Crest which separates the Illecillewaet
and Asulkan valleys, the ridge swings out across the valley bearing N. 8° W.,
and then swings around to N. 15° W. It is 200 to 300 feet across and some 50 to
60 feet above the valley floor, somewhat steeper toward the glacier. The blocks
are very coarse quartzites and schists, blackened with lichens, and presenting
angular outlines. The largest block noted was estimated to weigh 1,250 tons.
The usual filling of a moraine, gravel, sand, and clay, is practically absent. Upon
the eastern side, for a portion of its length, it is covered by a mass of broken
tree trunks which were swept from the side of Mt. Eagle by an avalanche (plate
xxxvn, figure i) some decades ago. Enough soil has accumulated about the rocks

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHF.RZER.

PLATE XXXVII.

FIG. I. — "Bear-den moraine" made conjointly by the lllecille>vaet and Asulkan Glaciers.
Strewn with timber avalanched from the right-hand mountain slope.

'

• :

FIG. 2. — Illecillewaet Glacier in 1897. Photographed by Notman & Son, Montreal. Compare with plate xxxvi.

*J OF THi

UNIVERSITY

OF

X
X

w

H
<

J
0,

c
n

OS

w

T

w
o

Q
-

O

O
in

•z

O
H
-

s

EH
Z
O
CJ

X
H

ui
u-

if
-j

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. Q7

to support a growth of raspberries, blueberries, etc., and also a few spruce 8 to
1 2 inches in diameter. The bulk of the material lies to the west of the glacial
brook and was derived from the eastern side of Glacier Crest and Mt. Lookout,
the cliffs of which have a northwest-southeast trend. The shape of the moraine
and the way in which the blocks have been deposited indicate, as noted by Prof.
Penck at the time of his visit, that the moraine was built conjointly by the former
Illecillewaet and Asulkan glaciers. The blocks contributed by the Asulkan
came from the western side of Glacier Crest and the Asulkan Ridge, and are
much less in amount than those derived from the eastern side and transported
by the Illecillewaet. The largest tree found growing inside of this moraine
was calculated to have been 520 years old when it died and from the condition
of its wood and bark to have been dead about 30 years.

CHAPTER VII.
ASULKAN GLACIER.

i. GENERAL CHARACTERISTICS.

LYING at the head of the Asulkan Valley, upon the opposite side of Glacier
Crest from the Illecillewaet Glacier (see plates xxxn and xxxm), is located the
Asulkan Glacier. Its broad expanse of snowfield extends in a semicircle from
Asulkan Ridge, past Leda, Pollux, and Castor to the northern extremity of the
Dome, faces to the northward, and under the sunlight is of dazzling whiteness.
The name is of Cree Indian origin and is generally said to mean "goat," but I am
assured that it really means "bridge." The nose of the glacier lies about three
miles from the station, reached by a picturesque and easy trail, except in the
upper part, where the trail becomes steep. The glacier itself may be safely visited
and studied without a guide, but no one should venture upon the neve" unattended,
as it is very treacherously crevassed. This glacier is the smallest and the most
southern and western of the series here reported upon, its nose lying in longitude
117° 28', west and latitude 51° 13', north.

The glacier consists of three streams, two of which are closely united and the
third separated from the other two except in the neVe" region where they are
all united. The length of this third stream, measured from the Asulkan Pass, is
about two miles, of which the first mile is ne"ve" and the lower mile is ordinarily
free from snow during the summer season (plate xxxix, figure i). The breadth
of the dissipator is about 1,800 feet in the upper part, but about the middle of its
course it makes an abrupt bend from the north to the northeast and tapers
gradually to a sharp nose. The eastern margin curves around gradually to the
nose, while the western side is curiously straight, cutting diagonally across
what appears to be the natural course of the glacier. There seems no apparent
reason for this abrupt bend in the glacier and for the remarkably straight western
margin of the ice, but the explanation will appear in what follows. This peculiar
contouring of this stream gives it the general form of a bear's paw — a polar bear

98 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

—in which the straight margin represents the sole. From near the heel of this
foot there extends southward a long, slender ridge of glaciated rock, carrying
more or less morainic matter, which separates this eastern ice stream from the
double ice mass immediately to the west (plate xxxix, figure i). Judging from
the line of crevasses and faulting across the ne"ve", there lies another similar ridge,
parallel with the first and about one-quarter mile to one-half mile to the west,
which separates this mass into two streams, each having its own separate nose, as
shown upon the map. This ridge is apparently the continuation of the line of
bedrock exposed along the right-hand margin of the westernmost ice stream.
The neVe1 line upon this glacier is about 7,000 feet above sea-level and the main
portion of the neVe" lies between this altitude and 8,000 feet. From the Asulkan
Pass (7,710 feet) to the nose of the easternmost stream the descent is 2,110 feet,
or at the rate of 1,055 ^eet to tne m^e- The altitude of the nose is 5,600 feet,
or some 800 feet higher than that of the Illecillewaet, due apparently to the smaller
volume of ice in the Asulkan and its dissipation at three separate points. The
altitudes of the two higher noses to the west are about 6,ooc feet, or the same as
the Victoria. So far as may be judged from the crevasses and fault ings, the
ice responds fully to the irregularities in its bed which indicates that it is
relatively thin. The surface slope of the western and middle streams is very
steep; that of the easternmost, or main, stream is much more gentle, amounting
in places to not more than 6°. Toward the nose the inclination becomes 25°
and then drops off to but a few degrees, so that it may be readily ascended.
Upon either side of the stream the marginal slopes are steep for a few hundred
feet back from the nose.

2. PIEDMONT CHARACTERISTICS.

If the reader has covered Chapter IV of this report he will have recognized
already the piedmont character of the Asulkan, which consists of three com-
mensal streams. The glacier is of peculiar interest because it is an illustration
of a piedmont glacier in its senile condition. It has reached its second childhood
and now illustrates the disintegration of a piedmont glacier into the component
streams, the union of which in its youth gave rise to the glacier itself. Every
glacier of this type begins with the independent development of a system of
Alpine glaciers, coordinate in importance, which coalesce laterally into a single
ice mass. The length of the glacier is determined by the length of the separate
streams composing it and its breadth by the number of streams and their com-
bined breadth. In the final stages of dissolution, which must come sooner or
later in its life history, the piedmont glacier shrivels back into the original Alpine
components. The eastern tributary has already separated sufficiently so that
it may be regarded as an independent glacier. The other two have separated
for a distance of about one-fourth of a mile, but the separation will not be com-
plete until the ridge of rock above noted has appeared at the surface of the ice.
The middle stream covered the ridge of rock, now exposed between it and the
eastern stream, and sent its nose down the valley as far as the drainage brook

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZF.R.
Mt. Donkin iintl Asulkan Pass.

PLATE XXXIX.
Pollux. Castor.

FIG. i. — General view of Asulkan Glacier in 1902. Copyrighted, 1902, by the Detroit ['holographic Co.
Donkin. Castor and Pollux. Dome. Bonney.

FIG. 2. — The Asulkan glaciers and snowfields from Avalanche Mt. (elevation 9,387

feet), showing a decadent piedmont glacier. Photographed in

1901 by Arthur O. Wheeler.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 99

shown upon the map. There it formed a series of terminal moraines upon its
eastern side, the eastern component standing at about the same level and forming a
similar series. The sudden bend noted in the eastern component, one-half mile
back from the nose, resulted from its pressure against the side of the middle
stream which it was unable to force aside. Conjointly they formed a straight
medial moraine from the bend to the nose. Upon the more rapid retreat of the
middle stream and its disappearance from this part of its bed, this moraine
became the left lateral of the easternmost stream (plate xxxix, figure i), and was
of such a massive character that it has continued to deflect the ice from its
natural course.

In plate xxxix, figure i, we have shown nearly the entire eastern and middle
streams of the Asulkan, and a portion of the ne"ve" of the western. A distant view
of the entire glacier is given in plate xxxix, figure 2, taken by Wheeler from the
summit of Avalanche Peak (9,387 feet) in 1 90 1 . The Dome may be recognized from
its contour and from it there is seen to be a broad ridge extending valleyward and
marking the western limit of the present Asulkan Glacier. To the right of this
ridge, along the eastern slopes of Mts. Afton (8,423 feet) and Abbott (7,710 feet),
four marked depressions occur, each containing small-sized glaciers. The
contour of the rocky slopes separating these amphitheaters, or cirques, as they
are termed, proves that at an earlier stage of glaciation these streams coalesced
laterally and united with the present Asulkan, forming a grand, hanging, piedmont
glacier, extending from the Asulkan Ridge to Mt. Abbott with at least nine or
ten main commensals. Previous to this stage they had united with others from
the head and opposite side of the valley into a grand Alpine glacier, which became
a tributary of the ancient Illecillewaet trunk glacier in Pleistocene time.

3. NOURISHMENT.

The neve" field of the present Asulkan is arranged in the form of a semicircular
belt, extending from the Asulkan Ridge upon the east around to the Dome upon
the west, having a length of about three and a half miles and an average breadth
of perhaps three-fourths of a mile. The area of this field is somewhere between
two and a half and three square miles, or less than half that of the Illecillewaet neVe"
field. The amount of precipitation over this field cannot be essentially differ-
ent from that given for the neighboring glacier (page 82). From an elevation
of about 7,000 feet the neVe" snows reach up to the crests of the bounding ridges,
in many places, attaining an elevation of 9,000 feet. It is possible to pick out, in a
general way, the neVe fields by which the separate ice streams are nourished.
The eastern stream receives its supply from Asulkan Ridge (9,100 feet) and from
the Pass (7,710 feet), the former moving westward down the oblique slope and
delivering its supply of ice and subglacial debris to the right side of the mainstream
flowing from the Pass. A still less amount is received from the opposite side
from the snow that accumulates upon the northern slope of the unnamed peak
(8,700 feet +) lying to the east of Leda. The Pass and upper neVe" are reached
by means of the right lateral moraine. Judging from the course of the transverse

100 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

crevasses, which lie at right angles to the main direction of flow, the neVe" that
accumulates between this minor peak and Leda (9,133 feet) moves northward
and nourishes the middle ice stream. The oblique course is taken probably
because of the continuation southward of the ridge of rock previously noted
as separating the middle and eastern streams. The presence of a similar ridge
beneath the ice deflects the ne"ve" snow and ice from the northern slopes of Castor
(9,108 feet) and Pollux (9,176 feet), to which is added that from the Dome
(9,029 feet), and thus is obtained the supply for the double-nosed western ice
stream. The middle stream is least well supplied at the present time with ice
from the neVe" field and has receded farthest. It seems very probable that
when the ice was thicker over the neV£ there was relatively less of it deflected
to the western stream by the subglacial ridge and this fact permitted the middle
stream to maintain the same length as the now better nourished eastern. The
superficial layers from Castor and Pollux could move directly across those which
the configuration of the bed deflected northward, but as the general elevation of
the neVe" was lowered a relatively greater and greater percentage of ice was
deflected to the western stream and the nose of the middle stream retreated
steadily some 2,200 feet up the slope, to an elevation at present 400 feet higher
than the nose of the eastern stream. Were the ice of all three streams now
concentrated into a single one it is probable that the nose would attain as low
an altitude as that of the Illecillewaet.

4. MORAINES.

Owing to the absence of high, overtowering cliffs, such as we find in the
case of the Wenkchemna and Victoria glaciers, the neVe" fields of the Asulkan
receive very little rock de'bris over their surfaces. In consequence, the ice itself,
except along the margins, is quite free from rock fragments. As in the case
of the neighboring Illecillewaet collecting basin, conditions are favorable for
receiving wind-blown dust from peaks and ridges towering above the snow.
This dust is distributed somewhat evenly over the snow and, when concentrated
by melting, gives rise to the stratification and imparts a soiled appearance to the
ice about the lower margins.

The right lateral moraine of the eastern ice stream makes its appearance just
east of the nose and south of the stream from Asulkan Ridge. It rises at once into
a conspicuous, sharp-crested ridge, extending south-southwestward, and bending
abruptly to the south-southeast, attaining the length of a mile before it dips
under the ne"vd snow. The lower portion of the moraine seems entirely free
from ice, the outer slope carrying fir and spruce 50 to 60 years of age. The inner
slope is more steep and has younger vegetation, indicating that the ice has within
a few decades withdrawn from the moraine. The crest rises to a height of 60 to
70 feet above the valley floor upon which the glacier rests. The rocks consist
largely of bruised and rounded quartzites and schists, which in the upper part
are embedded in a matrix of glacial clay. This material appears to come
from beneath the glacier that covers the western slope of Asulkan Ridge,

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XL.

FIG. i. — Left Asulkan moraine shedding its rocky covering and exposing the ice core.

FIG. 2.— Debris-covered nose of Asulkan Glacier, August, 1904. Glacier had been advancing for some three or four

years.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. IOI

by which it is delivered to the main stream from the Pass, as previously noted,
upon a level with its surface. This ne've'-covered glacier sustains the same relation
to the Asulkan that the Collie and Gordon glaciers do to the Yoho. Between
the crest of the moraine and the glacier there intervenes a steep boulder slope,
about 300 feet broad in the lower portion near the nose, but narrowing gradually
for a half-mile, when the moraine and ice meet. Opposite the nose, upon the
eastern side there is an outcrop of a silvery schist, with its strata upon edge,
which has been glaciated and plucked.

The development of the left lateral of the eastern stream from a former medial
has already been described and its very straight course obliquely across the valley
been accounted for. Originally, this moraine may have been largely subglacial,
or englacial, the material being derived from the basal layers of the ice. The
slope of the valley floor is northward, while this lower half-mile of the moraine
bears northeastward. After making the very abrupt bend noted, the moraine
continues for a quarter-mile farther, resting upon the rocky ridge of quartzite
and a greenish schist. This ridge raises the base of the middle stream above
the present surface of this portion of the eastern stream. The moraine consists
very largely of ground moraine, supplied apparently in large part by the middle
ice stream, but instead of clay the filling is a glacial sand. The finer material
may have been removed by currents too gentle to transport this sand. The
inner slope of the moraine is steep, the outer is more gentle down to the drainage
brook from the middle nose. The crest of the moraine rises 125 to 150 feet above
the floor of this valley. About one-quarter mile back from the nose this moraine
begins to shed its cover of rock debris, revealing in a most interesting manner
the real structure of such a moraine. From this point up the valley the moraine
is a typical, sharp-crested structure (plate xxxix, figure i), but here the debris has
begun to slip to either side, forming a double ridge with a continuous ice crest
between. Plate XL, figure i, gives a view of this exposed ice core, looking up the
glacier along the inner side. The highest portion of the ice ridge attains a height
of 25 to 30 feet, which is being rapidly acted upon by the sun in midsummer.
Where it has been longest exposed the ice has melted below the general level of the
glacier, forming a depression with a ridge of rock d6bris upon either side, the
outermost one being quite prominent. Into the depression the material from
either side has begun to roll and slide, thus protecting the ice at the bottom of the
depression from the sun. Had the thickness of the ice proved sufficient, in
time the rock debris would have gotten back, in large part, into the depression,
allowing the ice to melt upon either side and starting again the formation of a
single-crested, typical moraine. Thus it appears that moraines may, under certain
circumstances, pass through the same series of stages as those described for sur-
face lakelets upon page 5 7 of this report.

About the eastern nose there has been pushed up a ridge of ground moraine,
from 12 to 15 feet high, into which the ice nose plunges and buries itself. Upon
this ridge minor ridges, but a foot or two in height, also occur, as seen in plate XL,
figure 2 . At times the nose is so deeply buried that it is difficult to find it for pur-

102 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

poses of measurement. When the middle stream stood at this lower level, the two
built a series of three or four latero-terminal moraines which curve gently down
the valley from the lower ends of the laterals. The inner terminal of the middle
stream has the appearance of age, compared with the others lying just east, and
is being covered slowly with low shrubs and evergreens. From this difference in
age one would infer that the ridges of the series, lying to the east, had been built
by the eastern stream alone. The nose of the middle stream, especially upon its
eastern side, rests largely upon bedrock, more or less strewn with rock fragments.
The rock here, as elsewhere about the glacier, consists of quartzite and schist,
plucked and glaciated. Along its left side, back as far as it has become sepa-
rated from its neighbor, it has -built a sharp-crested, lateral moraine, which
in the lower half is double, curving gently down the bouldery slope. The inner
slope is steep, the outer long and more gentle. The boulders are rounded and
bruised but only occasionally well glaciated. The double western nose is similar
to the middle, in that it is steep, perched high up on the slope, has bedrock
exposed upon its eastern side, while upon the left it has built a short, sharp
lateral extending up to the ne've'. In front of the western and middle streams
there has been uncovered a steep slope of bouldery ground moraine so recently
that trees have not been able yet to get anything more than a start.

5. CREVASSES.

Owing to the irregularities in its bed, the steep slopes, and the apparent thin-
ness of the ice, the Asulkan streams are badly crevassed and faulted. The n6v6
fields of the western and middle streams cannot be traversed with any degree
of safety, while that of the eastern calls for the greatest skill in locating the
snow-covered death-traps (plate XLI). The crevasses in the ne"ve" portions are
mainly of the transverse type, caused by the rapid movement of the ice over
an irregular, steep slope. They stand approximately at right angles to the
direction of motion and furnish a clue as to the general movement of various
portions of the ne>6 field. Just below the neVe" line the eastern stream en-
counters, in its central portion, an obstruction by which the ice is shattered in every
direction, but mainly transversely (plate xxxix, figure i). The descent is not
rapid enough to constitute a cascade and the blocks, at first angular, become
sharpened by melting into seracs but retain their vertical position until melted
away at the base of the slope. The development of these seracs is well shown in
plate XLII, figure i. The ice exposed portion, or dissipator, of this stream shows
numerous marginal crevasses along either side of its course, those upon the
eastern side, in the lower portion, extending beyond the central axis. Those
upon the western side are not so strongly developed. It is upon the middle
stream that the dirt band crevasses occur figured in connection with the discus-
sion of this subject (plate xvii, figure i). The slope, however, is too steep for
their formation and ^reservation.

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZER.

PLATE XLI.

"

Vic,. I. — Stratification of Asulkan Glacier. The dates are added upon the sup-
position that the strata represent seasonal deposition. Copyrighted,
1902, by the Detroit Photographic Co.

FIG. 2. — General view of Asulkan Glacier in 1898. Reproduced through courtesy of the Messrs. Vaux. Compare with plate XXXIX,

figure I.

UNIVERSITY

OF

.

GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS. 103

6. ICE STRUCTURE.-

The ne've'-covered portion of the ice acquires a very perfect stratification as
the result of wind distributed dust and periodic melting over the surface. This
structure is beautifully shown in the crevasse walls and the faces of the numerous
faults in the ice. In the photograph of the Detroit Publishing Co., reproduced
in plate XLI, the successive layers, with the minor stratification seams, are clearly
shown. The correspondence of the strata upon opposite sides of the crevasse
shows that there had been no faulting. From his heel to the crown of his hat
this guide pictured is about six feet in length and, by comparison, we ascertain
that the strata shown range from three to twelve feet in thickness. The picture
was taken during the summer of 1902, and in looking at the uppermost stratum
it is forced upon one's belief that this represents the compacted snow that accum-
ulated over this spot during the season of 1901-2. Part of this snow was pre-
cipitated directly, part of it may have been drifted by wind action. It may have
lost some by wind action, as well, during the season of accumulation. It has
been compacted by melting, pressure, and occasional rain into a fine granular
ice. If we are right in supposing that this stratum represents the accumulation
during the season of 1901-2, minus the loss by the combined agencies, then the
stratum upon which it rests must have accumulated during the season of 1900-1.
Passing down the side of the crevasse we may thus assign dates to the successive
strata, finding that they reach back to the season of 1895-6. It is especially
interesting to note that the deposits supposed to have been laid down between
the summer of 1898 and that of 1902 average considerably thicker than those
between the summers of 1895 and 1898, since this dividing date falls very near
the supposed date of the beginning of the phase of increased precipitation in this
region. It is further to be noted that the stratum marked 1898-9 is the thickest
of the series. It is very unfortunate that our precipitation data are not fuller
for the locality. In order to serve the present purpose in establishing a rela-
tionship between the amount of precipitation and the thickness of the strata in
the ne"ve", a combination should be made of the last three months of the year with
the first nine of the following year. This would unite practically all of the snow-
fall and the rain and melting water of the following summer, as it is combined
in the stratum itself. In a paper upon the Canadian Pacific Railway, from
Laggan to Revelstoke,1 Mr. William Vaux gives the average snowfall for Glacier
House from 1895 to 1898 as 3 1 feet, based upon records kept by the station agent,
this being but 83 per cent, of the normal. From October, 1898, to May, 1899, in-
clusive, the snowfall alone amounted to 43 feet 8^ inches, being 1 7 per cent, above
the normal. The records are lacking up to 1 902 , for which year the Meteorological
Service reports 13.88 inches of rain and 347 inches of snow, or a total equivalent
of 40^ feet in snow, or 14 per cent, below the normal. By referring to plate XLI
it will be seen that the stratum assigned to the year 1898-1899 is the heaviest of
the series while that for 1901-1902 is light. From 1895 to 1898 the strata are

1 Proceedings of the Engineers' Club of Philadelphia, vol. xvii, 1900, p. 73.

104 GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS.

thin, corresponding to the average lighter snowfall above reported for these
years.

A still further confirmation of the conclusion reached above, that from the
middle of the year 1898 there has been a marked increase in the snowfall, is
furnished by the notes and photographs of the Messrs. Vaux, to be noted later.
Their photograph of 1898 shows a large amount of rock exposed along the slopes
of Leda, Pollux, and Castor, as well as between the eastern and middle ice streams.
In 1899 they note that the ne"ve" line is lower and the hanging glaciers are much more
active, giving rise to frequent avalanches, which were very infrequent in 1898.
When their photograph of 1898, reproduced here as plate XLI, figure 2, is com-
pared with that of the Detroit Publishing Co., taken in 1902 (plate xxxix, figure
i ) , the increase in the amount of neve" is striking, the presence of the bergschrunds,
in areas that were bare rock, indicating that glaciers had formed in the meantime
and that there is not simply a covering of loose snow, such as might fall in a
night. In looking over the broad snow expanse one does not think of there being
hanging glaciers upon the slopes of Castor and Pollux, as they seem to blend with
and be an integral part of the general ne've' field. In 1898 they were separated
sufficiently so that it was natural to think of them as being detached. In two
weeks of August, camping in plain sight of the region, in 1904 and 1905 , the writer
does not remember to have seen or heard a single avalanche from this quarter.
They were frequent in the summer of 1899, and, presumably, continued so un-
til the space between the lower neVe" and the upper became so filled in as to
prevent further slides. From all the evidence obtainable it seems most prob-
able that the major stratification planes in the Asulkan ne"ve" represent the breaks
between the successive year's snowfall, and that a phase of deficient precipita-
tion closed in this region about the middle of the year 1898, since which time
the average annual precipitation has been in excess of the normal.

The disturbance of the ice noted upon the eastern stream does not destroy
the stratification, since* it extends only part way across the stream and is not
intense. The strata are, however, more or less tilted and distorted. The lower
stratum is wedge-shaped, having apparently lost from its basal portion by sub-
glacial melting. The blue bands in this stratum are not parallel with its upper
surface, but cut it at angles of about 13° to 14°, being more nearly parallel with
the valley floor. In the stratum just above, the blue bands and stratification
planes were conformable. In general, the blue bands were found to be regularly
developed, quite in contrast with the stratification. The dirt stripes showed
well over the surface and margins of the eastern stream, some excessively thin
ones being observed and previously noted (page 54). About the nose, upon the
walls of some of the longitudinal crevasses, the blue bands were found to dip
back into the glacier at angles of 11° to 28°. About the eastern side they were
found to dip downwards and inwards, nine observations giving an average of
46°, with a range from 36° to 57°. The granules about the nose are small,
compared with those seen in the larger glaciers, and will average less than a
half-inch in diameter.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 105

7. DRAINAGE.

Owing to the crevassed condition of the ice the surface streams are small,
dropping into the glacier, or to the bottom, before they can develop any size.
Over the neVe" area the water resulting from melting, or from rains, is at once
absorbed. Over the ice exposed portions, during hours of melting, small rills
and surface brooks come into existence, carrying water with a temperature of
32°. No lakelets were noted upon the glaciers, or about the margins, but upon
the col, lying between Castor and the Dome, Mr. Wheeler found a lakelet of
sapphire blue water. Under ordinary conditions there is practically no marginal
drainage. In 1904, back some 800 feet from the nose of the eastern ice stream,
a small flow was visible for a short distance. From each of the three noses there
issue two to three drainage brooks, those from the eastern uniting with one
another and with the drainage from Asulkan Ridge, after which is received the
central flow from the middle portion of the glacier. That from the western com-
mensal, along with the drainage from the hanging glaciers lying farther to the west,
cascades into the Asulkan Valley, forming the "Seven Waterfalls." The flow
from the eastern nose is the strongest and carries the most sediment, considerably
more than the Illecillewaet. It fluctuates in volume during the day, reaching its
maximum in the late afternoon, or evening, and being lowest in the early morning.
The combined drainage from the middle and eastern portions of the glacier,
along with that received from the Asulkan Ridge to the eastward, has cut a
gorge 30 to 40 feet deep through the soft schist. This has the appearance of
having been done since the withdrawal of the ice, but it may have been started
by a subglacial stream. Under high velocity and charged with sharp, glacial
sediment the cutting power of water must be rapid upon a soft schist. Its action
upon quartzite boulders is well seen in the bed of the brook from Asulkan
Ridge.

During the last week in August in 1904 and 1905, the average of 28 observa-
tions upon the temperature of the water from the eastern nose was 32.42° F., the
range being from 32.0° to 33.0°. Two observations upon the water from the
middle nose, upon leaving the ice, averaged 33.0°, while from the third nose it was
32.8°. Before receiving the middle drainage the temperature of the brook was«j6.9°
and after the two had united just above the schist cut the temperature was 37.8°.
Passing down the valley some two miles, and receiving drainage from either slope,
the temperature at the bridge across the Asulkan Creek averaged 42.6°. The
water is here turbid but assuming more or less of a greenish cast. The stream
from the Asulkan Ridge before receiving the flow from the glacier was found to
average 36.5° (20 observations). These observations seemed to indicate that
the maximum temperature was attained between 11:00 A.M., and 1:00 P.M.,
and that as the volume of water increased as the day advanced the temperature
gradually fell. This is brought out in the table here given.

io6

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

TEMPERATURES OF STREAM FROM ASULKAN RIDGE.

1904.

Aug. 30.

7:10 A.M.

35-2'

Aug. 26.

7:15 A.M.

35-6'

Aug. 27.

9:10 A.M.

36.9'

Aug. 31.

10:15 A.M.

36.3'

Aug. 27.

1:10 P.M.

37.8'

Aug. 25.

6:05 P.M.

36.5'

Aug. 28.

• 6:15 P.M.

35. 6'

Aug. 25.

6:50 P.M.

35.*'

1905.

Aug. 28.

6:00 A.M.

35-8°

Aug. 28.

7:00 A.M.

36.1°

Aug. 29.

9:00 A.M.

37-o°

Aug. 27.

II 100 A.M.

38.70

Aug. 27.

2 :oo P.M.

37-9°

Aug. 27.

3:10 P.M.

37-4°

Aug. 27.

4:00 P.M.

37-o°

Aug. 27.

5 :oo P.M.

36.7°

Aug. 27.

6:00 P.M.

36-3°

Aug. 27.

7:00 P.M.

36.0°

Aug. 27.

8:00 P.M.

36.0°

8. FRONTAL CHANGES.

Points of reference for the study of the frontal behavior of the lower Asulkan
nose were established August 12, 1899, by the Messrs. Vaux and observations
and photographs have been repeatedly made by them since. At that time they
made an unsuccessful search for reference blocks previously marked by H. W.
Topham. One year previously (August 23, 1898) they had visited the glacier
and obtained a photograph from their "test rock," which was published, along
with a brief description of the glacier, in the paper previously referred to.1
A comparison of their test picture of 1898 (plate XLI, figure 2) with that of
1899 showed a slight shrinkage in the height and a slight increase in the breadth,
"while the position of the tongue had not changed to an appreciable extent."
The ice fall appeared to be less and they note that the ne"v£ line was lower, the
glaciers upon the slopes of Castor and Pollux more active, giving rise to a
number of avalanches, which seemed very infrequent in 1898. In marking the
position of the tongue at the time of their visit in 1899 three rocks were selected
in a line with the nose, the magnetic bearing of which was N. 85°35' E. One
rock was located upon the small, left lateral moraine, a second just below and to
the right of the nose, while the third lay upon the inner side of the higher
right lateral. In 1900 the Messrs. Vaux observed a retreat of 24 feet and " a
marked shrinkage in every dimension." From 1900 to 1901 these observers re-
ported an advance of 4 feet and for the two years 1901 to 1903 an additional
advance of 36 feet.2

This glacier was first visited by the writer September 17, 1903, at which time
it was found that the nose of the glacier lay 13^ feet beyond the Vaux line,
which was readily located by the two well marked end rocks. This would indicate
that the nose had retreated z\ feet between the date of Vaux's measurement in
1903 and September 17 of the same year. The stone that had been marked
near the nose had been pushed forward some 14 to 15 feet, turned on end and was
about to topple over. Upon August 27, 1904, the nose lay 12^ feet beyond the
line, indicating practically no change, when allowance is made for difference in

""Some Observations on the Illecellewaet and Asulkan Glaciers of British Columbia," Proceedings oj
the Acad. of Sci. of Phil, 1899, p. xai.

a " Variations of Glaciers," H. F. Reid, Jour, of GeoL, vol. xni, 1905, p. 316.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

dates of observation. Exactly one year later (August 27, 1905) the nose had
made a retreat of 34 feet from the position held in 1904, standing now 21$ feet
back from the reference line established in 1899. The nose consisted at this time
of a thin slab of ice, sloping to the west and coated with fine debris. A relatively
small amount of melting would cause a further recession of 30 to 35 feet. The
ice in the left lateral moraine was found to extend four feet beyond the reference
line and 25 £ feet beyond the nose. Owing to the rock cover it could not be
ascertained how much farther the morainic ice core extended. The movements
of this nose may be summarized as follows:

CHANGES IN THE NOSE OF THE ASULKAN GLACIER.

(Eastern Ice Stream.)

1898-1899. "Practically no change."

1899-1900. Recession of 24 feet.

1900—1901. Advance of 4 feet.

1901-1903. Average advance of 18 feet.

1903—1904. Retreat of i foot.

1904-1905. Retreat of 34 feet.

1905-1906. No change.

9. FORMER ACTIVITY.

a. Development and decadence. At a much earlier stage, presumably in
Pleistocene time, the combined snows of the Asulkan Valley united into a great
Alpine glacier, the ancient Asulkan, which was a tributary of the ancient Ille-
cillewaet, and this, in turn, a tributary of the great trunk glacier that flowed
southward in the Columbia Valley, to the west of the Selkirks. With the diminu-
tion of snowfall, and possibly also an amelioration of the climate, the glaciers
disappeared from the main valleys and withdrew into the tributary valleys
and alongside the steep, higher slopes. An Alpine glacier occupied the Asulkan
Valley from the Pass to where the valley joins the Illecillewaet, some four
miles in length, which was in part nourished by a hanging, piedmont glacier
extending from the Pass to Mt. Abbott. This glacier sustained, during this
stage, the same relation to the Asulkan lying in the valley, that the hanging
Victoria sustains to the lower ice stream. The effect of these great ice masses
upon the valley floor and sides was similar to that already discussed for the
Victoria and Yoho glaciers (pages 61 and 80).

b. Bear-den moraines. Just before the complete and final separation of the
Asulkan from the Illecillewaet, the Asulkan became loaded with very coarse,
angular rock fragments , and only a minimum of fine material. This was at the same
time that the Illecillewaet was similarly laden and, conjointly, they deposited
the massive bear-den moraine described upon page 96. The most of its
material was deposited upon its right, showing that it must have been received
from the western side of Glacier Crest and Mt. Lookout. The amount carried
was notably less than that brought down by its neighbor lying to the east of the
Crest. The ridges about the head of the valley and along the western side were
largely under snow and ice and could supply no such debris. After this load

108 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

of rock had been deposited the Asulkan began to retreat, withdrawing a distance
of 3,500 feet up the Asulkan Valley. The front now halted and there was built
a moraine of the ordinary type across the valley, consisting of fine and coarse
material, intermingled with but few coarse blocks. From this we conclude that
the glacier was, at this time, carrying the ordinary kind of load. The retreat was
resumed and in the meantime the glacier became a second time laden with coarse
fragments of the adjoining cliffs. At a distance of about 1,000 feet from the
previously formed moraine these blocks began to be deposited and were dropped
over a distance of some 500 to 600 feet, not so concentrated or imposing as the
outer bear-den moraine. For the next 1,500 feet these blocks were scattered
along the valley, implying that the supply over the surface had not been suf-
cient to bring about a halt. The retreat continued towards the head of the
valley and at a distance of 2,000 feet farther a halt occurred and a moraine of the
ordinary type was again built, with the usual quota of fine and coarse material.
From the time then that the Asulkan was about to separate from the Illecillewaet
it became twice loaded with coarse, angular fragments of quartzite, building a
moraine of the bear-den type. In the interval it carried material of the ordinary
kind found upon and within the ice and built a moraine of the ordinary type.
Subsequently to the formation of the second, straggling, bear-den moraine, it
has been carrying and depositing the usual class of material. It differs from
the Victoria in that the ancient moraine of the ordinary type was deposited
between the two bear-den moraines instead of outside the two, as in the case of
the latter.

c. Rate of retreat. The only possible data for any estimates upon the rate
of retreat up the valley must be drawn from a study of the forest trees and no
one realizes any more strongly than the writer how unreliable and misleading
such data may be. However, we may obtain an approximate minimum estimate
by this means, which may have some interest, if not real value. Some excep-
tionally large spruces and hemlocks are found near the mouth of the valley and
within the outer bear-den moraine. Based upon the average thickness of the
rings of growth, noted upon page 96, two of the largest seen should be 525 and 598
years of age, respectively. Toward the schist cut, at the head of the valley, the
rings of growth become coarser and the trees smaller, the difference in elevation
amounting to about 900 feet. One of the largest firs showed 161 rings of growth
and a still larger hemlock growing near was estimated to have lived about 250
years. Assuming that it required about the same length of time for the trees to
get started at either end of the valley, it took the Asulkan about 350 years to
retreat the two miles from the mouth of the valley to the schist cut, or at the
average rate of about 30 feet a year. From the schist cut to the present nose,
about one-quarter mile, the valley opens and the retreat must have been much
slower, owing to the volume of ice to be melted away. If we assume that it
required 50 years for the hemlock noted to get started, the minimum time in-
volved would be 300 years and the maximum average rate of retreat for this part
of the glacier would have been about 4.4 feet per annum. If the cut in the schist

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE— SHERZKR.

PLATE XLII.

IMG. [. — Development of ice seracs from glacial blocks, Asulkan Glacier, August, 1904.

FIG. 2. — Disrupted quartzite blocks, near head of Paradise Valley, Canadian Rockies. Illus-
trating plucking power of glaciers.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. IOQ

has been made entirely since the withdrawal of the ice from that part of the
valley, then the rate of cutting would be over an inch a year, which is probably
too fast for even water at high velocity, charged with glacial sediment and oper-
ating upon rather soft rock. This seems especially true when we consider that
the supply of water is much reduced, or possibly entirely shut off during the
greater part of the year. It is very probable, however, that the narrow gorge may
have been largely formed subglacially, while the glacier extended far down the
valley. Schist layers upon edge do not well record ice action, but even if such
evidence of glaciation was present it may have been destroyed by subsequent
weathering. It must be noted that the time of retreat determined as above
would represent only a minimum value and the rate of movement per annum
for a definite distance would represent a maximum.

CHAPTER VIII.
SUMMARY AND CONCLUSIONS.

IN the closing chapter it is desired to give a concise statement of the most
important results secured in the two seasons' work and the conclusions reached.
The writer desires further to express for the benefit of those who may be interested
his conviction concerning some of the theoretic questions that have arisen in
connection with the study of these Canadian glaciers.

i. PHYSIOGRAPHIC CHANGES IN THE REGION.

a. Mesozoic peneplain. From the close of the Archaean to the end of the
Laramie, conditions were very favorable for the accumulation of sedimentary
deposits in the region now covered by the Canadian Rockies and Selkirks.
Strata belonging to the Palasozoic and Mesozoic eras of the world's history reached
the extraordinary thickness, according to the work of Dawson and McConnell,
of 50,000 to 60,000 feet. Much of this was brought above sea-level during the
Mesozoic era and further sedimentation ceased except in certain restricted
regions in the eastern part of the area,where conditions were still favorable for
marine or fresh- water accumulations. During countless ages of exposure
to the manifold atmospheric agencies there was developed a broad Mesozoic
peneplain, extending in a direction to the west of north and sloping east-
ward and westward, determining the general direction of flow of the drainage
streams. It was during this stage, probably, that the mountains suffered
their greatest denudation, rather than since. The great Laramide revolu-
tion of the western United States and Canada completed the formation of
these mountains, the pressure coming from the west in the region under con-
sideration, and producing a series of parallel folds and troughs, with numerous
overthrust faults, all having a north-northwest to south-southeast trend. The
upheaval was slow enough so that many of the original drainage streams were
able to maintain their general direction of flow, cutting their way continuously

110 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

across the gradually rising ridges of the mountains and the lesser folds of the
foot-hills. In many cases, the troughs, lying between parallel ridges, or the
gaping crevasses in the rock strata parallel with them, captured the drainage
and a system of longitudinal river courses was developed, much younger than
the transverse system. As a result of these great erogenic movements, combined
with the atmospheric and aqueous agencies operating since, we have an uplifted
and dissected peneplain.

b. Pre-pleistocene erosion. With the completion of the mountains at the
close of the Mesozoic,the more or less sluggish streams of the ancient peneplain ac-
quired velocity and renewed their activity, cutting deeply into their former beds.
The newly born longitudinal streams incised still further the channels provided
for them and there were developed two systems of V-shaped valleys more or less
intimately connected. The agencies of weathering broadened the valleys
above and delivered the rock fragments to the stream below, by which tools
the water still further deepened its beds. This action went along slowly from
the beginning of Cenozoic time to the beginning of the Pleistocene, during which
time the roughly angular blocks were carved into jagged peaks and many
of the divides into sharp-crested ridges. The outline of the old peneplain is to
be recognized only when one ascends until his eye is on a level with its uplifted
surface, when peaks and ridges all blend into the common level that cuts the
sky at the limit of vision. See plates n, xvi, and xxxn.

c. Pleistocene erosion. The opening of the Pleistocene and the advent of
the glaciers introduced a new geological agent into the region. A reduction
in the mean annual temperature, combined with an increase in precipitation,
allowed the snow to accumulate about the higher peaks and ridges more rapidly
than it could melt away during the warmer season. Year after year the snow
banks thickened, sent their avalanches into the valleys faster than they could melt
away, and thus the mountains became enveloped in snow and ice. Glaciers
moved down from the more elevated valleys, joined forces with their neighbors,
grew in volume and power, took possession of the river valleys, and sent massive
tongues of ice far beyond the limits of the mountains. The valleys were filled
to depths of 4,000 feet from their floors, in certain cases, the actual elevation
rarely falling below 7,000 feet above sea-level. These ice streams exercised a
powerful effect upon the rock strata over which they passed; in general, rounding
and smoothing their outlines, cutting down prominences, and truncating moun-
tain spurs. In some cases where plucking was most active the rocks were made
still more jagged and irregular than the ice had found them. The lower half
of the valleys, which had been invaded by the ice, had their floors broadened
and their sides correspondingly steepened, giving this portion of the valley a U-
shaped cross-section. The upper portion, under less pressure of ice, still retains
more or less of its pre- pleistocene V-shaped form, the sides being simply smoothed
and fluted. The extension of these V-slopes until they intersect in the valley
may be assumed to mark the level, approximately, from which the glaciers began
deepening their beds. In the floors of the valleys at certain places rock-basins

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 1 1 1

were gouged out, either because of the structure or softness of the rock or be-
cause of the more vigorous ice action for a limited distance. At the heads of
the separate valleys, broad semicircular amphitheaters, or cirques, were cut out,
an interesting series of which is shown in plate xxxix, figure 2. at the right side
of the view.

From observations made upon the plucking power of glaciers in the various
valleys, the writer is quite prepared to admit the sufficiency of glaciers as engines
of erosion, especially where the ice is very thick, concentrated in its action, and op-
erates for long time over stratified, or much jointed formations. In addition
to the plucked mountain peaks observed in the Yoho Valley (page 79), there
is to be seen in Paradise Valley, lying between Lake Louise and Moraine
Lake, a very striking case of plucking, in which very heavily bedded quartzite
has been bodily removed. The upper stratum is 10 to 12 feet thick, and about
the margin of the stratum, the upper surface of which is very perfectly glaciated,
immense blocks, some of them as large as small houses, have been started a short
distance and then left. Apparently in the waning stages of the glacier the
ice had been unable to get hold of blocks which it had been able to pry loose
from the parent bed. This occurs near the head of the valley and it is difficult
to resist the conclusion that hundreds of feet of similar, or less resistant rock
may have been removed between this ledge and the mouth of the valley. Glaciers
with their basal layers shod with hard rock debris would be able to erode slowly.
The amount of erosion accomplished in this way would depend simply upon the
time, but it has probably always been small, when compared with that due to
plucking. It seems likely that pure ice can have only an insignificant effect
upon ordinary rock, when simply pushed across it. The greater the pressure the
more the melting, and unless disruption of the rocks occurs, the only effect
would be to give the rock a polish.

d. Pleistocene deposition. During the maximum stages of glaciation there
were so few overtowering cliffs above the neVe fields and the ice streams them-
selves that very little supra- and englacial material was carried. In consequence
during the stages of halt, that must have succeeded one another in the retreat
from the outermost position attained by the trunk streams, no great, conspicuous
moraines were formed. Not until the glaciers had retreated to near the heads
of the valleys do we find prominent terminal moraines. At this stage the level
of the ice and snow has dropped below the cliffs so that it is possible for the
glaciers to acquire, in many cases, a heavy load of rock debris upon their upper
surfaces. Glaciers like the Yoho have still been unable to build prominent
moraines from materials carried supraglacially. The detritus resulting from
the destruction of rock strata in the valleys and over the rocky slopes was
carried near the bases of the glaciers, or pushed and rolled along between the
ice and its bed. This resulted in the making of much ground moraine, much
of which remained in the valleys in places favorable for its lodgment. During
the maximum stages of glacial development much of ~this subglacial material was
carried beyond the mountains and deposited as till, or it found its way into the

112 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

rapid streams, where it was immediately assorted into boulders, cobbles, gravel,
sand, and clay. In these various forms it was built into the terraces, flood plains,
deltas, etc., which characterize the drainage streams. The finer materials made
their way to the Pacific and Hudson Bay. The argument against great glacial
erosion that the material removed cannot be found, seems to the writer to carry
little weight. If one looks far enough and is able to distinguish the Pleistocene
and post-pleistocene deposits from the earlier, it seems probable that enough
would be located to restore the mountains and valleys to the condition in which
the glaciers found them. In the Bow and Cascade valleys, near Banff, Wilcox
discovered two distinct till sheets, indicating that there were, at least, two main
advances of ice through this section of the mountains. Eastward from the moun-
tains McConnell and Dawson found three such sheets, derived either in whole
or in part from the Rockies.

2. PRECIPITATION.

a. Geographic distribution. Owing to the north to south trend of the four
mountain systems that here constitute the great Cordillera, their limited
breadth, their nearness to the warm waters of the Pacific, and the relation of the
region to the great cyclonic areas that enter from the Pacific, conditions are
favorable for an abundant precipitation upon the western slopes of the moun-
tain systems. The arrangement of the four systems being such that they
increase in height successively from the Coast Range to the Rockies, enables
all of them to .get a fair share of the available precipitation. The prevailing
winds are from the west and laden with moisture. In ascending the windward
slopes much of this moisture is precipitated as rain, or snow, owing to the expan-
sion and consequent cooling of the air. In the condensation of this moisture
its latent heat is liberated and raises the temperature of the air. In being
drawn down the leeward slope by the general cyclonic movement of the atmos-
phere, the air is still further warmed by the compression to which it is subjected,
its capacity for holding moisture is increased, and it reaches the same elevation
upon the leeward slope much dryer and warmer than it was at the corresponding
level upon the windward slope. This gives rise to the well-known "chinook
wind," the equivalent of the Alpine foehn. The Selkirks, lying to the west of the
Rockies, receive the heaviest precipitation, are more completely forested, expe-
rience more frequent avalanches of snow, and send their nave's and glaciers
to lower levels. The shifting of the centers of the cyclonic areas to the south
of this region would give rise to prevailing easterly winds, which in the winter
would be colder and dryer and in the summer warmer than those which now
prevail, and, without doubt, bring about the disappearance of glaciers from this
part of the mountains.

b. Climatic cycles. Precipitation records are too scanty and fragmentary
for safe generalization concerning the occurrence in this region of oscillations
known to occur in the other parts of the world. Still there are several lines of evi-
dence which indicate that a phase of reduced precipitation closed in the Selkirks

GLACIERvS OF THE CANADIAN ROCKIES AND SELKIRKS. 1 13

and Rockies about the year 1897 or 1898, and that since then the average
for the series of years is in excess of the normal. These lines of evidence consist
(i) in the records kept by the station agent at Glacier House of the snowfall,
and of the records of the Canadian Meteorological Service for Banff and Calgary.
With the exception of Agassiz, which appears to be one of Bruckner's "excep-
tional coast stations," the other records do not reach far enough back to be of
service. (2) The notes and photographs of the Messrs. Vaux in the Asulkan
Valley, made in 1898 and 1899, when contrasted with those of later date, indi-
cate the close of a series of years with 1897-8, during which the snowfall was
much less than since. (3) The thickness of the strata in the ne've' of the Asulkan
Glacier, assuming that they represent annual accumulations, indicates at the
point photographed in 1902 that three years of diminished precipitation closed
with 1897-8 and were followed by four years during which the average precipita-
tion was in excess. (4) In 1883 and 1884 Dawson found over a belt of country
140 miles long in the western part of the Rocky System, evidence of a recent
high- water stage of the lakes, which resulted in the killing of trees that must have
grown during a prolonged low- water stage that preceded. The condition of the
trees indicated that they had ' ' been killed within a few years. " If we assume that
the wet phase that gave rise to this condition of the lakes closed about the year
1880, then we should expect the inauguration of another wet phase about the
years 1897 or 1898. Finally (5), from the photographs that have been made of
the Illecillewaet Glacier we have proof of the long-time oscillations of the level
of the ice about the ne" v£ line, giving rise to

c. Ice waves. When photographs taken from the identical view-point in
1888, 1897, and 1905 (plates xxxvi and xxxvn) are compared, they show a
marked fluctuation in the height of the ice along the sky-line. The ice appears
to have been at a minimum about 1888 and to have been approaching the
same condition in 1905, possibly attaining it during the current season of 1906.
The crest of a wave, or impulse of ice from the ne've' appears to have reached the
sky-line about 1897 to 1899. The time from trough to crest would represent a
quarter of the period, that from trough to trough, half of the period of the com-
plete oscillation of the ice wave. In this case our data would indicate a period
of 36 to 40 years, agreeing well with the precipitation cycles to which these ice
waves are to be ascribed. It has been shown by the work of Finsterwalder,
Bltimcke, and Hess that the advance of a glacier is due to the progress of an ice
wave along its length, moving more rapidly than the ice itself. The Illecillewaet
in 1887 was experiencing about its nose the last stages in the effect of the arrival
of such a wave. The trough, then at the sky-line, moved valleyward and per-
mitted the retreat of the nose of the glacier, which retreat was probably at its
maximum about the year 1895 or 6, or some 8 or 9 years after the start. The
data for 1902 to 1905 seemed to indicate that an advance was about to be
inaugurated but the very marked retreat of 1905-6 shows that this ad-
vance has been somewhat delayed. When later the year is known at which date
the advance was most rapid we may figure the rate at which the wave travelled the

114 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

length of the glacier. It is a matter of much interest to try to connect the
crests and troughs of these ice waves with the corresponding wet and dry phases
of the precipitation cycles noted above. Since the crest of the wave arrived at
the sky-line in 1897-9, just at the close of the dry phase and the beginning of a
damp one, the wave must be referred back to the damp phase of the preceding
climatic cycle, closing in the late yo's, or early 8o's, so far as we may judge from
the observations of Dawson upon the level of the lakes in the western Rockies.
This would give the "reservoir lag" of one-quarter of the period, required by
Reid's calculations, and an additional 16 or 18 years for the impulse to reach the
crest of the rim. The trough resulting from the dry phase closing in 1897 appears
to have moved out from the reservoir more promptly, possibly owing to certain
local conditions.

3. PIEDMONT TYPE OF GLACIERS.

Three representatives of this unusual type of glacier were found, two of which,
the Wenkchemna and Asulkan, are here described; the third is the Horseshoe Glacier
at the head of Paradise Valley, in the Rockies. This type of glacier is always
compound, being made up of a series of glaciers of the common Alpine type, all
of coordinate importance, which coalesce laterally but retain their individuality
from n6v6 to nose. Since none of them are tributary to any of the others, but
independent in all essential respects, they are here referred to as commensal
streams, in order to indicate this relationship. These separate streams have
temporarily united forces and found strength in the union. In the case of the
Wenkchemna it is very probable that very few, if any, of the commensals could
exist separately. In its earliest stage of development the piedmont glacier begins
as a series of Alpine glaciers, either with or without tributaries, lying in neighbor-
ing valleys. With the increase in precipitation the level of the surface of the sepa-
rate streams rises until they cover the divides between adjacent streams, or the, at
first, separate Alpine glaciers reach out upon the pied-mont and there coalesce
laterally. In its senile condition, a stage to be reached sooner or later, the pied-
mont glacier returns to its condition of youth and disintegrates into its component
streams, as illustrated by the Asulkan of to-day. In the case of the Horseshoe
Glacier some sixteen different commensal streams may be recognized, the most
western four or five of which have almost completely separated from the
others. The glacier has a meager snow-field, the supply for which is ava-
lanched from the slopes of Mts. Ringrose (No. 10), Hungabee, Lefroy, and the
southern side of the Mitre. Observations upon the Wenkchemna showed that each
separate nose may have its own independent behavior and that the movements
of the glacier as a whole cannot be known unless data are collected for each
component stream.

4. PARASITIC GLACIER.

From the hanging glacier upon the eastern shoulder of Mt. Lefroy there is
avalanched to the back of the Mitre Glacier quantities of snow and ice, falling

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

a vertical distance of some 2,000 feet and accumulating along the base of the
cliff. The ice of the glacier is broken into fragments, some of it disintegrating
into its component granules and much of it ground into ice dust, destroying
completely the stratification and lamination of the hanging glacier. The ava-
lanches occur mainly during the late spring, summer, and early fall, and as a
result of -the spreading of the fragments from sliding and rolling there is made
each season a stratum of ice similar to those ordinarily found in the n6v6 region.
Regelation is complete and there arises what is known as a reconstructed, or
regenerated glacier, with its strata leading to and dipping towards the region
of accumulation. The weight of the ice here forces the lower strata to move
out at right angles to the cliff face and a forward movement is imparted to the
ice directly across the Mitre Glacier upon which it rests. This regenerated Lefroy
moves about one-half as fast as the underlying Mitre, so that before the latter has
reached the Victoria, the Lefroy has crossed to the opposite side of the valley.
Between the hanging Lefroy Glacier and its bed there is being manufactured a cer-
tain amount of ground-morainic material, which is incorporated into the strata
of the regenerated Lefroy, and moved across the valley as a result of its motion.
While this is taking place, however, the Mitre is carrying the entire Lefroy down
the valley and the actual motion of the debris is the resultant of these two
motions by which there is accumulated at the base of Mt. Aberdeen a great heap
of ground-morainic matter, with a dressing of angular material from the face of
the latter mountain. The ground moraine rests upon the back of the Mitre
and some of it is ridged parallel with its side, in which form it is dealt out to
the Victoria and constitutes the main bulk of its right lateral moraine. This
Lefroy Glacier is distinct from the Mitre, upon which it rests, in that it is a differ-
ent type, is nourished differently, has a different form, a distinct set of strata
unconformable with those of the Mitre, has a different direction of motion, a dif-
ferent rate of motion, and is accomplishing a wholly different geological work.
The glacier is parasitic in the sense that it is carried by its host and is nourished
from snow and ice that might otherwise be available for it. It is not parasitic in
the sense that it draws its sustenance from the Mitre itself.

It is probable that glaciers of this type are now, and have been, more common
than has been generally recognized. It seems likely that at a certain stage the
glacier in a hanging valley would sustain more or less of this relation to the trunk
glacier. By means of such a glacier we may account for the lateral transporta-
tion of materials across a valley and a transportation that would leave no record
upon the bedrock. If two distinct glaciers may occupy the same valley simul-
taneously, it seems probable that t\vo ice sheets of the continental type might
be superposed, flowing in different directions, the upper delivering material to
the lower.

5. BEAR-DEN MORAINES.

The bear-den type of moraine is so exceptional that some special explanation
must be found by which we may account for the accumulation of coarse mountain

Il6 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

fragments without the usual filling of fine materials. The size of the blocks
themselves is not so remarkable, knowing what a transporting agent a glacier
is, as the average size of the fragments making up the moraines. In the case of
four of the five glaciers studied, two of these moraines were found and only two.
The absence of them in the case of the Yojio is readily understood when the
lack of high cliffs is noted. Not one of the glaciers at the present time could
form such a moraine, no matter how prolonged the halt. The blocks are angular
and show no more glaciation than they might have received upon one face while
they were in their original position in the cliff. The blocks were carried upon
the ice and were not pushed or dragged along in front of, or beneath it. The fine
material was not removed by the action of running water, as might have been
done in other cases; but was absent from the first. If we are to account for these
moraines we must load the glaciers with a mass ot exceptionally coarse blocks
and only a minimum of fine debris. This cannot be done by assuming two periods
of excessive weathering for they would produce as much fine material as coarse,
and very probably a great deal more. The prevalence of the phenomenon pre-
vents our resorting to the ordinary rock slide for our explanation. In the case
of the Victoria it built a moraine of the ordinary type, then the two bear-den
moraines, and then the present modern moraine essentially like the first. The
Asulkan built its outer coarse moraine, then one of the ordinary type, then its
younger coarse moraine and subsequently a series of the common variety. An
examination of the various cliffs, in connection with each of the four glaciers,
from which the material was most certainly derived shows that they all have
a trend from north-northwest to west-northwest. A further suggestive fact
is that in all cases the bulk of the material fell to the eastward.

During the season of 1904 no plausible explanation occurred to the writer,
but upon leaving the field the idea of a double seismic disturbance came up and
was carried back to the mountains in 1905 . It seems now to be the only explana-
tion by which to account for the phenomenon. Slipping may have occurred
along some of the numerous fault planes traversing the eastern Rockies in a
north-northwest direction and the region crossed by westerly moving seismic
waves. From cliffs having a general northwest-southeast trend, blocks, already
much weathered, would be detached and thrown eastward, comparatively few
falling from the westerly facing cliffs. Glaciers most favorably situated for
acquiring a load by this means, as the Wenkchemna, have the most massive
deposits of the nature described; those unfavorably related to high cliffs, as the
Yoho, appear to have made none. The great blocks detached by the earth
jars fell into soft neVe", or upon the yielding ice, and were not ground into small
fragments as they usually are when they descend to the valley floors. The
protection afforded the ice by the material brought about a halt of the front,
until the blocks were deposited, when the general retreat was resumed.

It was reasoned that if the disturbances assumed reached from the Great Divide
to the Selkirks, then many other glaciers, equally favorably situated for acquiring
a load by this means, should show the same type of moraine, if they were not

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. II?

tributary to other ice streams at the time. Further we should expect to find
occasional rock slides of the same age as the moraines and cliff debris that did
not reach the back of a glacier. The latter material not being concentrated
would be inconspicuous. In 1905 there was but little time for a general examin-
tion of the region but visits were made to the Horseshoe and Geikie glaciers.
The former lies between the Wenkchemna and Victoria, with its main extent
of vertical cliff extending to the northeast, but with a considerable portion
extending from Hungabee to the Wastach Pass, with a westerly to northwesterly
trend. Opposite this portion of the glacier there is a deposit of very coarse
blocks, that were dropped upon the crest and outer slope of a still more ancient
moraine, consisting largely of a stony till. The number, however, is very meager
compared with those in the Valley of Ten Peaks, and would call for no exceptional
explanation. A low ridge of coarse blocks occurs just inside, showing best about
the front of the nearly detached western portion of the glacier, which is correlated
with the inner of the bear-den moraines. In the case of the Geikie Glacier, lying
at the head of Fish Creek Valley and nourished from the southern portion of the
Illecillewaet neVe" (plate xxxm) , no moraines of the type sought were found within
a distance of one and one-half miles of the nose. Although the cliffs are suffi-
ciently steep to have supplied the material their general trend is northeast-south-
west, i. e., in the direction of supposed earth movement, and they suffered
relatively little destruction. From the eastern face of Mt. Burgess there has
been dropped a mass of coarse rock, which more strongly suggests the morainic
deposit seen in the valleys than that seen anywhere else outside the reach of
the glaciers. In many of the talus slopes there are many coarse and fine
blocks, which look to be of nearly the same age, instead of showing the
gradation that we might expect. In their work referred' to upon page 4,
Collie and Stutfield describe a mass of rock debris in the valley of the
Athabasca (page 126) that may represent one of these ancient coarse
moraines or a modern one in process of forming. In referring to peaks
Woolley and Stutfield they say, "These two last mountains appeared to
have been conducting themselves in a most erratic manner in bygone ages.
A tremendous rock-fall had evidently taken place from their ugly bare lime-
stone cliffs; and the whole valley, nearly half a mile wide, was covered to a
depth of some hundreds of feet with boulders and debris. What had happened,
apparently, was this. The immense amount of rock that had fallen on the glacier
"below Peak Stutfield had prevented the ice from melting. Consequently the
glacier, filling up the valley to a depth of at least two hundred feet, had moved
bodily down; and its snout, a couple of hundred feet high, covered with blocks
of stone the size of small houses, was playing havoc with the pine-woods before
it and on either side. In our united experiences, extending over the Alps, the
Caucasus, the Himalaya, and other mountain ranges, we had never seen indica-
tions of a landslide on so colossal a scale." In a footnote they add, "The re-
mains of a similar landslide were afterwards noticed blocking the outlet to
Moraine Lake in Desolation Valley."

Il8 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

If the seismic theory furnishes the true explanation of these double massive
moraines, then we have a means of correlating the positions of the extremities
of all the glaciers showing them at these two stages in their history, also data
for determining their actual retreat since and their relative rates. Glaciers, that
were not tributary to others at the time, confined between steep cliffs, having
a northwest to southeast trend, may be expected to show such moraines. There
is the possibility that any particular glacier may have advanced since and have
overridden one, or both, as the Victoria and Wenkchemna have partially done
with the inner of their series. Numerous earthquakes must have occurred
during the long Pleistocene period, but the cliffs were so completely blanketed
in snow that we find no such records left in the trunk valleys. Similar moraines
should be found in other sections of the world, but they might originate from
the removal of the finer materials by running water, as well as by earthquakes
and simultaneous rock slides. As to the actual age of these moraines we may
only loosely speculate. The blocks look old and the schists and sandstones have
disintegrated more or less, in situ, but undoubtedly they were badly weathered
before the glaciers got possession of them. The age of the moraines is to be
expressed in centuries rather than thousands of years. Based upon our vegeta-
tion data we may conclude that the inner of the two moraines was completed
about five or six centuries ago and that the earthquake disturbance respon-
sible for it may have occurred two centuries earlier. The outer of the two
moraines seems to be about two centuries older.

6. SURFACE FEATURES.

a. Dirt bands, zones, and stripes. In Chapter III of this report the writer
has described and figured these three glacial features and has suggested that
certain terms, used rather indiscriminately for any one, be restricted to a single
feature. The first two are very often confused, one with the other, but are so
essentially different in their real nature, if not always in their appearance, that
they should be sharply separated and differently named. The dirt zones, or
simply the zones, when the foreign matter is not present to discolor them, are
the outcropping edges of the strata of which the glacier is composed. They show
to best advantage about the nose and lower margins of the glacier that is suffi-
ciently free from debris, as broad, parallel zones encircling the lower extremity
and passing around to the sides where they disappear. They are usually convex
down-stream, but the form they assume is determined by the configuration of the'
glacier's extremity. In case the stratification in the glacier is absent for any
cause, there can be no zones seen.

The dirt bands are entirely superficial and result from the collection of fine
debris in long hollows or troughs that first extend transversely across the glacier,
but which become convex down-stream from the more rapid central motion of the
ice. They occur in series, roughly parallel and regularly spaced, and assume
finally a pointed, or hyperbolic form, which probably suggested to Schlagintweit
the term "ogiven." The name "dirt band," however, was originally assigned

GLACIERS OP THE CANADIAN ROCKIES AND SELKIRKS. I 19

to them by Forbes, their discoverer, and is in more general use. The trans-
verse, parallel troughs, in which the dirt bands have their origin, arise from the
incomplete healing of .ransverse crevasses which occur at the crest of a steep
ice slope. The lips of a crevasse, exposed to intense solar action are rounded
more or less and when the crevasse closes there is left a trough which marks the
position of the original crevasse. Into this depression wind-blown dust collects
and is washed from the adjacent slopes. By absorbing heat this dust may em-
phasize the depression slightly and may render the ice somewhat spongy, as
pointed out by Tyndall. If the ice slope is too steep, a cascade results and the
ice is too much shattered to show the bands, or to allow them to form. If the
slope is steep, but regular, with much melting over the surface, the site of
the bands will be destroyed before any complete series can develop. Conditions
are most favorable for their production upon the face of a moderately steep slope,
which is immediately followed by a long stretch of gently inclined ice. They
sustain no necessary relation, whatever, to the dirt zones, being present when
the zones are absent. When both zones and bands are present they may be
comformable for a greater or less distance and may be difficult to distinguish
from one another. In the case of such a glacier as the parasitic Lefroy the zones
and dirt bands may be discordant and intersect at high angles. There is reason
for thinking that the dirt bands are produced annually, only the summer formed
crevasses furnishing the necessary troughs, while the few winter crevasses com-
pletely and perfectly heal in passing down the slope. If this proves to be the
case we have a means of determining the approximate yearly motion of the ice
along the slope and a clue to the extent of the longitudinal compression, or
extension, of the ice subsequently.

Where the edges of the blue bands, embedded in the more porous whiter ice,
outcrop upon the surface, particularly along the margins of the glacier pressing
firmly against the valley wall, there is developed a further miniature banding.
The firmer blue ice melts less rapidly than the more vesicular layers and a series
of parallel ridges and troughs results, the course, distance, and average breadth
of which is determined by the ice structure itself. In the narrow troughs the fine
dirt collects and the ice is marked with a series of delicate parallel dirt streaks.
Tyndall compared them with the marks left in a gravel walk by a garden rake.
Drygalski describes them under the name of Schmutzbdnder, but this term must
be reserved for the true dirt bands of Forbes. Dirt stripes suggests their appear-
ance and will enable them to be distinguished from all the other dirt features. In
that they owe their existence to the actual structure of the ice they have some
relationship with the dirt zones, but in that the dirt of which they are composed
is purely superficial, they are more nearly related to the dirt bands. They are
to be seen at only a short distance, while the zones and bands are best brought
out from a distant, elevated view.

6. Differential melting effects. Under favorable circumstances an interesting
series of stages may be passed through by dust wells; dirt, sand, and gravel
cones; boulder mounds; lakelets and morainic ridges. This was first worked out

I2O GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

by Russell upon the Malaspina for the lakelets and boulder mounds, but it
applies also to the other features as well. Dust wells may persist through a
season, or a series of seasons presumably, the dirt patches to which they owe
their existence being continuously retained in the miniature wells. Although
very shallow at any one time, their total depth might measure many feet. From
wind action and small trickles of water more dust is being added slowly and in
time there may be enough to protect the bottom, instead of causing its melting.
The dirt now appears at the surface and the ice beneath melts less rapidly than
the unprotected adjacent ice, giving rise to a miniature cone, marking the original
site of the well. Such cones are found of various sizes and covered with dirt,
sand, or gravel. By lateral melting the slopes eventually become so steep that
the veneering slides off, or it may be washed down by heavy rains and distributed
about the base of the ice cone. The bare ice is now attacked by the sun and a
hollow is produced where the cone stood, about the rim of which stands more or
less of the material by which it was covered. This material rolls and slides back
into the depression as the sides are widened and steepened by melting. When
enough has been concentrated at the bottom and about the sides to prevent
further melting, the adjacent ice which has lost its protective cover, just in
proportion as the depression has gained, now melts away to a level with the
bottom and then still lower, causing the material collected in the basin to again
assume the form of the cone. The miniature examples of this action might pass
through these stages several times in the course of the season, while the boulder
mounds and lakelets would require many seasons for the completion of a single
cycle. In the case of a medial moraine, or a lateral resting upon ice of sufficient
thickness, the same stages may be passed through, except that when the material
is shed it assumes the form of a double ridge, between which the elongated
trough is developed and into which the de"bris may slide to produce a single
ridge again. In this way the superficial debris of a glacier may be subjected
to much tossing and bruising before it comes to rest in the frontal or ground
moraine. In the case of a debris-covered ice surface all that is necessary to start
the process is to have the material unevenly distributed, a little thinner or a
little thicker patch of foreign matter.

7. ICE STRUCTURE

a. Stratification. From a comparison of the thickness of the strata in the
Asulkan with the available records of snowfall it seems probable that the strata in
this glacier, as well as in the Illecillewaet and Yoho glaciers, represent the
annual accumulation of snow in the region. The fall snows are combined with
those of the following winter and spring, compacted by the summer's melting
and rainfall into a white, porous stratum of granular ice. At any given place
upon the ne"ve" by means of wind action a stratum may have gained, or lost in
thickness. Owing to the deposition of the snow in successive layers and the
periodic distribution of wind-blown rock debris, each stratum acquires a more
or less distinct lamination; conformable with the stratum itself. During the

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 121

summer melting the fine dirt is concentrated at the surface, forming a soiled
streak which contrasts strongly with the fresh snowfall of the fall. The water
resulting from the surface melting and rainfall sinks into the stratum and con-
tributes to the growth of the n6v6 granules, forming a crust of different texture
and color, by which the strata may be distinguished when no dust is present. In
the case of a regenerated glacier, such as the Lefroy, the stratification results
from periodic avalanching of snow and ice during the late spring, summer, and
early fall. The strata may vary much in thickness and have no immediate con-
nection with the amount of precipitation. They may become charged through-
out with ground-morainic material and give rise to very distinct zoning. When
a glacier is fed in part by neVe snow, and in part by avalanches from hanging
glaciers the stratification may appear very irregular, as in the case of the Victoria.
In passing an ice cascade the stratification and lamination may be completely
destroyed, or the uppermost strata may be destroyed and the lower more or
less perfectly preserved, as pointed out by Reid. It is not supposable that the
stratification could be thus destroyed and the more delicate lamination preserved.
In the case of the regenerated Lefroy the stratification is restored, after having
been lost, but it is not possible to restore the lamination completely, or regularly,
in the case of such a glacier. It should be noted in this connection that under
exceptional conditions shearing planes may be developed in the body of the
glacier which do not coincide with the limiting planes of the depositional strata.
In this way there may be acquired another type of secondary stratification
having no relation whatever to that which originates in the neVe".

b. Shearing. Observations upon the oblique front of the Victoria in 1904
indicated that the upper strata were moving bodily over those upon which they
rested. The upper strata projected more and more daily, when there was not
enough additional debris in the lower to account for the phenomenon by differ-
ential melting. A small amount of sand and fine gravel, washed down from
above, collected in the lee of the upper projecting layers. Some days this was
in small enough quantity to accelerate the melting of a narrow strip of ice upon
which it rested, but quite as often melting was retarded by the material. At
one place where the shearing action seemed pronounced three heavy spikes were
driven into the base of the upper stratum and three corresponding ones in the
face of the subjacent layer. These spikes were six inches in length and were
driven horizontally into the ice until their heads were flush with the surface,
about eighteen inches apart. The average surface slope of the ice was 46° and
the vertical height of the ice 50 to 52 feet. The upper stratum had a thickness
of about three feet, the lower two feet, and each contained, apparently, about
the same amount of foreign matter, and this small in amount. At the beginning
of the observations the upper stratum projected 19.7 inches beyond the lower
(July 21), and by August 3, 25.6 inches, showing a gain in the 13 days of about
six inches of the upper beyond the lower. The spikes were visited daily and
reset and showed that while the upper stratum was advancing with reference
to the lower it was also melting back more rapidly, because of its more exposed

122 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

position. The average daily melting about the spikes in the upper stratum was
0.23 of an inch in excess of that about those in the lower and proved that a differ-
ential movement of the strata was taking place.

c. Blue bands. It seems highly desirable to distinguish the minor stratifica-
tion seams, originating in the ne"ve", from the blue bands, blue veins, or ribbon
structures, that have had an entirely different origin. The first step toward
such distinction is to have a separate term for each of the two types of structure
and the writer suggests that lamina be used exclusively for the minor layers of
which the strata are composed and that blue bands,1 already in such general use,
be restricted to the structures commonly included under the term, however
they may have been produced. When they are each made the object of com-
parative study it should be possible to distinguish them. We should naturally
expect the laminse to become less and less distinct toward the nose, and to appear
continuous, while we find the blue bands there showing very typically and being
discontinuous. The same stratum might show both structures, either conform-
able, or cutting one another at various angles. In the case of simple glaciers,
Agassiz and Reid have succeeded in tracing the laminae from the neVe" to the nose.
The structures seen in the Canadian glaciers are blue bands, rather than
laminas, since they are developed in great perfection where the strata have been
completely destroyed, as in the case of the regenerated Lefroy and almost
obliterated as in the Yoho and Illecillewaet glaciers. In general their position is
at right angles to what may be assumed to be, or to have been, the direction of
maximum pressure. They are seen best along the margins where the glacier is
closely confined between rocky walls, extending parallel with the sides, dipping
downward and inward at a steep angle. Beneath the medial moraine upon the
Victoria they are vertical to fan-shaped. At the foot of ice cascades they may
extend crosswise of the glacier. Having the same origin and being essentially
alike, it does not seem wise to use different terms by which to separate these,
such as marginal structure, longitudinal structure, and transverse structure, as
suggested by Tyndall. Contorted patterns and faultings are to be accounted for
by assuming differential movements in the ice after the formation of the bands.

When followed for a short distance, in either direction, blue bands are found
to thin out to an edge and disappear, showing that they have a very flat,
lenticular shape. Separate bands overlap and are felted together as are the
bands in a schistose or gneissic rock. They strongly suggest schistosity in rocks
and not stratification. The ice of which each band is composed is more compact,
more free from air bubbles, and a deeper blue than the ice in which it is embedded.
That they have been produced by pressure and stand at right angles to it, when in
process of formation, as demonstrated by Tyndall, seems most probable. That

1 The term band alone, or banding, as suggested by the glacial conference in August, 1899, is not fully
satisfactory since it does not distinguish this structure from the dirt bands of Forbes. The following terms
have been applied to this structure by various writers; Bandstruktur, Banderung, Blaubanderung, Blatter-
struktur, Blaublatterung, Blaublatterstruktur, blaue Bander, blaue Streifen, Schieferung, Schichtung,
parallele Struktur, stucture rubanee, structure lamellaire, ribboned structure, blue veins, blue leaves, blue
bands, lamellae, laminae, lamination, and stratification.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 123

these bands represent portions of the glacier that have been completely liquified
by pressure, allowing the air bubbles to escape, seems, to the writer, very im-
probable, for four reasons, (i .) Blue bands occur in the basal layers, parallel with
the valley floor. The thickness of the ice of an ordinary glacier is not sufficient
to induce general melting by its simple weight. (2.) If the granular structure of
the glacier is completely destroyed by melting, it cannot be reproduced by simple
freezing, and still the granules are best developed in the blue band areas. (3.) Oc-
casional granules may be found which extend from the blue bands into the ad-
jacent vesicular ice. (4.) Water freezing in cavities in the body of a glacier should
form a series of prisms, standing with their main axes at right angles to the ice
surfaces bounding the cavity. Such filled cavities are found in the ice but they
do not constitute blue bands.

d. Ice dykes. In connection with the Lefroy Glacier chiefly, there were
noted in the early summer what appeared to be former crevasses, filled with ice and
forming ice dykes in the body of the glacier. Some of these were cut by crevasses,
testifying to their greater relative age and suggesting that they might persist
from one season to another. A few of the dykes contained granular ice, the
granules being moderately coarse, and were assumed to have been formed by the
filling in of crevasses with ice avalanched from the hanging glacier upon Mt.
Lefroy. Most of the dykes, however, were completely filled with a double tier
of ice prisms, having their bases attached to the walls of the crevasse and
extending horizontally out into the cavity, at approximately right angles.
Generally the prisms met at the centre those from the opposite face of the crevasse
and their inner ends interlocked. Sometimes a space was left between the oppo-
site tiers of prisms. Ellipsoidal shaped spaces were also found completely filled
with radially arranged prisms meeting at the centre. The explanation given for
these features is that they were formed by the freezing of water in crevasses,
and other cavities, in the spring, or early summer, while the glacier still retained
a sufficient degree of its winter's temperature. The water was supplied by the
early melting, or rains, and the freezing surfaces were the walls of the crevasse,
instead of the lower stratum of the atmosphere, as is usually the. case. Since in
freezing, water forms a series of parallel prisms, with their axes lying, as a rule,
at right angles to the surface of refrigeration, these prisms have the abnor-
mal horizontal position, instead of the usual vertical one. Although the upper
part of the dyke may have been lost by melting, there was no evidence that a
horizontal stratum of ice had formed across the top from freezing induced directly
by the atmosphere.

Drygalski has argued that it is pressure that determines the direction
that the crystalline plates will assume when water is freezing, and that ttie
main prismatic axes will lie parallel with this pressure. In the case of
the ice of a lakelet or basin, after it has once been enclosed by the ice
cover, the under side will be subjected to an upward pressure owing to
the gradual expansion of the water as it is brought to the temperature of
freezing. But the orientation of the basal plates, parallel with the upper

124 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

surface, began before the cover was completely formed and hence before
such pressure could have come into operation. As pointed out by Miigge
they also assume this position when an opening through the ice cover is
artificially maintained. Miigge believes that the plates are simply floating
in their position of equilibrium and that the pressure has nothing to do
with the orientation of the plates. That this, however, is not the cause of
the orientation is shown by the position of the columns in the ice dykes
above described, where the plates have formed in a vertical position, while in
the case of the ellipsoidal water-filled cavities they have formed at all angles
between the vertical and the horizontal. In the case of the ice dykes the forma-
tion of an ice cover would have given rise to a lateral pressure, as well as an
upward one, and the position of the plates in the horizontal columns would have
been in harmony with the view of Drygalski. No trace of this cover, however,
was seen, and it seems probable that the columns would still have formed at
right angles to the cold walls of the crevasse, under none other than hydrostatic
pressure.

Some experiments still in progress in the freezing of water in variously shaped
vessels lead the author to believe that the basal plates are placed parallel to the
surface of refrigeration, independently of pressure or position of equilibrium.
The actual congealing temperature enters quiet water at right angles to this
freezing surface, regardless of its position, and each successive plane of mole-
cules in turn feels the effect of the crystallizing force. The result is that sheets
of molecules are successively frozen parallel with the requisite isothermal surface
as it slowly works its way into the body of the water. The orientation of the
plates is facilitated by the fact that in making the ice crystal the molecules ar-
range themselves more readily (because of the superior crystallizing force) in
the plane of the secondary axes than in the direction of the principal axis. This
is shown by the form of the snowflake which has been produced supposedly under
conditions in which the crystal was free to grow in any direction, so far as the
supply of moisture and suitable temperature are concerned. As is well known
the molecules are arranged mainly about the short main axis in the plane of
the secondary axes. The principle is illustrated further by the frost crystals which
form upon the window-pane, with cold air upon one side and a relatively warm,
moist atmosphere upon the other. At first only a very thin layer of moisture,
parallel with the surface of the glass, can congeal, and in this layer the molecules
at once arrange themselves in the plane of the secondary axes. As the atmos-
phere supplying the moisture becomes cooled for some distance back from the
glass the crystals may grow more or less irregularly. That the cohesive force in
the ice crystal is much more powerful in the direction of the basal planes than
in the direction of the principal axis, is demonstrated in the experiments to be
noted later (p. 130). Pressures in a direction at right angles to the main axis
will cause the basal plates to slide over one another, as in a bunch of tickets,
but no such shearing action can be secured when the direction of pressure is par-
allel with this axis. According to the view of the writer the temperature con-

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 125

dition for the crystallization of the water is supplied successively parallel to the
refrigerating surface, whatever may be its position or form, and the molecules
yield to the relatively more powerful forces which are operative in the planes of
the secondary axes.

e. Glacial granules. Glacial ice, which has not been subjected to a melting
temperature, is firm, solid, and, apparently, homogeneous, except for air bubbles
and foreign matter that it may contain. It is brittle, breaks without cleavage,
and in quantity, when pure, has a rich blue color by transmitted light. Sub-
jected slowly to a melting temperature there is developed a system of delicate
capillary tubes, which form a network throughout all the ice affected, and ex-
tend into the body of the glacier a number of feet. These tubes outline the
granules, more or less perfectly, of which the entire glacier is composed. These
granules are irregular polyhedrons, of variable size, with curved faces which
interlock with one another. Ordinarily there are no spaces between them that
can be recognized and there is no cementing material to bind the granules
together. They are observed to increase in size from the neVe" to the nose in any
particular glacier and there can be no doubt but that the granules formerly in
the ne"ve are directly related to those seen in the lower part of the glacier. In the
Canadian glaciers studied the largest granules were seen in the basal layers
about the nose; the Asulkan, the smallest of the glaciers, having the smallest
average granules, and the Yoho, the largest glacier, having the largest average
granules. When subjected to considerable melting the capillary tubes become
"irregular and very thin spaces open between the faces of adjoining granules,
allowing the granules eventually to fall apart, or to be easily pulled apart.

Each granule is an incomplete ice crystal, incomplete because its development
has been interfered with by the neighboring crystals. Belonging to the hex-
agonal system of minerals, it has a single main axis, which is also its principal
optic axis. In common with all known ice crystals it appears to be made up of a
bundle of very thin plates, placed with their flat faces together, the axis standing
at right angles to these plates. When the granules have melted apart the very
delicate edges of these plates, or more probably sets of these plates, may often
be recognized extending as delicate parallel lines about the granule and thus indi-
cating the positions of the planes of the secondary axes. These lines are known as
" Forel's stripes." They are referred to by Mxigge as the "Translations Streifung,"
and were regarded by him as due to the partial shearing of the basal planes over
one another. They are found, however, in newly forming crystals of ordinary
lake or pond ice which have not been subjected to any shearing stress. Within
the body of the granule there are seen, at times, circular disks of excessive
thinness, with their flat faces perfectly parallel and all at right angles to the optic
axis. They are of silvery whiteness and appear like "flattened air bubbles,"
as they were originally described by Agassiz (plate vi of Atlas, figure 10). These
are "Tyndall's melting figures" and are cavities, "vacuous space," con-
taining nothing more than water vapor, resulting from the internal contrac-
tion of the water, as it changes from its solid to its denser liquid condition.

I 26 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

The melting begins at certain points between the crystalline plates and spreads
in a direction parallel to them, instead of across, or through them. The planes
of these melting figures are parallel to Forel's stripes, and either feature when seen
may be used for orienting the crystal. In addition to the stripes of Forel, there
is to be seen a very conspicuous system of parallel ridges and furrows, covering
the outside of softened granules, which can have no connection whatever with the
crystalline structure. The ridges are either continuous, or consist of a series of
regularly placed points, forming a wavy, irregular pattern about the crystal.
The appearance suggests that seen upon the inside of one's finger-tips and thumb.
It shows itself when the adjacent faces of the granules begin to separate and is
due to differential melting at the surface, but it is far from clear what could give
rise to such a regularly irregular pattern.

By means of the polariscope it was found that there is a tendency towards the
orientation of the granules about the nose of the Victoria, Yoho, and Illecillewaet
glaciers, the other two not being tested. The Victoria shows distinct stratifica-
tion about the oblique front, the Yoho indistinct, and in the case of the Illecille-
waet, the stratification about the nose seems to have been completely destroyed.
Vertical sections of the ice were prepared, cut crosswise and lengthwise, and these
were compared with horizontal sections and oblique sections. It was found that
there is a marked tendency to arrange the optic axes of the granules in the basal
layers near the nose in a vertical position, from one-fourth to one-third of them
being estimated to be so oriented. The cause of this orientation is not yet
apparent, but connected, undoubtedly, with the method of growth of the granules
themselves. In order to account for the orientation which he found in the
Greenland glaciers, Drygalski assumed that the granules were separately melted
and refrozen with their axes parallel with the direction of pressure, which he
considers at right angles to the strata. If it is true that the direction of pressure
determines the position that the crystalline plates will assume, and hence the
position of the optic axes, which the writer seriously questions, then the space
occupied by a single crystal, which has been completely melted, should contain
a large number of radially arranged prisms, each standing at approximately right
angles to the portion of the ice surface to which its base is attached. Owing
to the law of transmission of forces by a liquid the pressure is equal in all directions
whether this pressure arises from the weight of the superincumbent ice, or because
of the expansion of the water in the closed cavity just before freezing. Drygalski
is in error in supposing that the pressure experienced by the liquified granule is
vertical only, since, if confined, the water would press outward in all directions.
In case the position of the refrigerating surface, or surfaces, is the cause of the
orientation of the plates, then in the closed cavity occupied by the liquid granule,
there should be formed a mass of radially arranged prisms, similar to those
observed by the writer upon the Lefroy and by Agassiz upon the Aar. In
either case, the cavity should be filled with small radially arranged prisms and
not by a single crystal with its axis in a vertical position. This furnishes rather
conclusive evidence that granules and blue bands never have existed in a com-

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 127

pletely liquified condition. That they may have been melted partly upon
one face and frozen upon another, or the water derived from one by melting
added to another granule, is in harmony with known properties of ice. Pulfrich
found that when an ice crystal was pressed against a wet surface of glass and
allowed to freeze the water between the ice and plate was incorporated into
the crystal, so as to make a homogeneous mass.

Much interest is attached to the methods of granular development, since the
more modern theories of glacial movement are more or less dependent thereon.
It has been shown by Emden, Drygalski, Crammer and others that when neVe"
granules are made into a water slush, such as might originate from excessive melt-
ing, or heavy rainfall, the granules grow in size and, under favorable conditions,
quite rapidly. In the n6v6 as well as in the- body of the glacier, however, there
must be a maximum limit which the granules may attain by this means for the
ice will presently become too compact for more water to enter and no space will
be left for the growth of individual crystals.1 That further growth of the granules
does not take place by the simple freezing together of neighboring granules, is
conclusively shown by the homogeneous structure of the mature granule. That
new granules cannot originate by the complete and simultaneous melting of a
number of adjacent smaller ones, is believed to have been just shown in the
preceding paragraph. Of the various theories of granular growth remaining
we may recognize three divisions, based upon evaporation, melting, and "dry
union."

First. — It has been shown by Chamberlin and Salisbury that in dry granular
snow, kept continually below the freezing temperature, certain granules will
grow in size at the expense of the others, presumably by the giving off and con-
gealing of water vapor.2 In the porous snow of the neVe" it seems probable that
the principle would be operative and that the granules would diminish in numbers
and increase in size, even when not immersed in water. For the body of the
glacier, with the granules in such intimate contact, the authors do not believe
that evaporation and condensation can take place to any appreciable extent.

Second. — -Making use of the principle of Thompson that ice may be melted
by pressure, without any change in temperature, many investigators, as Mugge,
Drygalski, Chamberlin, Crammer, etc., have accounted for the growth of the gran-
ules in the main body of the glacier by assuming a partial or complete melting
and refreezing. Those granules which owing to their location are subjected to
the greatest pressure, or internal friction, or those portions of granules similarly
affected will melt, thus redistributing the pressure and allowing the free molecules
to attach themselves to the most favorably located granules. The liquefaction of
the granules may be confined to their outer surfaces, or, as Drygalski believes, take
place locally in the bodies of the granules. Chamberlin believes that the granules

' See Hagenbach-Bischoff's criticism of Forel's infiltration theory, "Weiteres Uber Gletschereis," Ver-
handlungen der Naturforschenden Gesellschaft in Basel, vm, 1889, p. 822.

''Geology, vol. I, Chamberlin and Salisbury, p. 296 (Chamberlin, Peet, and Perisho). "A Contribu-
tion to the Theory of Glacial Motion," Chamberlin, Decennial Publications of the Univ. of Chicago, vol.
IX, p. 194.

128 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

are subjected to more or less of a rotary movement and a sliding along their
limiting surfaces, by which the internal stresses of the glacier are undergoing
constant readjustment and the ice mass permitted to move under the in-
fluence of gravity. These views of granular growth would call for constant
changes in the form, size, and number of the granules and in their relative
position.

If the principles underlying these views — melting under pressure or friction-
were alone operative in granular growth there should occur a much larger number
of smaller granules mingled with the larger in the basal layers about the nose of
a glacier. Owing to the manner in which the granules are keyed together the strain
upon the smaller would be relieved as they diminished in size and would be
transferred to the faces of the larger neighbors. Lying in between the coarser
granules we should expect a considerable number of these smaller remnants,
but such occur only somewhat sparingly. The remarkably well preserved
blue bands in the basal layers about the nose of the glacier furnish conclusive
evidence, it seems to the writer that the granules have not been destroyed since
the bands were produced and that they have not materially shifted their position
with reference to their neighbors. Upon the surface of the lower Asulkan
Glacier these bands were found so thin that thirty were included within the dis-
tance of four inches, several necessarily cutting across adjacent granules. Any
perceptible shifting of the granules as the result of sliding or rotation would give
rise to faulting of these bands, while their destruction by either slow or rapid
melting would cause abrupt gaps in the continuity of these bands. The preser-
vation of the depositional laminae from the neVe" to the nose would seem impossible
if the granules are being destroyed and reformed or rotated out of their original
position with respect to their neighbors.

Third. — The "dry union" of granules described on page 40 of this report
accounts for the reduction in the number and an increase in their size toward
the nose of the glacier. According to this theory the molecules of the yielding
granule give up their own crystalline arrangement and without any apparent
melting are immediately incorporated into the body of the controlling granule.
Heim's view was that such a union could occur only when the main axes of the two
granules were placed in approximately parallel positions,1 but the experiments
of Hagenbach-Bischoff showed that such union could occur regardless of the posi-
tion of the axes,2 and this he regarded as the true cause of granular growth
in the glacier. This view was accepted by Emden in his prize essay, " Ueber das
Gletscherkorn," published in 1890. It furnishes the simplest theory of granular
growth, not of glacial motion; accounts for whatever uniformity exists in the
size of the granules, calls for no shifting of the granule relative to its neighbors,
and hence permits the continuity of laminae and blue bands.

'Handbuch der Gletscherkunde, 1885, s. 330.

'Verhandlungen der Naturforschenden Gesellschaft in Basel. Bd. vii, 1888, s. 192; Bd. viii., :88g, s. 633
und 821. . Archives des Sciences physiques e>. naturelles, T. xxiii., 1890, p. 373.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. I 29

8. THEORIES OP GLACIAL MOTION.

The one question that continually arises in the minds of all glacial students,
with tantalizing frequency is — what is the nature of this glacial motion? Are
we any nearer an acceptable hypothesis than we were nearly seventy years
ago when the serious study of glaciers was begun ? Possibly ! Without attempt-
ing the discussion here of the various theories that have been proposed, the writer
desires to record his convictions after his Pleistocene studies in the Lake Erie
region and his four consecutive seasons about the Canadian glaciers. All are
now agreed that sliding, expansion by freezing or changes in temperature,
general melting under pressure and regelation cannot fully and completely
account for the known facts of glacial motion. Probably all will admit that,
under certain circumstances, every one of these factors may find its application.
In the neVe" region it is possible that a certain amount of rolling and sliding may
occur amongst the granules, producing some motion, such as we may see in a pile
of beans or peas. Farther down in the glacier, although the granules are inti-
mately interlocked, it is quite probable that they would permit of a certain
amount, possibly considerable, motion between their adjacent faces. If we
introduce the idea of a partial melting of the granules, those portions of them
subjected to especial stress, or friction, will yield under the action of gravity
working from above, or behind, and will permit other granules to yield. Upon
relief of pressure, the water will be frozen to the original granule or distributed
to neighboring granules, as discussed in the preceding paragraph, and thus
the movement of the glacier may arise entirely from the alteration and growth
of its component granules. We must assume that the heat necessary for the
partial liquefaction of the granules is developed within the glacier as the result
of pressure and friction and not that it is derived from the atmosphere, or the bed.1
However, any heat communicated from such a source will make it that much
easier for internal changes to take place. As presented by Chamber lin and
Salisbury, this theory of granular change accounts more satisfactorily for the
glacial phenomena observed than any other, in which no molecular movement of
the solid granules is assumed.

The original idea of plasticity ascribed to glaciers by Rendu and developed
by Forbes is based upon the conception that the molecules of firm ice will yield
continuously to a stress, without producing visible rupture. The stress may be
of the nature of a thrust, or of tension. This theory has been rejected by the most
prominent physicists who have turned their attention to the problem of glacial
motion, because of their unwillingness to admit that ice could possess this and cer-
tain other properties apparently inconsistent with plasticity. The difficulty so far
as rigidity alone is concerned is removed by our knowledge that such substances
as lead, tin, and iron may be made to flow by pressure, under ordinary tempera-
tures. As soon as direct experiments were made to test the plasticity of ice it

1 In a recent "pamphlet entitled, "The" Viscous vs. the Granular Theory of Glacial Motion," Mr. O. W.
Willcox has endeavored to show that the heat developed through pressure or impact of the granules would
be conducted away as rapidly as it is generated (p. 18).

130 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

was found that bars of ice frozen in a mould, or cut from a glacier, may be bent,
elongated, compressed, and twisted, without visible rupture, even when kept
continuously below the freezing temperature. These experiments were made by
Main, McConnel, Kock, and Mxigge and show that solid ice, made up of a collection
of irregular crystals, is decidedly plastic. McConnel calculated that the amount
of extension required of the ice in the Rhone Glacier, because of the more rapid
central movement when compared with its sides, amounted to 0.0029 millimeter
per hour, for each 10 centimeters of length. In his experiments with bars of
glacial ice, but one of the three bars tested showed as small amount as this and
that for only a portion of the experiment. In the case of single crystals they
were found capable of continuous yielding without rupture, providing the pressure
was applied at right angles to the optic axis, the movement appearing to consist
of a sliding between adjacent crystalline plates. When the force of compression,
or tension, was applied parallel with the axis, the result was exceedingly small, or
nil.1 The verification of these results by other investigators leads to the conclu-
sion that ice is capable of showing a certain type of plasticity, although different
from that ascribed to it by Rendu and Forbes. An amorphous plastic substance
yields under a suitable stress in any direction without visible rupture. A crystal-
line substance which maintains its definite molecular arrangement will be
limited in the number of directions in which it may yield. If ice crystallized in
cubes it seems likely that it might have yielded without rupture in three direc-
tions. Had it crystallized in square prisms we may now conceive of a movement
in two directions. In the hexagonal system in which it actually crystallizes
the molecular cohesion measured in the direction of the main axis is of a suffi-
ciently different nature from that at right angles to it to permit of a gliding of the
basal plates without rupture and the destruction of their molecular arrangement.
This is plasticity in a crystalline substance. The experimental results were
obtained with moderate stresses and much below the freezing temperature. In
the case of a glacier under great stress and a temperature near the freezing
point it seems absolutely necessary chat the glacial granules should manifest
this property to a greater or less extent, regardless of the actual mechanics of
glacial movement.

A number of phenomena, noted upon the Canadian glaciers, has convinced
the writer that a certain amount and kind of plasticity is a fundamental property
of glacial ice. The complicated patterns occasionally shown by the blue bands
are such as might arise from plasticity, but not from melting or rotating of the
granules. When similar effects are seen in ordinary rocks they are commonly
referred back to a plastic condition of the matrix. It was noted that when the
difference in the rate of movement between the centre and margins of the glacier
is sufficiently small, there are no marginal crevasses to be seen. This implies that
the ice permits a certain amount of stretching, without visible rupture. If ice
were absolutely incapable of yielding under tension, any appreciable difference be-

>" On the Plasticity of Glacier and other Ice," James C. McConnel, Proc. Roy. Soc. of London, vol. 44,
1888, p. 331; "On the Plasticity of an Ice Crystal, " vol. 49, 1891, p. 323.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

tween the rate of movement of center and sides must open marginal crevasses.
In the rock scorings seen near the nose of the Illecillewaet, the frontal and lateral
grooves about the knobs and trails suggest very strongly a plastic condition
of the ice. In the case of the same glacier, but farther up between the ice and
the left lateral moraine, the ice was seen moving over a knob of bedrock not
fully exposed in 1904. Transverse crevasses formed above while the ice melted
below and there were formed strips of ice, 20 to 25 feet long, supported at either
end. Moving downward these strips were suspended in the air, and in
the course of ten days in September one bar had sagged so as to be very no-
ticeable to the eye, without forming any crevasse large enough to be noticed, or
to permit of the destruction of the ice bar itself. This phenomenon seemed to
indicate that the ice could, to some extent, yield to a tensional stress.

Now that the question of the plasticity of granular ice has been settled by
experiment, what objections are there that may be urged against its application
to the glaciers ? Crevasses and faultings indicate simply that there is a limit to
its plasticity, as indeed there is to the most typically plastic solid. The tendency
to flow must be greater in the basal layers, but it does not necessarily follow that
with the friction of the bed to combat, the velocity here will be greater than or even
equal to that of the upper layers. The hold which glaciers have upon rocks in
their basal layers is probably not a firm one. It seems to the writer that the
chatter-marks, crescentic gouges, the shape and often sudden termination of
the coarse striae, as well as the faceted condition of the boulders in the ground
moraine, all indicate that the glacier was persistent, rather than firm, and that
it very often lost its grip. All of the phenomena of rock scoring, the subglacial
fluting of the ice, the compression of the ice at the base of a slope, the phenomenon
of shearing, and the mounting of reversed slopes prove that portions of the ice
move bodily under the influence of a more or less rigid thrust from behind. This
necessary amount of rigidity in the ice is not inconsistent with the degree of plas-
ticity ascribed to it. When the flow is not sufficiently rapid at any point the
ice must be thrust forward bodily. The most forcible argument against this
modification of the viscous theory of glacial movement is brought forward by
Chamberlin. It would seem that the granules should be distorted noticeably in
the direction of flow. That this distortion is not more apparent may be due to
the complete mechanism of granular growth. It may be disguised by the shear-
ing of the granules in the direction of the basal planes and their later dry union
by the principle of Hagenbach-Bischoff.

9. COLOR OF ICE AND GLACIAL WATER.

The exquisite richness and variety of coloring seen in glaciers and glacial
lakes constantly arouses the wonder and admiration of those privileged to gaze
upon them. The colored photograph fails to reproduce it and it eludes the
brush of even the most skilful artist. An explanation of the cause of this color-
ation can scarcely fail to be of interest. In 1904 a study was made of several

132 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS

of the lakes by means of standard solutions of copper and nickel sulphate and an
instrument devised from a stereoscope. By using measured amounts of the solu-
tions and mixing with pure water it was possible to match the water and to express
its color as an equation, from which the depth and shade of color might be at
any time reproduced. The solutions were prepared by dissolving 30 grams of the
chemically pure salt in 100 cubic centimeters of distilled water and were placed
in a thin glass" cell," with parallel sides, the inside measure of which was
8 millimeters. A reflector was so arranged that the water could be viewed
directly, while the fluid mixture was seen by reflected diffused light. The proper
proportions could then be obtained by experiment. The shade of color changed
somewhat by the condition of the sky, the position of the observer, the time
of day, and the strength of the wind. In the case of Moraine Lake the depth
of blue was too intense in the quiet of the early morning to be matched by
the pure copper sulphate solution, in a cell of the above thickness. A slight
breeze sprang up and lightened the shade sufficiently. The table below gives the
color simply at the time of observation and under the conditions then prevailing.

COLOR OBSERVATIONS UPON ROCKY MOUNTAIN LAKES.

Lake. Date. Time. Proportions in cubic centimeters. Sky.

Lake Louise Aug. 3. 8:45 A-M- 29 c-c- green + 10 c.c. blue + 65 c.c. water. Clear but hazy.

Emerald Lake Aug. 16. 8:00 A.M. 10 c.c. green + 10 c.c. blue + 50 c.c. water. Fair.
Moraine Lake Aug. 10. 10:00 A.M. 3 c.c. green + 30 c.c. blue + 23 c.c. water. Sunny.

As the season advances a marked change occurs in the color of the water of
Lake Louise, there being much more blue in the lake in the early part of the sum-
mer and more green towards the fall. Mr. Robert Campbell informed me that a de-
cided change has occurred in the color of the water of Emerald Lake in the last
few years, it being more of an emerald when he first saw it and now he considers
it more of a turquoise. In discussing this change with Mr. Bell-Smith, the
Canadian artist, I learned that he had observed the same change since 1888, there
being a very noticeable increase in the amount of blue.

Some simple observations and experiments in the field made clear to the
writer the cause of the differences in color of the water of different lakes and the
changes that occur in the same lake. It had been remarked before, but it was left
for Bunsen to demonstrate that absolutely pure water is blue, as seen by trans-
mitted light.1 The colors with the longer wave-lengths, as yellow, orange,

' " Ueber den innern Zusammenhang der pseudovulkanischen Erscheinungen Islands," Annalen der
Chemie und Pharmacie, Bd. LXII, 1847, p. i. Previous to the time of Bunsen the illustrious scientists, Newton
Humboldt, Davy, Arago, and Forbes had speculated upon the problem but with only meager results. Later
Beetz, Tyndall, Bezold, Boas, and Aitkin had more or less completely grasped the idea of selective absorption
and gave us a satisfactory theory of the various modifications of the color of water in nature.

" Ueber die Farbe des Wassers, " von W. Beetz, Annalen der Physik und Chemie, Bd. cxv, 1862, s.
137 zu 147-

The Glaciers of the Alps, chapter 6, Pt. n, Color of Water and Ice, 1860, Tyndall.

Lectures on Light, Tyndall, 1877, p. 35.

Theory of Color, Wilhelm Von Bezold. Translated by Koehler, 1876, pp. 41 and 67.

" On the Color of the Mediterranean and Other Waters," Aitkin, Proc. Roy. Sac. of Edinburgh, xi,
pp. 473 to 4»3-

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. '33

and red, are absorbed if passed through water of sufficient thickness. While of
the colors at the other end of the spectrum, with the short wave-lengths, blue is
the one which water is chiefly able to transmit, violet and green being also trans-
mitted, but less perfectly. Bodies of pure water of a volume sufficient to absorb
the longer waves of light reflected from the bottom, but not so deep as to absorb
it all, will appear blue. This blue is not reflected from the sky, although the con-
dition of the sky will affect the tint. Lakelets in the neVe", such as the one dis-
covered upon Sapphire Col by Mr. Wheeler, are a rich blue; those upon the ice
may be blue, or not, depending upon their freedom from sediment, being liable
to change rapidly. Moraine Lake (plate xxv) owes its exquisite blue color to its
purity and depth. Water in the form of ice possesses still the same power
to transmit the colors with the shorter wave-lengths, violet, indigo, blue and
green, with the preference for blue. If a mixture of these four colors, or of all
the others which compound white light, be passed through a block of pure ice,
of sufficient thickness, none but the blue will emerge. If no light whatever is
being transmitted through either ice, or water, it will look black, or will show
whatever color of light is being reflected from its surface.

From this blue as the fundamental and natural color of ice and water by
transmitted light we meet with many modifications in nature. Finely divided
ice, as snow and ne"ve", presents innumerable reflecting surfaces from which light
of any and all colors is sent to the eye. The same is true of water lashed into
foam, or in any finely divided state, as fog, cloud, or condensed steam. In
ordinary light these forms of water and ice appear white, but in the gorgeous
colors of the sunrise and sunsets they transmit to the eye by reflection the greatest
variety of color. Ne"ve" begins to show a bluish tinge as soon as the transmitted
light begins to predominate over that which is being reflected. The water which
issues from the glacier is generally charged with sediment, and if this is much
in amount, its color will determine the color of the water of the drainage brook.
It generally appears a milky, or creamy, white but may be a dirty gray. With
the deposition of the coarser yellowish sediment and the retention of the very
finest, if the volume of water is considerable, the water assumes a greenish tinge,
as seen in the Asulkan and Illecillewaet streams. With the loss of this sediment
the stream acquires more and more of its natural blue. When this glacial sedi-
ment is introduced into a lake in sufficient, but not too great, quantity the water
becomes charged with finely divided rock particles in suspension. These
particles are able to reflect the longer waves of the spectrum, particularly yellow,
but also the closely related green and orange, while they very effectually cut out
the shorter wave-lengths giving rise to violet, indigo, and blue. The result is that
there are introduced into the water innumerable reflecting faces which are capa-
ble of sending to the eye only those colors that lie at the centre of the spectrum
and towards the red end. But the only light that is available for reflection is
that which has already passed through the water once and had its yellow, orange,
and red to a greater or less extent filtered out. That which remains to be
reflected by the foreign particles will pass again through the water and will suffer

134 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS.

still further absorption. Of the blue and green rays, which make their way
readily through the water, only the green rays are reflected by the particles and
these alone reach the eye. The quantity of light is thus much reduced in amount,
but with the sun shining upon the lake the green becomes quite vivid if the water
carries the requisite amount and kind of foreign particles. If the foreign particles
were pure white they would be capable of reflecting all colors equally well, and
the rich blue of the water would be brought out in perfection.

In the case of Lake Louise, we have a variable amount of sediment entering the
lake during the year, and consequently a seasonal variation in the color of its water.
Witlrthe very slack drainage during the winter, the sediment, in considerable part,
settles and the water in the spring shows more of its own blue color. With the in-
creased activity and melting of the Victoria Glacier the supply of sediment deliv-
ered to the glacier increases as the summer advances and the water becomes a richer
and richer green. About the delta at the head of the lake the sediment is so abun-
dant and lies so near the surface that the water is unable to absorb the yellow, and
the color of the sediment itself is seen with little or no modification. In the case
of the change in color noted for Emerald Lake we may infer that the drainage
stream at its head has been carrying less sediment than formerly. It is not
improbable that the diminished activity of the inlet may be connected with the
stage of diminished precipitation recently closed and that the rich emerald
green of the lake in 1888 and earlier was connected with the stage of increased
precipitation, which is supposed to have closed in the early 8o's. With an
increase now in the average annual precipitation it will be interesting to see
whether the lake returns to its former shade of color.

The introduction of green or yellow, organic solid matter, animal or vegetable,
into the body of water would have the same effect. If the lake is sufficiently
shallow and the bottom covered with green vegetation, or yellow sediment,
the water will not be able in the short transmission to cut out the green and this
color may appear in lakes of water free from sediment. About the margin of
Moraine Lake the water has a greenish cast for this reason. Organic matters
in solution quite generally give water a yellowish to brownish color, as seen
naturally in bogs and artificially in tea, coffee, cider, beer, etc. With the above
principles in mind one may infer from a glimpse of a distant lake the condition
of the water and the state of activity of glaciers whose drainage streams empty
into it. If upon rounding the shoulder of Mt. Temple, in entering the Valley of
Ten Peaks, the first glimpse of Moraine Lake showed that a rich green had been
substituted for its superb blue, one might safely infer that the Wenkchemna
Glacier had begun to erode its bed although a neighboring rock slide might
temporarily give such a result. Even the names of lakes in a glacial region
are suggestive of the amount of glacial activity in the valley, such as Sap-
phire Lake, Turquoise Lake, and Emerald Lake.

The same principles of coloring apply to ice as well as to water. When highly
charged with sediment of any particular color, its own natural color is obscured
and the color of the sediment is sent to the eye with little or no modification.

GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 135

If, however, yellowish sediment is distributed through the ice in proper propor-
tions the case is identical with that discussed for water, and the ice assumes a
greenish cast. In the case of the solid ice the sediment can not be assorted and
evenly distributed as in the case of water and hence only greenish patches and
streaks occur, just where conditions are favorable. It has not seemed appropriate
to any one to apply the names "emerald," or "turquoise," to a glacier while
"sapphire" would not be very distinctive. By applying the principles here
set forth we may account for the coloration of glaciers, the lakes in their neigh-
borhood; the gorgeous pools of blue and green water in the Yellowstone Park
and similar regions, the blue color of the ocean and the seas, the green and final
yellow strip as we approach the shore and such phenomena as the blue and green
grottoes of the island of Capri. No one has yet satisfactorily explained how a
considerable body of pure water may appear limpid.

THE END.

14 DAY USE

RETURN TO DESK FROM WHICH BORROWED |

LOAN DEPT.

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

U-

-4&D4MTO-3 .— 33-3-f¥4-£

I IN

HER cm.

N\.

ft f el LD 2

MAY 5J978

ifK r.l» MAY 5

BIC£3J9za- J63

- —

CIR. DEC 1 2 1973

-WTTft

(H241slO)476B

OAN2

General Library
Universiry of California U
Berkeley I"

U C BERKELEY LIBRARIES

I

5 €9

THE UNIVERSITY OF CALIFORNIA LIBRARY