4th Report Michigan Academy of Science pp. 168-173 (1904)

On January 27 of the present year (1902) a shower of dust fell over a wide area in Western Michigan. The dust was mingled with snow and the mixture was commonly mentioned as yellow snow. The fall of the dust continued during a considerable part of the afternoon,—three to seven hours according to observers at different stations.

The fall occurred during the subsidence of the anticyclone following a storm of great energy that crossed the continent during several days preceding. The storm moved from northern California southeastward to Kansas, thence curving northeastward to Lake Superior. During and following this storm strong winds were prevalent over a wide region that came under its influence, although the winds were not exceptionally high. On the date of the dust fall and the preceding day the winds over the lakes and Mississippi valley were westerly.

Along the coast the dust is known to have fallen as far south as Holland and at Hart northward, indicating an extreme width of at least sixty miles. It is known to have extended inland to Sparta, about thirty miles from the lake shore. The territory bounded by a line slightly outlying these points is not less than 1,600 square miles in area. It appears that the precipitation was somewhat variable in quantity at different places and there are some indications that the amount was greater twenty to thirty miles inland than at Muskegon.

Naturally there were many speculations as to the real nature of the phenomenon. The dust was variously mentioned in public print as (1) local surface material, especially from the dunes along Lake Michigan; (2) Volcanic dust; (3) Meteoric matter; (4) Earthy material raised from the surface beyond the lake. The first supposition was easily disproved by microscopical measurements of the grains, which showed its extreme fineness as compared with local soil particles. Moreover, the fall was observed on Lake Michigan windward from the land. The presence of considerable organic matter showed that it could not have come from a volcano or from outer space. Dr. A. C. Lane, State Geologist, pronounced the dust identical with the loess soil of the Mississippi and Missouri river valleys. Mr. W. A. Orton of the National Bureau of Plant Industry called it fine silt. Thus the problem was narrowed down to finding the exact source of the material and how it could have been raised and transported so great a distance under conditions that did not seem exceptional.

Some attempts were made at collecting data from which an approximate estimate of the total deposit could be formed. Snow containing the dust was taken from measured spaces and the dry residue weighed with results as follows:

Number Place of collection Collector Area in
square feet
Dry weight
in grams
Weight per
square foot
1 Muskegon C. D. McLouth 1 1.855 1.855g  
2 Ravenna C. E. Alberts 1 3.484 3.464g  
3 Muskegon C. D. McLouth 4 3.417 .854g  
  Average per square foot 2.058g  

On the basis of the average found the precipitation on a square mile would be about six and one-third tons; on the total area as indicated above the amount would be about ten thousand tons.

I also weighed two residues which, though not affording accurate data, are significant.

Mr. G. A. Rumsey, of Slocum, about twenty miles due east from Muskegon, filled a common drinking glass with the snow and obtained there-from a dry residue that weighed 0.29 grams.

Ray Dawes, a pupil in Muskegon high school, collected a large panful of the snow found clinging to trunks of standing trees on the windward side. This melted to three quarts of water and yielded sixty-five grams, approximate weight.

The following notes are quoted from some of the correspondents to whom samples and inquiries were sent.

C. E. Alberts Ravenna.—"The dust almost spoiled the sleighing."

T. L. Coates, Sparta.—"There are several pounds of it (the dust) in my cistern."

C. E. Ruthruff, Montague.—"That deposited here was coarser than the sample you sent, and had a decidedly red color." (Statement not verified by a specimen).

Vernon G. Mayo, Newaygo.—"On an inclined walk usually ashed to prevent slipping, it was abundantly deposited, so no ashes were needed."

D. E. Austin, Blue Lake Township, near Whitehall.—"The sand or dust storm was of seven or eight hours duration. * * * The dust coming in swirls was deposited at right angles to the course of the storm, in some places scarcely a trace, then again sufficient to check the motion of a sleigh. * * * In breaks of the snow storm the dust would appear like great yellow smoke columns above the level of the main storm clouds."

In physical forum the dried material resembles flour. The particles, which are generally angular and irregular, range in diameter from .005 to .00005 of an inch, few grains reaching the larger dimension. Its color is a deep shade of yellow-orange. It sinks readily in water and is semi-pasty when wet.

In composition the material is chiefly silica. About five per cent, according to Mr. 0m-ton, is organic matter, such as plant hairs, pieces of wood fiber, desmids, etc.

As is well known, dust storms are common in many parts of the world,

In Michigan they seem to be rare but are not unknown. Mr. L. L. Coates informs me, on authority of Mr. George H. Porter, of Sparta, that a similar storm occurred at that place in 1873. On February 18, 1896, Mr. Herman P. Thomas of Cassopolis witnessed a dustfall there. He weighed a collection from a measured space from which he computed 245 pounds per acre, a quantity greatly exceeding that indicated by measurements in the recent instance.

The Monthly Weather Review of January, 1895, contains a detailed account of a dust storm in parts of Indiana and Kentucky with results of careful analyses of numerous samples of the material.

In all these instances the material seems to have been nearly or quite identical with that of the recent storm.

1. A map showing approximately the extent of the dust shower in western Michigan, January 27, 1902.
Places at which the dust is known to have fallen are marked with x.
Extent from north to south along the lake shore about 60 miles.
Greatest extent eastward from shore about 30 miles.
Average amount of dustfall per square foot based on three measurements 2.058 grams.
Estimated amount of dust deposited on one square mile, 6.33 tons.
Amount computed for area of 1,600 square miles, about 10,000 tons.

In attempting to explain the raising and transporting for so great a distance of this mass of soil all readily available data were, acquired relating to (1) wind directions and velocities, (2) distribution of loess, (3) snow covering in the loess regions. The officials of the U. S. Weather Bureau both at Washington and at the Grand Haven station kindly furnished an abundance of meteorological data, by maps and otherwise. These were diligently studied to find indications of a possible bare spot of loess where the prevailing wind could have done the work.

2. Map showing wind directions and velocities in relation to loess soil regions January 26, 1902. Data taken from reports of U. S. Weather Bureau.
Snow covering was essentially as shown on map of January 27.
The loess is shaded.

The known extent of loess in the United States is in comparatively narrow belts along the rivers Mississippi, Missouri and Ohio. Its nearest approach to Grand Haven is in southern Wisconsin, distant 150 miles. It extends westward across Missouri and southern Iowa beyond Omaha and Sioux City, more than 500 miles from the east shore of Lake Michigan.

3. Map showing wind directions and velocities, also approximate extent of snow covering, in relation to loess soil regions January 25, 1902.
Meteorological data taken from reports of U. S. Weather Bureau.
Approximate boundary of the loess is shown by dotted line.


4. Map showing wind directions and velocities, also extent of snow covering, in relation to loess soil regions, January 27, 1902.
Meteorological data taken from reports of weather Bureau.
The loess is shaded.

On January 20, just one week before the dustfall, the loess region, except in the Ohio valley was mostly under a thin cover of snow. On the date of the shower the area of snow had been greatly extended, especially westward, and the depth considerably increased. Much of' the loess was then under six inches and more of snow, but the tract nearest to us in southern Wisconsin was outside the three inch limit and some of it may have been practically bare. In middle Nebraska, South Dakota and in southern and western Illinois the snow covering was also light and in these regions there may have been some exposed soil. These conditions had remained essentially unchanged during 48 hours or more preceding the dustfall. It remained then to choose whether the dust had probably been raised in remote parts and its long transit made longer in time and space by shifting winds or whether it had been transported from the nearest locality in the course of a few hours.

On January 25, when the storm center was far west, the winds were light and mostly from eastward over the Mississippi and lower Missouri valleys. The snow cover had not been materially increased since the twentieth but there was not sufficient energy in the winds to disturb an exposed surface of the lightest soil. On the following morning, after most of the new snow had fallen, the winds were decidedly stronger and mostly front westward. The highest velocity reported by the Weather Bureau was 36 miles per hour at Duluth. Other significant records were: St. Paul, 20; Milwaukee, 14; Des Moines, 12; Davenport, 12; Sioux City, 20; Omaha, 18; Chicago, 18; Kansas City, 18. These velocities are not high but they indicate a general movement of air currents of considerable force.

On the twenty-seventh the general direction of the winds was from the west but velocities had become much lower west of the Mississippi. In the lake region velocities remained about as on the previous date, as follows: Grand Haven, 26; Milwaukee, 12; Green Bay, 14; Chicago, 16. These it must be understood were not maximum velocities but velocities at moment of taking data for the daily report. At Grand Haven on this date the maximum 1-minute record was 42 miles per hour at 9:17 a. m. and the maximum 5-minute record was 37 miles per hour at noon. Quite probably there was a similar rise in velocities above the morning records at stations on the vest side of Lake Michigan, and these strong winds may have caught up the material in southern Wisconsin during the forenoon and swept it across the lake at 50 miles per hour in the upper cloud regions.

It is plain that the evidence I have been able to accumulate up to time of writing is quite insufficient to make a satisfactory demonstration. At first thought it seemed that only a spiral wind of tornado violence could have lifted such a mass high in air where it could he sustained and carried on air currents that probably did not exceed a velocity of 50 miles an hour, but no disturbance of that nature is indicated in any of the weather reports. The dust was raised and carried by winds that are frequently exceeded in violence and equalled in constancy. It would seem therefore that we should have about twenty dust showers each year rather than one in about twenty years.

Besides persons and institutions named before I am indebted to Mr. Frank Leverett of the U. S. Geological Survey for important information on extent of the loess, also to numerous correspondents through whom I have been able to approximate the limits of the fall.

Muskegon, March 26, 1902.