Geological Survey Bulletin 395 Radioactivity of the Thermal Waters of Yellowstone National Park

GAS ACTIVITIES.

Thorium emanation.—The tabulated results show that marked differences exist in the activity of gases from different springs. In most cases the activity is due entirely to radium emanation. In at least 16 of the 82 gas sources examined thorium emanation was found associated with that of radium. These are the first springs on this Continent which to our knowledge have been found to contain thorium emanation, and the number reported in Europe is small. The initial activity due to thorium emanation varied from a mere trace to fifteen times that due to radium emanation in the same gas. Owing to the rapid rate of decay of the thorium emanation the results relative to it are only approximately quantitative and represent the minimum quantity. The thorium emanation was most marked at Norris Geyser basin and at Terrace Springs near the mouth of Gibbon River. Thorium emanation was not found alone in any gas. Some radium emanation always accompanied it. Actinium emanation was looked for, but without success. Its extremely rapid rate of decay (half value in four seconds) makes its detection difficult in small amounts under field conditions.

PLATE III. A. OLD FAITHFUL GEYSER, UPPER GEYSER BASIN. B. HOT SPRING, NORRIS GEYSER BASIN, WHERE THORIUM WAS FIRST DISCOVERED IN THE UNITED STATES.

Relative activity.—In only one of the ten localities visited in the park were the escaping gases found, on the whole practically inactive. This locality is the well-known group of springs in the northern part of the park, known as Mammoth Hot Springs. Since the volume of the escaping gases from these highly carbonated waters is relatively large, this fact probably accounts for the extremely low concentration of the emanation in the gases and waters of this locality. In this connection it should be mentioned that the most active gas recorded (pool near Model Geyser, Upper Geyser basin) issued very slowly, and that the next most active gas (pool near Clepsytra Geyser, Lower Geyser basin) required nearly an hour for the collection of a sample of 50 cubic centimeters in a bottle having a mouth 3 centimeters wide. The latter spring gives the appearance of discharging gas freely, but most of the bubbles proved to be steam.

Gases from European springs.—Some of the well-known thermal and cold springs on the European Continent are more or less charged with gases, like many of those of Yellowstone Park. The radioactive properties of the gases escaping from the European springs have been extensively studied by different investigators, and numerous quantitative results are thus available for comparison. The following table gives representative data for gases from springs in various European localities and for different localities in Yellowstone Park. The values for the activity in the third column of the table are expressed in electrostatic units.

TABLE 8.—Radioactivity of gases from springs in Europe and in Yellowstone Park.

Although the values for the activity as given by different observers are based upon somewhat different systems of reckoning (some investigators, Mache, for example, do not include the ionization effect of the disintegration products), still the values given show that the gases from the springs in Yellowstone Park are fully as active as those from European sources.

Helium not determined—In several of the more active European springs it is known that helium is one of the constituents of the escaping gas. No tests were made by us to ascertain the presence of helium in any of the gases tested. The marked radioactive properties of the gases from many of the springs would make an examination of the gases for helium and the other rare gases very desirable.

ACTIVITY OF WATERS.

Inferences from different results.—The tabulated results clearly show that the thermal waters in different localities and even in the same locality differ considerably in activity. In some samples no activity was detected. Several cold springs were also tested. These likewise showed marked differences in activity, and their average somewhat exceeds that of the thermal springs. It is evident then that the temperature of the waters is independent of their activity. Similar variations in activity were found by Boltwood^a in his extensive tests of the radioactive properties of the waters of Hot Springs, Ark., and by us in connection with the examination of some natural waters of Missouri.^b The acidity or alkalinity of the water seems to have no connection with its activity.

---

^aAm, Jour, Sci., vol. 20. 1905, p. 128.
^b Trans. Am, Electrochem. Soc., vol. 8, 1905, p. 291.

---

Is the activity of the waters due to radium emanation which has diffused into them in the course of their underground flow, or is it produced by the radium that the waters carry in solution? As those water residues which were examined yielded only traces of radium emanation, we may conclude that the radium salts in solution are responsible for only a small part of the total activity, most of which is due to radioactive gases that have diffused into the waters.

Some of the differences observed in the activities are readily explained. The water of boiling springs and therefore of geysers almost invariably gave negative tests. Such a temperature is not favorable for the retention of dissolved gases. The escaping steam and the consequent agitation of the waters are both effective in removing any radioactive gases that may have been absorbed by the water.

The same effect may be caused in another way. If a large volume of gas is evolved from a spring, whose temperature may lie considerably below the boiling point, the emanation is swept out of the water by the gas. Such a condition also is not favorable for the accumulation of radioactive gases. In such a spring the activity of the gas also will prove to be low, although a considerable amount of emanation may have been removed by it. If these conditions were not considered such a spring might be classified as inactive, though its residue on evaporation might be found to contain appreciable quantities of radium. Most of the springs tested at Mammoth Hot Springs are examples of this condition. All of the springs on the Main Terrace gave negative results for both waters and gases, within the limits of error of our instruments under field conditions, and yet the residues and deposits, as Tables 4 and 5 show, greatly exceed in activity those of springs in other localities, the waters of which were found to be radioactive in a marked degree.

At all events the results for waters and gases show that the underlying rock in the park contains radium and in some localities thorium in sufficient quantities to make the waters and gases decidedly radioactive.

Comparison with European waters.—The radioactive properties of European thermal and mineral waters have been studied extensively by different investigators. From the numerous quantitative data a few have been selected and arranged in tabular form. Some of the values obtained by Boltwood in his examination of the waters of Hot Springs, Ark., are incorporated in the table, as well as values obtained for some Yellowstone Park waters.

TABLE 9.—Radioactivity of spring waters in Europe and in the United States.

A comparison of the activities of the thermal waters of Yellowstone Park with those of European springs, as given by the foregoing table, shows that with a few exceptions the quantity of radium emanation carried is of the same order of magnitude and that marked variations exist. It is seen that the European springs in localities where deposits of pitchblende occur are distinctly more radioactive. In making comparisons of this kind, however, it must be borne in mind that the temperature of most of the park waters is near the boiling point, and that this factor is unfavorable for the accumulation of the emanation. It is worthy of note, in this connection, that the gases escaping from many of the springs in the park contain about the same quantity of radium emanation as the most active European thermal springs, which are located in regions where extensive deposits of radioactive minerals occur. It is therefore probable that many of the park waters traverse radium-bearing minerals in the course of their underground flow.

The waters of Hot Springs, Ark., are probably fully as active as those of the park, but their temperatures average considerably lower. The gases rising from two of the springs at Hot Springs were found by Boltwood^a to be less active than equal volumes of gases separated from the waters hy boiling, a result in direct contrast with our observations on the thermal springs of the park.

---

^aAm. Jour. Sci., vol. 20, 1905, p. 128.

---

A discussion of the results in the foregoing table with respect to the hydrography of the different localities lies beyond the scope of this paper.

PLATE IV. MORNING GLORY POOL, UPPER GEYSER BASIN.

RADIUM CONTENT OF DEPOSITS AND ROCKS.

Comparative activity of different rocks.—At Mammoth Hot Springs the waters are calcareous, the dissolved material being derived from the underlying Jurassic and Cretaceous. At the other basins the waters are siliceous (some alkaline, some acid), the dissolved material being derived from the rhyolite of which the park plateau is largely formed. In both classes of deposits radium was detected.

The deposits of travertine at Mammoth Hot Springs contain radium in appreciable quantities. The activity of the residue obtained by evaporating the water is about half that of the deposits, showing that the radioactive material is largely deposited with the components that come out of solution first. This fact alone would account in a large measure for moderate differences in the radium content of the deposits formed by different springs of similar origin. The values obtained at Mammoth Hot Springs for the radium content of the deposits from the actively flowing springs, however, are of the same order of magnitude. The radium content of these deposits averages about 8 x 10^-12 grams per gram of material, a value two and one-half times greater than the radium content of the rock from which the deposits are probably derived, and about five times the average radium content of the igneous rock of the park. To account for the concentration of radium in these spring deposits more experimental work is necessary. An investigation of the solvent action of hot carbonated water on radium-bearing rock samples at different pressures might throw some light on the question. Again, the fact that radium in the Mammoth Hot Spring deposits is largely deposited with the components that first come out of solution may be connected with the numerous growths of algae in these hot waters, which may act as precipitating reagents.

The siliceous deposits from some of the springs in the other basins of thermal activity likewise show a higher radium content than the igneous rock of the vicinity. In some cases the content is several times greater, but this result is not general. In fact, the nearly pure silica deposits from most of the geysers and the older springs contain less than half as much radium as the underlying rhyolite. The location of the springs appears to be related to the radium content of the residue. The residues with a radium content greater than that of adjoining rocks come, as a rule, from relatively new springs which are close to the outcropping rock. The disintegrated rock material from the so-called mud pots, paint pots, and mud volcanoes were also found to contain more radium than the park rhyolite.

Radioactive equilibrium.—Another question naturally arises in this connection: Is the radium thus deposited in radioactive equilibrium with its parent uranium? Some experimental evidence obtained indicates that far less uranium is deposited than the radium content requires. Of course, the quantity of uranium considered here is too small to be detected by chemical analysis, but if radium were present in the equilibrium ratio the activity of the deposits would remain practically constant for years. Now, the radium present in the old terraces, long since dry, is considerably below the average value for the recent deposits. A sample of travertine from Terrace Mountain, the scene of the first hot-spring activity at Mammoth Hot Springs, contains only about 1 per cent of the radium found in the present-day deposits of Jupiter terrace and Main terrace. Again, a sample of travertine from Hotel terrace, on which the government fort and buildings stand, carried only about 40 per cent of the radium found in most of the present-day deposits. Of course, these results might possibly be attributed to loss of radioactive material through weathering; but the hypothesis that radium and uranium are not deposited in radioactive equilibrium also furnishes a satisfactory explanation for the fact.

Estimate of age of deposits.—If we start with a given quantity of radium, separated from its parent uranium, at the end of 2,600 years, according to recent determinations, one-half of it would be transformed into inactive products, at the end of another such period there would be only one-fourth of the original amount, and so on. If we assume, then, that the uranium deposited with the radium is far below the equilibrium quantity, the experimental results obtained follow. Moreover, from the data thus in hand the approximate age of the old deposits can be calculated. The application of these data to the question of the age of the deposits of course involves the assumption that the original radium content of the travertine in the old terraces was approximately the same as in the new. Such an assumption is not far from the truth. The similarity of the waters of this locality as a whole in regard to their radioactive content, as well as in other respects, is very close and indicates a common origin. Moreover, some of the old terraces are well within the zone of present activity. The travertine of Terrace Mountain is overlain by glacial bowlders. Since its activity is only 1 per cent of that of the recent deposits, its age is about 20,000 years,^a which would also be the approximate time that has elapsed since the glacial period in the park.

---

^aThis would he 14,000 years if we accept 2,000 years as the half-time period of radium.

---

Comparative test for uranium by analysis.—Another line of experiments bearing on this question was suggested by some results obtained quite recently in determining the radium present in some of the "tufa" deposits made by the springs at Hot Springs, Ark. These deposits, like the travertine at Mammoth Hot Springs, are calcareous. The radium content^b of one of these deposits indicated the presence of about one-half of 1 per cent of uranium, a quantity well within the range of analytical methods. Investigation of the question from this standpoint would be of interest. The deposit in question, which was obtained from the Palace Spring, is three times as active as the famous Bath Springs of England tested by Strutt.

---

^bTrans. Am. Electrochem. Soc., vol. 12, 1907, p. 247.

---

RELATION TO HYDROTHERMAL ACTIVITY.

Assuming that the thermal activity in the park is due to recent extrusive lava flow, or intrusive lava masses, the question arises to what extent the radium content of these rocks affects their rate of cooling. To what extent is the radium responsible for the continued hydrothermal activity? We have attempted no calculations with the data obtained from the Yellowstone Park rocks, for we do not know the volume of water discharged by these springs, and so lack data for calculating the heat dissipated by the water. At Hot Springs, Ark., the present flow from the springs is given by Haywood and Weed^c in their reports on the Hot Springs of Arkansas, at 850,000 gallons daily. A simple calculation shows that the heat dissipated by the springs, flowing at the present rate during ten thousand years, would be furnished by approximately one cubic mile of solid lava cooling from its fusion point to the temperature of boiling water. Assuming that the radium present in the rocks is even somewhat greater than the average value found by Strutt^a and by us, for igneous rocks (say 5 X 10^-12 grams per gram of rock), the heat supplied by it per cubic mile of rock is about one-tenth of 1 per cent of the heat dissipated by the water.

---

^cHaywood, J. K., Report of an analysis of thc waters of the hot springs on the Hot Springs Reservation, Ark.; and Weed, W. R. Geological sketch of the Hot Springs district: 57th Cong., 1st sess., S. Doc. No. 282, 1902, pp. 31, 93.

^aProc. Roy. Soc., vol. 77, p. 472.

---

The rock samples tested for radium gave values that average only a little more than Strutt's average for rhyolites and basalts—about 2 X 10^-12 grams radium per gram of rock. These data certainly seem to indicate that the hydrothermal activity so manifest in the park is not connected with localized deposits of radium. In the above calculations the question of heat lost by diffusion and other factors is not taken into consideration, but after allowing a generous margin for error we do not see how more than 1 per cent of the heat required for the hydrothermal action can be ascribed to the radium content of the rock. It is true that a sample of limestone from Mammoth Hot Springs, which Arnold Hague considers typical of the underlying Mesozoic limestones, is more active than the most active of Strutt's sedimentary rocks,^b a sample of oolite from the neighborhood of the Bath thermal springs of England, but this coincidence alone is not of sufficient importance to justify a verdict.

---

^bIdem, vol. 78, p. 150.

---