By T.A. Jaggar, Jr., Ph.D.

Cases of asphyxiation by gas have been very frequently reported of late years, and we commonly associate with such reports the idea of a second-rate hotel and an unsophisticated countryman who blows out the gas. Such incidents we connect with the supercivilization of the nineteenth century, but it is none the less true that Nature furnishes similar accidents, and that in regions far remote from the haunts of men. In the heart of the Rocky Mountains of Wyoming, unknown to either the tourist or the trapper, there is a natural hostelry for the wild inhabitants of the forest, where, with food, drink, and shelter all in sight, the poor creatures are tempted one after another into a bath of invisible poisonous vapor, where they sink down to add their bones to the fossil records of an interminable list of similar tragedies, dating back to a period long preceding the records of human history.

It was the writer's privilege, as a member of the expedition of the United States Geological Survey of the Yellowstone Park, under the direction of Mr. Arnold Hague, to visit and for the first time to photograph this remarkable locality. A similar visit was last made by members of the survey in the summer of 1888, and an account of the discovery of Death Gulch was published in Science (February 15, 1889) under the title A Deadly Gas Spring in the Yellowstone Park, by Mr. Walter Harvey Weed. The following extracts from Mr. Weed's paper indicate concisely the general character of the gulch, and the description of the death-trap as it then appeared offers interesting material for comparison with its condition as observed in the summer of 1897.

Death Gulch is a small and gloomy ravine in the northeast corner of the Yellowstone National Park. "In this region the lavas which fill the ancient basin of the park rest upon the flanks of mountains formed of fragmentary volcanic ejecta, ... while the hydrothermal forces of the central portion of the park show but feeble manifestations of their energy in the almost extinct hot-spring areas of Soda Butte, Lamar River, Cache Creek, and Miller Creek." Although hot water no longer flows from these vents, "gaseous emanations are now given off in considerable volume." On Cache Creek, about two miles above its confluence with Lamar River, are deposits of altered and crystalline travertine, with pools in the creek violently effervescing locally. This is due to the copious emission of gas. Above these deposits "the creek cuts into a bank of sulphur and gravel cemented by this material, and a few yards beyond is the dÉbouchure of a small lateral gully coming down from the mountainside. In its bottom is a small stream of clear and cold water, sour with sulphuric acid, and flowing down a narrow and steep channel cut in beds of dark-gray volcanic tuff. Ascending this gulch, the sides, closing together, become very steep slopes of white, decomposed rock.... The only springs now flowing are small oozes of water issuing from the base of these slopes, or from the channel bed, forming a thick, creamy, white deposit about the vents, and covering the stream bed. This deposit consists largely of sulphate of alumina.... About one hundred and fifty feet above the main stream these oozing springs of acid water cease, but the character of the gulch remains the same. The odor of sulphur now becomes stronger, though producing no other effect than a slight irritation of the lungs.

"The gulch ends, or rather begins, in a scoop or basin about two hundred and fifty feet above Cache Creek, and just below this was found the fresh body of a large bear, a silver-tip grizzly, with the remains of a companion in an advanced stage of decomposition above him. Near by were the skeletons of four more bears, with the bones of an elk a yard or two above, while in the bottom of the pocket were the fresh remains of several squirrels, rock hares, and other small animals, besides numerous dead butterflies and insects. The body of the grizzly was carefully examined for bullet holes or other marks of injury, but showed no traces of violence, the only indication being a few drops of blood under the nose. It was evident that he had met his death but a short time before, as the carcass was still perfectly fresh, though offensive enough at the time of a later visit. The remains of a cinnamon bear just above and alongside of this were in an advanced state of decomposition, while the other skeletons were almost denuded of flesh, though the claws and much of the hair remained. It was apparent that these animals, as well as the squirrels and insects, had not met their death by violence, but had been asphyxiated by the irrespirable gas given off in the gulch. The hollows were tested for carbonic-acid gas with lighted tapers without proving its presence, but the strong smell of sulphur, and a choking sensation of the lungs, indicated the presence of noxious gases, while the strong wind prevailing at the time, together with the open nature of the ravine, must have caused a rapid diffusion of the vapors.

"This place differs, therefore, very materially from the famous Death Valley of Java and similar places, in being simply a V-shaped trench, not over seventy-five feet deep, cut in the mountain slope, and not a hollow or cave. That the gas at times accumulates in the pocket at the head of the gulch is, however, proved by the dead squirrels, etc., found on its bottom. It is not probable, however, that the gas ever accumulates here to a considerable depth, owing to the open nature of the place, and the fact that the gulch draining it would carry off the gas, which would, from its density, tend to flow down the ravine. This offers an explanation of the death of the bears, whose remains occur not in this basin, but where it narrows to form the ravine, for it is here that the layer of gas would be deepest, and has proved sufficient to suffocate the first bear, who was probably attracted by the remains of the elk, or perhaps of the smaller victims of the invisible gas; and he, in turn, has doubtless served as bait for others who have in turn succumbed. Though the gulch has doubtless served as a death-trap for a very long period of time, these skeletons and bodies must be the remains of only the most recent victims, for the ravine is so narrow and the fall so great that the channel must be cleared out every few years, if not annually. The change wrought by the water during a single rainstorm, which occurred in the interval between Mr. Weed's first and second visits, was so considerable that it seems probable that the floods of early spring, when the snows are melting under the hot sun of this region, must be powerful enough to wash everything down to the cone of dÉbris at the mouth of the gulch." Mr. Arnold Hague, on the occasion of his visit, was more successful in obtaining evidence of the presence of carbonic-dioxide gas. He writes: "The day I went up the ravine I was able in two places to extinguish a long brown paper taper. The day I was there it was very calm, and where I made the test the water was trickling down a narrow gorge shut in by shelving rocks above."

It was at noon on the 22d of July in the summer of 1897 that we made camp near the mouth of Cache Creek, about three miles southeast of the military post and mail station of Soda Butte. In company with Dr. Francis P. King I at once started up the creek, keeping the left bank, that we might not miss the gulch, which joins the valley of Cache Creek from the southern side. We had a toilsome climb through timber and over steep embankments, cut by the creek in a loose conglomerate, and after going about a mile and a half we noticed that some of these banks were stained with whitish and yellow deposits of alum and sulphur, indicating that we were nearing the old hot-spring district. Soon a caved-in cone of travertine was seen, with crystalline calcite and sulphur in the cavities, and the bed of the creek was more or less completely whitened by these deposits, while here and there could be seen along the banks oozing "paint-pots" of calcareous mud, in one case inky black, with deposits of varicolored salts about its rim, and a steady ebullition of gas bubbles rising from the bottom. In other cases these pools were crystal clear, and always cold. The vegetation, which below had been dense close to the creek's bank, here became more scanty, especially on the southern side, where the bare rock was exposed and seen to be a volcanic breccia, much decomposed and stained with solfataric deposits. A mound of coarse dÉbris seen just above on this side indicated the presence of a lateral ravine, which from its situation and character we decided was probably the gulch sought for. A strong odor of sulphureted hydrogen had been perceptible for some time, and when we entered the gully the fumes became oppressive, causing a heavy burning sensation in the throat and lungs. The ravine proved to be as described, a V-shaped trench cut in the volcanic rock, about fifty feet in depth, with very steep bare whitish slopes, narrowing to a stony rill bed that ascended steeply back into the mountain side.

Climbing through this trough, a frightfully weird and dismal place, utterly without life, and occupied by only a tiny streamlet and an appalling odor, we at length discovered some brown furry masses lying scattered about the floor of the ravine about a quarter of a mile from the point where we had left Cache Creek. Approaching cautiously, it became quickly evident that we had before us a large group of huge recumbent bears; the one nearest to us was lying with his nose between his paws, facing us, and so exactly like a huge dog asleep that it did not seem possible that it was the sleep of death. To make sure, I threw a pebble at the animal, striking him on the flank; the distended skin resounded like a drumhead, and the only response was a belch of poisonous gas that almost overwhelmed us. Closer examination showed that the animal was a young silver-tip grizzly (Ursus horribilis); a few drops of thick, dark-red blood stained his nostrils and the ground beneath. There proved to be five other carcasses, all bears, in various stages of decay; careful search revealed oval areas of hair and bones that represented two other bears, making a total of eight carcasses in all. Seven were grizzlies, one was a cinnamon bear (Ursus americanus). One huge grizzly was so recent a victim that his tracks were still visible in the white, earthy slopes, leading down to the spot where he had met his death. In no case were any marks of violence seen, and there can be no question that death was occasioned by the gas. The wind was blowing directly up the ravine during our visit, and we failed to get any test for carbonic acid, though we exhausted all our matches in the effort, plunging the flames into hollows of the rill bed in various parts of its course; they invariably burned brightly, and showed not the slightest tendency to extinguish. The dilution of the gas in such a breeze would be inevitable, however; that the gas was present was attested by the peculiar oppression on the lungs that was felt during the entire period that we were in the gulch, and which only wore off gradually on our return to camp. I suffered from a slight headache in consequence for several hours.

There was no difference in the appearance of the portion of the gulch where the eight bears had met their end and the region above and below. A hundred yards or more up stream the solfataric deposits become less abundant, and the timber grows close to the brook; a short distance beyond this the gulch ends. No bodies were found above, and only bears were found in the locality described. It will be observed that Weed's experience differs in this respect from ours, and the appearance of the place was somewhat different: he found elk and small animals in addition to the bears, and describes the death-trap as occupying the mouth of the basin at the head of the gulch, above the point where the last springs of acid water cease. The rill observed by us has its source far above the animals; indeed, it trickles directly through the worm-eaten carcass of the cinnamon bear—a thought by no means comforting when we realized that the water supply for our camp was drawn from the creek only a short distance down the valley.

It is not impossible that there may be two or three of these gullies having similar properties. That we should have found only bears may perhaps be accounted for on the ground that the first victim for this season was a bear, and his carcass frightened away all animals except those of his own family. For an illustration of a process of accumulation of the bones of large vertebrates, with all the conditions present necessary for fossilization, no finer example can be found in the world than Death Gulch; year after year the snow slides and spring floods wash down this fresh supply of entrapped carcasses to be buried in the waste cones and alluvial bottoms of Cache Creek and Lamar River. Probably the stream-formed conglomerate that we noted as we ascended the creek is locally filled with these remains.

The gas is probably generated by the action of the acid water on the ancient limestones that here underlie the lavas at no great depth; outcrops of these limestones occur only a few miles away at the mouth of Soda Butte Creek. This gas must emanate from fissures in the rock just above the bears, and on still nights it may accumulate to a depth of two or three feet in the ravine, settling in a heavy, wavy stratum, and probably rolling slowly down the bed of the rill into the valley below. The accompanying photographs were made during our visit.