THE AGE OF THE EARTH.1

By the Right Hon. LORD RAYLEIGH, F. R. S., and Others.

The subject which we have met to consider to-day is encumbered with past controversy. It can not be denied, I am afraid, that exponents of particular views in the past have laid too much emphasis on their own particular way of looking at the problem without making enough allowance for human fallibility. I shall try, so far as possible, to avoid this pitfall. There has been a tendency on all sides for specialists in one branch of science to consider themselves free to disregard evidence drawn from a class of considerations with which they are not familiar. I am sure that this is not the road to truth. In attempting a problem of this kind, when we seek to plumb into the depths of time, far beyond human experience, we can not afford to neglect evidence drawn from any quarter, even if it is not the kind of evidence which we find it most congenial to contemplate. A parallel case is that of a jury of plain men in a murder trial. They may know nothing of medical jurisprudence, postmortem examinations, and so on. They may even consider the subject repellent; but that does not exempt them from the duty of fully considering and weighing such evidence to the best of their ability. The witnesses in the trial have, however, to limit themselves to matters with which they are personally conversant. I will try to give my evidence within these limits.

The phrase "age of the earth," though rather vague, is perhaps definite enough for our purpose. What we want to know is, how long has the earth's surface been fitted for the habitation of living beings; or, alternatively, how long has it taken to accumulate the known series of geological formations? These questions are not the same, but I do not think that we shall need to insist on the distinction this morning.

Lord Kelvin's arguments depended on attempts to limit the length of time during which the earth's surface temperature could have re

1 Contributions to a joint discussion of the Sections of Mathematical and Physical Science, Geology, Zoology, and Botany of the British Association at Edinburgh on Sept. 13, 1921. Reprinted by permission from Nature, Oct. 27, 1921.

mained substantially the same as at present, and he attacked this problem from two different points of view. In the first place he attempted to set a limit of time to the duration of the sun's heat; and secondly, from consideration of the earth's internal heat, he argued back to the time when the surface was too hot for the presence of living beings. I have heard a suggestion that there is some mutual inconsistency in these two lines of argument-consideration of the sun's heat makes the past temperature too low; consideration of the earth's heat makes it too high-but I do not think that this criticism is more than superficially plausible. The point was rather that from either of these arguments a condition widely different from the present would be reached, and therefore that, even if there were some unrecognized flaw in one of the arguments, the other would stand. Possibly, looking back into the remote past, a condition of the earth's surface is imaginable where the mean temperature was much the same as at present, heat coming from the earth's interior in compensation for a diminished radiation from the sun, but I feel sure you will all agree with me that we can not get more time by special pleading of this kind. The fossiliferous rocks have, without doubt, been accumulated under conditions of solar radiation not essentially different from the present. One simple consideration is that the plants in the coal measures obviously had green leaves, and that these could not function without a full allowance of solar radiation.

We have then to consider whether Lord Kelvin's arguments can stand in the light of present knowledge. I think that we must admit that they can not.

First as regards the earth's heat, it is now generally known that the premises of Lord Kelvin's calculation, carefully particularized by him, are upset by the discovery of radio-active substances in the earth. In 1906 I made a determination of the amount of radium in the superficial parts of the earth which are alone accessible. From radium analysis we can calculate the amount of uranium and other associated substances and the thermal output from them, and the result is to show that if we suppose the same radium content to extend to a depth of some 20 miles, the whole output of heat would be accounted for without assuming that any of it comes from the store of primeval heat as postulated by Lord Kelvin. It is without doubt difficult to understand why the output of heat is not greater, for it would certainly be expected that the rocky crust of earth would be more than 20 miles thick, to say nothing of any radium there might be in the unknown interior.

Can we at present infer anything definite from the earth's internal heat as to the possible duration of geological time? I think practi

cally not. It appears certain that the radioactive materials present in the earth are generating at least as much heat as is now leaking out from the earth into space. If they are generating more than this (and there is evidence to suggest that they are), the temperature must, according to all received views, be rising. In a word, we are puzzled to explain the existing state of things, and cannot use it as a firm basis from which to explore the past.

Next, as to the sun's heat. Lord Kelvin's argument was that we knew of no possible source at all adequate to supply the existing output of solar energy except secular contraction, and even this source of supply was not enough to account for more than 20,000,000 years of solar heat in the past. It is impossible to condemn on principle arguments of this kind. We often must, and do, rely on them in science as in everyday affairs; but a certain reserve is always needed on the ground that there are more things in heaven and earth than are dreamt of in our philosophy. Knowledge which has accumulated since Lord Kelvin's time has driven us back on this alternative.

The sun is only one of the host of stars, and if we find it impossible to account adequately for their radiation by contraction it evidently will not do to assume that the sun is limited to this source of supply.

Now, some of the stars (the giant red stars), though of about the same mass as the sun, are radiating energy at something like one thousand times the rate that the sun does. They ought, according to the contraction theory, to have expended a considerable fraction of their total energy in historical times. No one will maintain that this has occurred, and if not there must be some source of supply other than contraction. It is not necessary for our immediate purpose to inquire what this source is. It is enough to note that its existence invalidates Lord Kelvin's estimate of the age of the sun's heat.

Modern knowledge in radioactivity has given what appears, if separately considered, to be a firm and satisfactory basis for the estimation of geological time. Uranium, for example, goes through a series of changes (radium is one of the stages in its progress), changing eventually into an isotope of lead—that is, an element chemically indistinguishable from lead, except by a slight difference of atomic weight and (practically at least) inseparable from ordinary lead by chemical means if once mixed with it. The isotope of lead in question has probably an atomic weight of 206 exactly, as contrasted with an atomic weight of 207.1 for ordinary lead. This is much less than the atomic weight of uranium (238.5), and the difference represents approximately the weight of helium atoms, which are the débris shed at the various stages of the transformation.

2 Ordinary lead may partly consist of it, but this is not yet certain, and not very important for the immediate purpose.

Further, it is well established that a gram of uranium as found along with its products in rocks and minerals is now changing at a rate represented by the production of 1.88X10-11 grams of helium and 1.22×10-10 grams of lead isotope per annum. We have not time this morning to consider the methods by which these figures have been reached. It must suffice to say that in the case of helium it amounts practically to direct observation, while in the case of lead isotope the evidence, though less direct, is very strong, and, so far as I am aware, is not contested by any student of the subject. I have said that this is the rate at which one gram of uranium as found in the earth is producing helium and lead isotope at present. It is important to inquire whether one gram of uranium did the same in the past. This we can not, of course, determine directly. It is certain that nothing we can do in a laboratory in the way of change of temperature and pressure can alter the rate sensibly, and enough has been done in this way to make it unlikely that any pressures and temperatures encountered in the superficial parts of the earth could have such an effect. It has been suggested by Professor Joly that the absolute age of a gram of uranium may affect its rate of disintegration. All possibilities should be considered, but this suggestion derives no support from the behavior of the short-lived radioactive substances the behavior of which we can watch.

Upon the whole, therefore, it would seem that in the disintegration of a gram of uranium we have a process the rate of which can be relied upon to have been the same in the past as we now observe it to be.

The application is either to individual uranium minerals or to the earth's crust as a whole. Taking first the minerals containing uranium, these are found in all cases to contain helium and lead. The helium in them, which appears to be retained mechanically, may safely be treated as wholly a radioactive product. The lead in some cases conforms closely to the expected atomic weight of 206, about one unit lower than common lead, and in such cases we may safely regard the whole of it as a product of uranium disintegration.

Thus take the broggerite found in the pre-Cambrian rocks at Moss, Norway. The lead in this mineral has an atomic weight of 206.06 as determined by Hönigschmied and Fräulein St. Horovitz. The ratio of lead to uranium is 1:3. Taking the lead as all produced by uranium at the rate above given, we get an age of 925,000,000 years. Some minerals from other archæan rocks in Norway give a rather longer age.

In other cases there is some complication, owing to the fact that thorium is associated with uranium in the mineral and that it, too, produces helium and an isotope of lead of atomic weight probably 208 exactly, about one unit higher than common lead.