THE SALINITY AND AGE OF THE OCEANS
Halley’s proposal in 1715—Joly’s application of modern
data in 1899—Further studies by Sollas, Clarke, and
Becker—Data of the problem—Corrections for disseminated
sodium, wind-borne sodium, human agencies, and marine
erosion—Uniformity not capable of proof—Importance of
cyclic circulation of sodium—Origin of chlorine now fixed
in salt—Hour-glass method applied—Land areas of the past
and their elevation compared with those of to-day.
In 1715 the famous astronomer Edmund Halley published a paper which he
entitled, _A short Account of the Cause of the Saltness of the Ocean,
and of the several Lakes that emit no rivers; with a Proposal by help
thereof to discover the Age of the World_. He showed that since the
water removed from lakes by evaporation is perfectly fresh, “the saline
particles brought in by the rivers remain behind, while the fresh
evaporate; and hence it is evident that the salt in the lakes will be
continually augmented and the water grow salter and salter.” Applying
the same principle to the oceans, he wrote, “It is not improbable but
that the ocean itself is become salt from the same cause, and we are
thereby furnished with an argument for estimating the duration of
all things.” Two hundred years ago it seemed hopeless to attempt to
determine the annual increment of salt added to the oceans, but Halley
used his argument “to refute the ancient notion some have of late
entertained of the eternity of the world.” The paper was completely
forgotten, until Dr. G. F. Becker again drew attention to it a year or
two ago.
As we saw in the last chapter, the analysis of river waters has now
made possible a determination of the annual amount of material carried
in solution by the rivers into the oceans. T. Mellard Reade was the
first to contemplate the application of solvent denudation to the
measurement of geological time, and the data he gathered together
in support of his arguments came as a great surprise to those who
had concentrated attention merely on the mechanical work of erosion.
No independent advance, however, was possible until 1899, when Joly
pointed out that of the many elements which enter into the composition
of sea water, sodium alone tends to accumulate. All the others are
sooner or later rejected, associating themselves with the detrital
sediments, or forming chemical or organic sediments by their ultimate
precipitation. Joly then proceeded to use sodium as the age-index of
the oceans. He assumed that the annual increment Naᵣ of sodium added
to the oceans by all the rivers of the world has remained practically
constant throughout geological time. If Naₒ represents the total amount
of sodium now accumulated, the ratio Naₒ/Naᵣ gives the time which has
elapsed since the oceans first existed and denudation began to wear
down the lands. Joly’s first estimate was from 80 to 90 million years,
and shortly afterwards he increased this to 100 million years. Sollas
attacked the problemafresh in 1909, and from a most careful survey of
all the data and a detailed enquiry into every phase of the subject,
he concluded that the most probable estimate of the age of the oceans
would appear to lie between 80 and 150 million years.
In 1910 a further study was made by Clarke and Becker. The latter
departed from the uniformitarian basis on which all the other
calculations had been founded. He inferred that sodium accumulation
progressed more rapidly in the past than at present. All the original
sodium must have been derived from igneous rocks, and Becker considers
that at the time when the oceans were first possible the surface of the
earth must have consisted exclusively of such rocks. At the present
day three-quarters of the land areas are covered by sedimentary rocks
which can supply no further important additions to the sodium content
of the ocean. Assuming that the production of sodium has been always
proportional to the area of igneous exposures, and that the total land
area of the globe has averaged 80% of the present area, he finds that
the age indicated is about 70 million years.
We may now proceed to examine the problem in detail. The fundamental
data on which the method is based may be summarised as follows:
Mean density of the ocean (Murray) 1·026
Volume of the ocean (Murray) 323,800,000 cu. miles
” ” (Joly) 339,248,000 ”
” ” (Clarke) 302,000,000 ”
” ” (Karsten) 307,496,000 ”
Clarke now accepts the latter as being the best estimate, and we
therefore calculate that the
Mass of the ocean = 1,178,270 × 10¹² tons.
Total salinity of the ocean (Dittmar) 3·5%
Sodium of the ocean (Dittmar) 1·08%
Accumulated sodium, Naₒ = 12,600 × 10¹² tons.
Annual increment, Naᵣ = 156,000,000 tons.
From these values the quotient Naₒ/Naᵣ gives the age of the oceans to a
first approximation as 80·8 million years. This figure, however, cannot
be regarded as final. Two of the assumptions on which it is based are
that all the sodium liberated from igneous rocks is contained in the
ocean, and that all the sodium carried annually to the ocean has been
liberated from such rocks for the first time. The obvious corrections
to be applied will increase the numerator of the ratio and decrease
the denominator, thereby increasing the age estimate. The degree to
which the numerator must be augmented is, as far as we know, relatively
insignificant. Saline deposits, such as those of the Stassfurt
district, are only of trivial importance. The salt of the ocean, if
spread regularly over the whole land areas, would form a layer about
120 feet thick, and all the beds of rock salt which have accumulated by
evaporation become negligible beside this vast quantity. The amount of
sodium in ground waters and disseminated through the sedimentary rocks
is to be taken into consideration more carefully. Marine sediments at
the time of their formation are saturated with sea water, and, when
raised up to form land, they must, therefore, be strongly charged with
salt. The actual proportion of sodium abstracted from the ocean in this
way cannot at one time be very great, probably not more than 1% of the
whole. Throughout geological time, however, a certain amount of sodium
has been in cyclic circulation between land and sea. A rough idea of
the influence of this circulation on the age estimate may be gained
from a knowledge of the total volume of the sedimentary rocks, and of
their average pore space.
Neither of these factors is yet known with the desirable accuracy, but
as far as present needs are concerned, the total bulk of the sediments,
excluding deep-sea deposits, may be placed at 70 million cubic miles,
and their pore space at 10%. The latter figure is probably too low,
for fresh sediments have a pore space of 40% to 50% of their volume,
and in consolidated sandstones 20% is common. Many of the sediments
have been denuded and re-deposited time after time; on the average,
probably three times. On the other hand, if the ocean has progressively
increased in salinity, its average sodium content must have been
about half that of the present. Basing our calculations on these
considerations, the total volume of sediments which have ever existed
is about 210 million cubic miles; the pore space to be filled by sea
water would then be 21 million cubic miles. The total amount of sodium
precipitated within the rocks by evaporation could not therefore exceed
420 million million tons, and at the present rate of denudation its
removal would require nearly three million years. To what extent this
result would be increased by absorption effects, whereby interstitial
salt solutions are concentrated in order that they may be in
equilibrium with normal sea water, cannot be estimated. This corrective
factor would probably not be large.
A still more important cyclic circulation is brought about by
wind-borne sea salt. Fine particles of spray are swept from the foaming
crests of waves by the wind and carried often many miles inland. Near
the coast the salt blown over the land is naturally most abundant. It
falls in sea fret and rain, and is ultimately returned to the sea in
the drainage from the land. Fortunately, the amount of sodium supplied
in this way can be estimated by analyses of rainwater and a measurement
of the rainfall. As before, the cyclic sodium must be excluded from
that entering into the denominator as primary sodium. Joly allows an
additive time correction of 10%, Clarke of 7%, and Becker of 6%. The
time estimate is therefore increased by a further five or six million
years.
Clarke has suggested that the present rate of sodium accumulation has
been accelerated by human agencies. Nearly six million tons of sodium
are annually produced as common salt, and, as sewage and chemical
refuse, much of this is again returned to the oceans. Subtracting this
from the denominator, the age is increased by other three million years.
A number of corrections may be made to decrease the first rough
estimate of the ocean’s age. Solvent denudation due to marine erosion
must not be forgotten. Experiments by Joly have shown that sea water
is several times (2½-14) more active as a solvent than fresh water. He
attributes to its action over the tide-swept strand—covering a total
area of about 60,000 square miles—a supply of sodium amounting to 3%
of that derived from the normal drainage areas. This is a very liberal
estimate, and in applying a negative correction of three million years
we are granting all that can reasonably be claimed.
Finally, there are other possibilities of uncertain magnitude which
deserve mention. The ocean may conceivably have contained sodium
before the fluviatile contributions commenced. By volcanic eruptions,
sodium-bearing materials may be cast widespread over the sea.
_Juvenile_ waters expelled from igneous magmas may presumably
afford a supply of sodium. The importance of saline waters derived from
the heated interior of the earth has been particularly insisted upon
by the great geologist Suess. As a source of sodium, such waters are
probably of little moment, for the whole amount of sodium in the ocean
is already more than explained by the erosion of igneous rocks. Solvent
denudation involves a loss from average igneous rock of 1·57% due to
sodium abstraction alone. According to Clarke, the average sodium
content of igneous rocks is 2·52%, and the amount retained by the
detrital material which goes to make up the sediments is 0·95% of the
original rock. Now, if the total mass of sodium which has accumulated
in geological time is 12·6 × 10¹⁵ tons and this represents 1·57% of the
igneous rock in which it originally occurred, it is evident that the
mass of the igneous rocks which have been denuded away amounts to 800
× 10¹⁵ tons. The mass of one cubic mile of rock may be taken as 10,800
million tons, and hence the volume of igneous rock which has been
removed is 74 million cubic miles. The sediments derived from the
latter ought therefore to occupy a volume of the same order. A number
of independent estimates of the volume of sediments existing on the
land areas have been attempted, and while no great accuracy can be
hoped for, it is significant that all agree in assigning a volume
somewhat greater than 74 million cubic miles. The sediments hidden
under the oceans would add still further to the total. If any reliance
at all is to be placed on these estimates, it is clear that the sodium
in the ocean ought to be more than it is; at any rate, there would
appear to be little room for sodium derived from other sources, such as
the interior of the earth.
We may sum up the results of this discussion as follows:
Approximate age, Naₒ/Naᵣ = 81 million years.
Correction for
(_a_) Disseminated Sodium = 3 million years.
(_b_) Wind-borne Sodium = 6 ” ”
(_c_) Human agencies = 3 ” ”
(_d_) Marine erosion = -3 ” ”
--------------------
Age of the oceans 90 million years.
Another assumption on which this computation is founded must now be
examined; that is, the practical uniformity of the annual addition of
sodium throughout the period involved. Little weight can be attached to
the argument that the source of supply has been gradually impoverished
in alkalies, for the soils from limestones, which are among the poorest
of the alkali-bearing rocks, may contain more sodium than do the soils
derived from granites. The sedimentary rocks being less able to resist
erosion are more rapidly disintegrated and removed than are the igneous
rocks. The composition of streams considered in relation to the rocks
which they drain is of interest here. Hanamann has shown that in the
Elbe and its tributaries three times more sodium is carried away from a
square mile of Cretaceous sediments than from an equal area of granite.
In the same connection, it is well known that the waters flowing
through stratified rocks carry more salt than those from igneous rocks.
The greater exposure of the latter in former land areas, if such were
the case, would certainly not contribute a greater sodium income to
the ocean than if sedimentary rocks prevailed. Joly considers that an
approximate uniformity has been preserved whatever the distribution of
rock types may have been. It would not, however, be difficult to prove
by actual analyses that a greater area of igneous rocks would result in
a reduction of the sodium increment, and that the present rate may be
too high because of the predominance of sedimentaries.
This unexpected conclusion raises the question of the validity of the
method at its most critical point. Can we be sure that the cyclic
circulation of sodium has been taken sufficiently into consideration?
The only source to which the anomalous proportion of sodium from
sedimentary rocks can be traced, apart from the factors already dealt
with, is constituted by the ground waters. These waters often contain
minute quantities of salt of which the origin is very obscure. Whence
came the chlorine? Igneous rocks themselves only contain about 2% of
the chlorine necessary to convert the liberated sodium into salt. The
emission of chlorine from volcanoes and solfataras probably affords
the necessary supply, but most of this never reaches the ground waters
directly. Becker, dealing with the same difficulty in the case of river
water, calculates that 40% of the annual sodium increment is combined
with chlorine of which the origin cannot be traced. He accordingly
rejects this amount as not being of normal or primary origin. If it
has all been derived from the oceans, which he regards as an extreme
possibility, it must be subtracted from the apparent income, since it
is simply a circulation of capital.
A _maximum_ figure for the age of the oceans (still on the
assumption of uniformity) ought theoretically to be attainable by
accepting the whole of the sodium-chloride of the rivers as cyclic. We
have the following figures:
Chlorine = 138 million tons.
Sodium combined with chlorine = 87 ” ”
Sodium uncombined with chlorine = 69 ” ”
-----------------
Total Sodium, Naᵣ = 156 million tons.
The unchloridised sodium implies an age of 180 million years, an
estimate which would in general be condemned as excessive.
The difficulties presented by sediments and ground waters are almost
insuperable, and it may be safer to adopt a means by which they can
be entirely disregarded. The sodium lost from igneous rocks amounts
by mass to 1·57% of the latter. Igneous and metamorphic rocks are at
present exposed over one-fifth of the drainage areas, or 8,000,000
square miles. Many of the metamorphic rocks are partially or wholly
of sedimentary origin, and ought not to enter into the computation.
We have already seen that the average rate of denudation is one foot
in 8600 years, and at this rate, one cubic mile of primary rock, or
10,800 million tons, would be denuded away in 4·54 years. In one year
the supply of sodium would therefore be 37 million tons, and the time
required for the accumulation of the existing store of sodium in the
oceans would be 340 million years.
We meet with a curious discrepancy at this point. The actual amount
of unchloridised sodium is 69 million tons, while that which is drawn
from the igneous rocks cannot exceed 36 million tons. In each case
sub-aerial denudation alone is being considered, so that the failure
to take marine denudation into account does not affect the matter.
The figures imply that a great deal of unchloridised sodium is drawn
from the sedimentaries. Even if the extreme assumption is made for the
moment, that the igneous rocks lose all their sodium, 2·52%, the annual
supply could only reach 60 million tons. The discrepancy is evidently
due to the facts that sediments freshly formed from primary rocks
retain a considerable percentage of sodium, and that it is only after
the material has been re-assorted, perhaps several times, that the
element is withdrawn to the extent indicated by Clarke’s average
figures. At the present day, even when the igneous rocks have been
deprived of 37 million tons of sodium, the sedimentary rocks, to bring
up the total to 69 million tons, must lose more than half of their
primary sodium. It is interesting to observe that if all the sodium
now carried each year to the oceans—156 million tons—were derived
directly from the rocks, igneous and sedimentary, the latter would be
obliged to lose nearly twice as much sodium as they actually contain!
This remarkable conclusion indicates either that our statistics of
the annual production of sediments are hopelessly wrong, or else that
the chloridised sodium is almost wholly cyclic. The latter inference,
supported as it is by the impossibility of otherwise accounting for the
chlorine, thus receives further support.
In the application of the above data to the age question, it has been
tacitly assumed that the present exposures of primary rocks are neither
greater nor less in area than the average of such exposures since the
oceans began.
It is impossible to know what the average may have been, for while
the original lands must have been entirely formed of igneous or
pseudo-igneous planetesimal material, it is equally certain that
their area was but a fraction of that of the present land areas. The
evolution of the earth’s surface features has tended to intensify the
difference between the ocean depths and the mountain heights. The ocean
basins have gradually deepened and become more stable, and their
relative permanence is generally admitted. The lands are now more
extensive than ever they have been; initially it is probable that only
low and scattered islands emerged from the shallow primitive seas. We
are faced here with many vague problems. The volume of the seas may
have been less than it is to-day; the original rocks were probably not
granitic in type, and were certainly poor in sodium; carbon-dioxide
seems to have been the chief atmospheric companion of nitrogen, and
solvent denudation would be correspondingly more vigorous. How far
all these considerations affect the point at issue cannot be decided.
To discuss them would lead us into a complex maze of speculation. The
original basis of our calculation may afford as near an approach to
the truth as could be arrived at by any other reasonable hypothesis of
changing land areas, and of their composition.
On this view it is possible to calculate a _minimum_ figure for
the age of the oceans, but it must be remembered that the results
based on an assumption, which is merely a uniformitarian guess, are
themselves equally uncertain. If the igneous rocks had lost all their
sodium, we have already found that the annual supply would be 60
million tons, and the sediments would be destitute of that element.
The age would then be at least 210 million years. The contradiction
between this minimum figure and the maximum deduced from the total
unchloridised sodium, viz. 180 million years, is explicable on one or
both of two further possibilities, that the average exposure of igneous
rocks in the past has been greater than that of the present day,
or that the amount of primary sodium now being derived from the
sedimentary rocks is considerably above the average for the past. Our
interpretation of the data is made particularly difficult because
of the widespread covering of glacial detritus, rich in felspathic
constituents and easily eroded.
The contradictory results become still worse when it is remembered that
the rate of denudation—one foot in 8600 years—is probably too high to
be safely applied to areas of igneous rocks. Analysis of river waters
draining such areas indicate a rate which is only about a half of the
above. But even this correction is surpassed in importance by one
arising out of the possibility that the present standard of measurement
may be misleading in not being typical of geological time. The mean
height of the drainage areas is subject to much fluctuation. By
denudation it is steadily reduced; by earth movements, and particularly
during the periods of intense diastrophism, it may become unusually
great. It seems not unlikely that we are now near a period of extreme
continental elevation, and that the geological processes which are
thereby quickened up, cannot be accepted as affording a true standard.
This suggestion will be dealt with further in the next chapter.
The high figures—210-340 million years—given above, must not be
supposed to possess any serious value. The whole discussion merely
serves to betray the uncertainty of the method and the doubtful
applicability of even the most accurate data. For the present we can
only conclude that our knowledge of the part played by sodium and
chlorine in the constant redistribution of the materials of the earth’s
crust is still lamentably imperfect, and that quantitative deductions
drawn from it must be regarded as being purely provisional.