USGS Logo Geological Survey Professional Paper 1180
Grinnell and Sperry Glaciers, Glacier National Park, Montana: A Record of Vanishing Ice



Sperry Glacier, in Glacier National Park, Mont., is 9 miles south and slightly west of Grinnell Glacier and 25 miles south of the international boundary between the United States and Canada, at lat 48°37'18"N., long 113°45'24"W. (fig. 1). It occupies a cirque on the northwest slope of Gunsight Mountain on the Pacific side of the Continental Divide, which follows the crest of Gunsight Mountain to its highest point and then bends sharply to the southeast.

The glacier is slightly less than 10 miles, by trail, from the McDonald Hotel near the upper end of Lake McDonald. It is accessible by horse travel for 9 miles, to an approximate elevation of 7,800 feet, and from this point by a well-marked foot trail. Sperry Chalets are 6.5 miles from the McDonald Hotel and 3 miles from the glacier, near the headwaters of Sprague Creek.


Sperry Glacier was named in honor of Lyman Beecher Sperry, M.D., of Oberlin College, Oberlin, Ohio. Sperry (fig. 11) camped with several others in the Avalanche Basin in 1894, and deduced from the milkiness of the water in Avalanche Lake that it came from a glacier. He, with several members of the party, succeeded in scaling the cliffs at the head of Avalanche Basin and reached a vantage point on the side of the Little Matterhorn, a small isolated peak in the saddle between Edwards Mountain and Mount Brown (fig. 1), from which the glacier was visible. Albert L. Sperry, a nephew of Dr. Sperry's, was a member of the party in 1894 and described the impression gained from their first view of the glacier as follows (Sperry, 1938, p. 51):

While standing upon that peak overlooking the terrain above the rim wall, we got the thrill of thrills, for there lay the glacier, shriveled and shrunken from its former size, almost senile, with its back against the mountain walls to the east of it, putting up its last fight for life, It was still what seemed to be a lusty giant, but it was dying, dying, dying, every score of years and as it receded, it was spewing at its mouth the accumulations buried within its bosom for centuries.

FIGURE 11.—Lyman Beecher Sperry (1841-1923), physician lecturer, and author, was the discoverer of Sperry Glacier. Sperry Glacier, Sperry Glacier Basin, and Sperry Chalets are named in his honor. During the summers 1894-1906, Dr. Sperry camped, climbed, and explored in the Montana Rockies, particularly in the area set aside by the Congress in 1910 as Glacier National Park. This photograph, taken in 1904 at some unspecified location in the Park area, was by George H. Dean, student of Dr. Sperry's, who accompanied him on the 1904 trip. Photograph 6274 from the archives of Glacier National Park, courtesy of the National Park Service.

Dr. Sperry, with a party including his nephew, returned to Avalanche Basin in 1895. They scaled the rim wall by following a ravine at its east end (figs. 12, 13), reached Mary Baker Lake and went on to the front of the glacier but did not go onto it.

FIGURE 12.—Sperry Glacier viewed from Avalanche Basin in 1894 or 1895 (Sperry, 1938, p. 42). Avalanche Lake is in lower foreground. Terminus of glacier is at head of central stream cascading down the cliff.

FIGURE 13.—Crevasses on Sperry Glacier, 1897 (Sperry, 1938, p. 130). Crevasses of this magnitude were not observed during the period covered by the present report.

The glacier was actually reached first by Dr. Sperry and his party in 1897. Their route was the same general route of the present trail. It took them near the present site of the Sperry Chalets and through the saddle between Gunsight Mountain and Edwards Mountain.


The changes that have occurred on Sperry Glacier since its discovery can be visualized by reference to selected photographs. The earliest known photographs were published in the book, "Avalanche," by Albert L. Sperry (1938) (this report, figs. 12, 13).

The earliest available photograph showing nearly all the glacier was taken by Francois Matthes in 1901 during the mapping of the Chief Mountain quadrangle (fig. 14).

FIGURE 14.—Sperry Glacier, October 18, 1901, by Francois Matthes. This earliest available photograph showing nearly the entire glacier was presumably taken from a point on the Continental Divide about 1 mile south of Hidden Lake. Surface details of the glacier are obscured by a light snowfall but its extent then is clearly evident. The end of the horse trail and the beginning of the foot trail to the glacier are just beyond the pass in the upper central part of the view. Lakes that are shown on recent maps did not exist at the time of this photograph; they formed as the glacier retreated. Photograph 7166 from Glacier National Park files, courtesy of the National Park Service.

Comparative 1897 and 1969 views are shown in figure 15, and 1913 and 1956 views in figures 16 and 17.

FIGURE 15.—Comparative views of terminal area of Sperry Glacier, in 1897 and 1969. A, 1897 (Sperry, 1938, p. 132); B, August 23, 1969 (W.A. Blenkarm, USGS). In the earlier view the glacier extends to the moraines whereas in the later view no part of the glacier is visible. The position of the 1969 terminus is a considerable distance to the right of both views.

FIGURE 16.—Panoramic views of Sperry Glacier from the side of the Little Matterhorn in 1913 and 1956, showing terminal recession and overall shrinkage during the 43-year interval. A, August 15, 1913, photographs by W.C. Alden (710, 711, 712); B, August 22, 1956, photographs by D.H. Robinson, National Park Service. The 1956 view also shows that the ridge extending downward from the high point of Gunsight Mountain was entirely bare. The snow-covered areas in the cirque to the right of this ridge are probably snowfields rather than glacier ice. The pronounced decrease in the glacier area on the side of Edwards Mountain is also apparent. (click on image for a PDF version)

FIGURE 17.—1913 and 1956 views of terminus and terminal moraines of Sperry Glacier. A, August 16, 1913, photographed by W.C. Alden (716); B, August 22, 1956, photographs by D.H. Robinson, National Park Service. Head of Avalanche Basin in in left center edge of views. Terminal recession during the 43-year interval is evident. Alden (1914, p. 16) noted that the pond in the lower foreground of the views was reddish in color, whereas Robinson reported in 1956 that it was milky white. The area now occupied by lakes shown on recent maps was concealed by the glacier in 1913.


In 1901 the area of Sperry Glacier, as shown on the Chief Mountain quadrangle map, was about 800 acres. Its western edge was almost in the saddle between Gunsight Mountain and Edwards Mountain (fig. 1). This edge is now two-thirds mile from the saddle. The lower extremity of the glacier at the western edge extended into the broad notch between Edwards Mountain and the Little Matterhorn, a small isolated peak in the saddle area between Edwards Mountain and Mount Brown (fig. 1). Drainage from this part of the glacier flowed through this notch into the head of Synder Creek as late as 1913 (Alden, 1914) but probably did not continue for many years thereafter.

In 1901 the terminus was defined by a series of moraines (pl. 2) extending across the valley. The east and upper edges were defined by the steep walls of the cirque.

The area of Sperry Glacier has decreased since it was first mapped in 1901:

YearSource of data AreaReduction since previous measurement
TotalAverage per year

1901Chief Mountain quadrangle map800------
1938Sperry Glacier map (Dyson)39041011.1
1946Sperry Glacier map (Dyson)330607.4
1950Sperry Glacier map (USGS)384256.2
1960Sperry Glacier map (USGS)207181.8

Much of the decrease in area between 1901 and 1938 was due to the disappearance of 75-100 acres of ice from the western part of the cirque. Dyson's 1946 mapping covered only the terminal area. The difference between the 1938 and 1946 values, therefore, represents primarily the decrease in area as a result of terminal recession. Changes in area of the upper parts of the glacier were probably insignificant, owing to the steep walls of the cirque. The rate of shrinkage slowed markedly after about the mid-1940's. The average annual decrease in area was 10.5 acres during the 45 years 1901-46, whereas it was only 3 acres from 1946 to 1960. The slower rate of loss in recent years can be attributable partly to a slight short-term warming, but, more significant, the remnant glacier is probably approaching hydrologic equilibrium with the long-term climatic trend.

Dyson (1948, p. 97) described the Sperry Glacier as being in 1938 "the largest in Glacier National Park and, with the possible exception of one or more of the Dinwoody Glaciers in Wyoming's Wind River Range, the largest in the Rocky Mountains south of the Canadian boundary." This may have been true in 1938, but owing to Sperry Glacier's more rapid rate of shrinkage, it has lost that distinction to Grinnell Glacier.


Dyson (1948, p. 97) estimated, by comparing 1913 photographs with his detailed mapping, that recession of the terminus from 1913 to 1938 totaled 1,533 feet measured along 5,700 feet of the front—an average annual recession of 61 feet.

Markers for recession measurements were set by the Park Naturalist in 1931 but no data were obtained during 1932-34. New markers were set in 1935 and the recession as reported by the Park Naturalist for the 10-year period 1935-45 was 641 feet, or an average of 64 feet per year. This matches closely Dyson's estimate of 61 feet annually for 1913-38. Dyson's 1938 mapping did not show the markers used for the recession measurements, so a correlation with the mapping was not possible. The 1938 mapping was, however, the first detailed mapping of the terminus since the 1901 mapping of the Chief Mountain quadrangle.

The method of determining terminal recession by planetable mapping, as described for the Grinnell Glacier, was begun for the Sperry Glacier in 1945. The terminus was remapped, in whole or in part, in 1947, 1949, 1950, 1952, 1956, 1961, 1966, and 1969. The total and average annual rates of recession for about the central half-mile section of the terminus for selected periods are tabulated below.

Recession (ft)

Total for


Below an elevation of about 7,500 feet, the period of record was characterized by a continual recession of the terminus and a lowering of the surface; above this elevation, however, the surface elevation of the ice apparently increased.

Sperry Glacier has not, for several centuries, extended far beyond its present limits. This is evident from G.M. Baden's 1960 age determinations (written commun., 1961) on three old trees located north and west of the glacier. The locations and ages of the three trees are as follows:

From BM 7375 (fig. 18)
Age (yr)
of tree,
in 1960
BearingDistance (ft)

N. 60° W.2,8507,120350+
N. 22° W.4,9256,560226
N. 23° W.4,5256,640190


The advance of the glacier during a 24-hour period was measured by Alden (1914, p. 15) in August 1913 within four ice caves at the front of the glacier. Movement varied from 1/4 to 3/4 inch; the rate of movement apparently reflected the prevailing temperature that day—rapid on a sunny warm day, and slow on a cold blustery day.

So far as I can determine, further information on movement was not obtained until 1949 when M.E. Beatty, Park Naturalist, and I mapped the locations of 5 prominent rocks and marked 4 of them with identifying numbers 49-1 through 49-4. The fifth rock was marked in 1963. Their locations were periodically redetermined; and all five were redetermined in 1969, giving a 20-year record of movement from 5 points (pl. 2, table 15).

TABLE 15.—Movement of marked rocks on Sperry Glacier

[Locations shown on plate 2]

Location No. Period Movement (ft)
AverageTotal annual



Movement in the central part of the glacier averaged about 13 feet per year, as recorded by rocks 49-1 through 49-4. Movement closer to the east edge of the glacier averaged 17 feet per year, as recorded from the unmarked rock. Rates of movement also increased toward the west edge of the glacier, as indicated by the 3-year record for rocks 66-1 and 66-2 and ablation stake 1-63, and by the annual results for stakes 1-61, 1-65, and 7-65. Stake 7-65 was the lower part of an ice auger lost in 1965 but exposed in 1966.

Measurements of stakes 2-61 and 3-61, at higher elevations on the glacier, indicated considerably greater rates of movement there than at lower elevations. Even though appreciable errors may have been introduced by difficulties in field measurement, the measurements confirm the previous finding that movement was greater near the glacier's west edge than in the central part.



Changes in surface elevation of Sperry Glacier have been recorded by periodic measurements of four profiles (pl. 2). Profile A—A' extends across the entire glacier, approximately at right angles to the direction; profiles B3B', C3C', and D3D' are roughly parallel to the direction of flow. The mean elevations of segments of the transverse profile and profile B3B' were determined for comparative purposes (tables 16, 17). In comparing the observations, the reader should bear in mind the differing dates of measurement; for example, the 1958 and 1959 measurements were made in mid-August, whereas the 1957 and 1961 measurements were made later, in mid-September.

TABLE 16.—Mean elevations, in feet, of segments of profile A—A', Sperry Glacier, on specific dates

[Elevations are above assumed datum for glacier surveys. Profile shown on plate 2. Distances measured from reference point, planetable bench mark 7699]

Distance (ft)
100—1,1001,100—2,100 2,100—3,100100—3,100
Mean elevation (ft) of segment



1966Aug.24, 257,608.17,531.57,592,07,577,2

Net change since 1949+8.6-4.5-9.2-1.7

TABLE 17.—Mean elevations, in feet, of segments of profile B3B', Sperry Glacier, on specific dates

[Elevations are above assumed datum for glacier surveys. Distances measured upglacier and downglacier from intersection of profile B3B' with profile A3A' (pl. 2), 1,965 ft from reference point, planetable bench mark 7699. ---, no data]

Downglacier from intersection
Upglacier from intersection
Distance (ft)1,700-1,5001,500-1,000 1,000-500500-00-500500-800800-1,000
Mean elevation ( ft) of segment




The transverse profile (A—A') is slightly more than 3,100 feet long. The profile shows a fairly uniform downward slope from the west edge for a distance of 1,700-1,800 feet, then a rise to the top of an ice ridge about 250 feet from the east edge. A depression occurs between the ridge and the east edge of the glacier.

Plate 2 shows that the 1969 ice surface was at a higher elevation than in 1949 for a distance of 800 feet from the west edge of the glacier, with a maximum difference of 20 feet. In most of the east half of the profile the 1969 ice surface was at a lower elevation than in 1949, with a maximum difference of 20 feet. Along the 3,000-foot section of the transverse profile (excluding a short distance at each end), the net change was a lowering of 1.7 feet (table 16). The fact that one section of the glacier showed an increase in surface elevation while another section showed a decrease does not necessarily indicate variation in amount of snowfall over the area. The differences in snow accumulation on various sections of the glacier probably reflect differences in wind patterns, which are influenced by the rugged terrain and the consequent drifting, rather than reflecting varying elevations.

Profile B3B' (pl. 2) is almost at right angles to the transverse profile, intersecting it at a point near mid-glacier. From the intersection, measurements extended upglacier 800-1,000 feet and downglacier to the terminus.

Along profile B3B' (table 17) upglacier from the intersection point both positive and negative changes in surface elevation have been measured. The surface was higher in 1969 than in 1950.

Downglacier from the intersection point the surface elevation has generally shown a continual lowering (table 17). During the 19-year period 1950-69 the terminus receded 725 feet along the profile alinement. In 1950, the ice at the location of the terminus in 1969 was 115 feet thick.

Profile C3C' (pl. 2) near the west edge of the glacier originates at bench mark 7375. It was first measured in 1947; then it crossed a prominent ice ridge that stood 40 feet higher than the depression or trough just upglacier. The successive profiles along this line (pl. 2) show a continual lowering of the ridge, which by 1969 had completely disappeared. The ridge crest in 1947 was 90 feet above the corresponding point in 1969. Upglacier (southeast) from the former trough or depression the ice surface has remained much the same since 1950.

Profile D3D', originating at the same point as the transverse profile, bench mark 7699, was first measured in 1958. During the 11 years the terminal position advanced more than 200 feet and the mean surface elevation rose about 20 feet. The increase in surface elevation recorded at this profile corresponds with the increase at the transverse profile's southwest half, which was higher in 1969 than in 1950. The area crossed by profile D3D' has become more crevassed in recent years as a result of the renewed ice activity.

Along the transverse profile (A3A') surface elevation fluctuated, whereas downglacier from it there was a continual shrinkage, as well as pronounced terminal recession. Upglacier there was some increase in surface elevation, as shown by the results at profile B3B' from 1950 to 1969 and at profile D3D' in 1958 and 1969. This increase in surface elevation—glacier thickening—is also indicated by a comparison of the 1950 and 1960 maps. The 7,700- to 8,100-foot contours were farther downglacier in 1960 than in 1950. The 7,600- and the 8,300-8,600-foot contours were in virtually the same positions on the two maps.

The great thickness of glacier ice in 1913 is evident in Alden's photograph (fig. 18) and his related comments. In describing the structure of the glacier he (1914, p. 14) referred to the stratification as follows:

"The best view * * * of the bedded structure of the glacier is to be had where the east side of the glacier rounds a salient of the cirque wall at a point about two-fifths of a mile south-southeast of the front of the most easterly marginal lobe of the glacier. On the northwest side of the salient instead of the ice crowding against the rock slope there is a great chasm, or moat * * *, one side of which is formed by the rock wall. The other is a smooth, curving wall of stratified ice, 150 feet or more in height * * *."

FIGURE 18.— Moat at east side of Sperry Glacier, showing stratified ice wall, August 16, 1913 (Alden, 1914, fig. 3).

The cirque-wall salient mentioned by Alden appears, on aerial photographs, to be roughly opposite the 1950 location of the terminus. The ice wall described and illustrated by Alden (1914) was no longer evident in 1944 when I first observed this glacier. From Alden's description, one can infer that the ice surface in 1944 was approximately 150 feet lower than in 1913. This would represent an average annual decrease in surface elevation of the glacier in this particular area of about 5 feet.


Ablation measurements on Sperry Glacier since first obtained in 1961 are summarized in tables 18 and 19. (The procedure for setting the stakes was the same as at Grinnell Glacier, see page 14.) The ablation stakes (pl. 2) were usually placed during the latter part of July or early August and readings were made in late August or early September, covering only a part of the ablation season. Ablation was considerable during May, June, and early July. The values in table 18 provide a basis for estimating total ablation during the melting season. Such estimates must, of course, take into consideration temperatures prevailing during the estimating period.

TABLE 18,—Seasonal measurements of ablation at Sperry Glacier for selected years

[Location of stakes shown on plate 2.---, no data. Stakes 7 and 9-12 emplaced in 1965. No measurements in 1964]

Period of
Period of

1July 24—Sept. 145210.1July 21—Aug. 26367.3
2July 24—Sept. 14527.6July 21—Aug. 26366.2
3July 24—Sept. 14527.6July 21—Aug. 26367.6
3A---------July 21—Aug. 26367.0

4July 24—Sept. 14528.9July 21—Aug. 26367.6
5July 24—Sept. 14529.2---------
6July 24—Sept. 145210.8---------

1July 31—Sept. 8397.4Aug. 7—Oct. 13675.3
2July 31—Sept. 8396.8---------
3Aug. 1—Oct. 18789.8---------
3AAug. 2—Oct. 18776.4---------

3BAug. 2—Oct. 18778.4---------
4Aug. 2—Oct. 187710.2---------
5Aug. 4—Oct. 187511.7---------
6Aug. 4—Oct. 187410.4---------

7---------Aug. 7—Oct. 13675.5+
9---------Aug. 7—Oct. 13674.6
10---------Aug. 7—Oct. 13674.6
11---------Aug. 7—Oct. 13674.5
12---------Aug. 7—Oct. 13674.2

TABLE 19.—Annual ablation measurements at Sperry Glacier

(Location of stakes shown on plate 2)

Period of

1Sept. 14, 1961—Aug. 8, 19623281.7

July 31, 1963—Aug. 24, 19661,12012.6

July 23, 1964—Aug. 24, 19667625.8

Aug. 6, 1965—Aug. 24, 19663839.5

Aug. 24, 1966—Sept. 8, 19673808.7
2July 21, 1962—Sept. 8, 19671,8754.0

3Aug. 7, 1965—Aug. 25, 1966383-1.9

Aug. 25, 1966—Sept. 8, 19673797.9
4Sept. 14, 1961—Aug. 26, 1962346.9

July 31, 1963—Aug. 25, 19661,1219.5

Aug. 7, 1965—Sept. 8, 1967762-1,4

6July 23, 1964—Aug. 25, 19667635.4
8Aug. 7, 1965—Aug. 26, 19663848.1
9Aug. 7, 1965—Aug. 25, 1966383-1.9

Aug. 25, 1966—Sept. 8, 19673797.9
10Aug. 7, 1965—Sept. 8, 19677626.9
11Aug. 7, 1965—Sept. 8, 19677623.1

Table 1 shows July and August mean temperatures at Sperry Chalets, about 1,000-1,500 feet lower than the glacier. If the usual decrease in temperature with increasing elevation is applicable, temperatures over the glacier probably were about 5°F less than at Sperry Chalets.

<<< Previous <<< Contents >>> Next >>>

Last Updated: 08-Jul-2008