496 
NATURE 
[APRIL 14, 1923 
spasmodic instead of continuous, the general result 
the order of 10o-®° dynes per sq. cm., whereas to 
elevate the Rockies something like 10° dynes per 
sq. cm. would be required. 
A hypothesis that may be of use in accounting for 
continental drift, if the latter is considered to be 
indicated by the geological evidence, is based on 
Jeans’s proof (Proc. Roy. Soc. 93, 1917, 413-417) 
that the earth is stable with regard to first harmonic 
deformations. The fact that most of the land is in 
one hemisphere indicates that a first harmonic de- 
formation exists, and must therefore be tending to 
die down; the only possible means of destroying 
the asymmetry being for the continents to break 
up and spread out so as to get as far apart as possible. 
If, then, we are prepared to admit that the continents 
were once all united into one mass, it is probable 
that they would have broken up and separated 
widely, since the stresses in them must have been 
comparable with the pressure at sea-level due to 
the weight of a continent, which is at any rate a 
moderate fraction of the strength of rocks. Wegener’s 
suggestion that India has moved towards the main 
mass is, of course, inconsistent with this hypothesis. 
The possibility that the continents were formerly 
united has been regarded by Mr. Crook (NATURE, 
February 24, p. 255) as in harmony with Osmond 
Fisher’s theory of the origin of the Pacific. The 
latter theory, however, is open to a serious objection. 
The birth of the moon on the resonance theory 
would require a violent distortion of the earth, 
sufficient to shatter into fragments any crust that 
might have already been formed, and these would 
distribute themselves symmetrically over the liquid 
interior at once instead of waiting a thousand million 
years to do it. 
Prof, Sollas’s suggestion, mentioned by Dr. Evans 
in Nature of March 24, p. 393, that there are traces 
in the earth of the incipient formation of a second 
satellite, is not in quantitative accordance with the 
resonance theory of the origin of the moon. It is 
practically certain that the earth-moon system, when 
combined into one body, did not rotate sufficiently 
fast for instability, but it is just possible that it could 
have rotated fast enough for resonance to magnify 
the solar semidiurnal tide to such an extent as to 
rupture the mass into two parts. If the moon was 
formed in this way, however, it must have taken 
away with it so much angular momentum that the 
earth could never again have approached conditions 
suitable for either resonance or instability. 
HAROLD JEFFREYS. 
St. John’s College, Cambridge. 

The Life-Cycle of the Eel in Relation to 
Wegener’s Hypothesis. 
THE argument in Dr. Wemyss Fulton’s very 
interesting letter in NarurE of March 17, p. 359, 
must be divided into two parts. First, it is pointed 
out that the gradual recession of the east and west 
coast-lines of the North Atlantic Ocean from one 
another would explain in a very satisfactory manner 
the evolution of the amazing migrations of the 
larval eel. Secondly, it is assumed that Wegener’s 
continental drift is the only method of effecting 
that gradual recession. It is possible to concur 
heartily with the first thesis, without admitting the 
second. 
Suess explained the North Atlantic Ocean as 
having been formed, during the later ages of the 
Cainozoic era, by successive foundering of portions 
of a pre-existing land surface. Except that the 
recession of the two coasts would then have been 
NO. 2789, VOL. I11] 

from the point of view of the inhabitants of the sea 
would be just the same as if the two continents were 
drifting apart. While acknowledging our indebted- 
ness to Dr. Fulton for pointing out how the life- 
history of the eel fits in with the other evidences of a 
gradually widening Atlantic, we need not admit that 
these wonderful migrations prove continental drift 
any more than the migration of birds across the 
Mediterranean proves that Africa has drifted away 
from Europe. A. Moriey DaAvIEs. 
Imperial College of Science and Technology, 
South Kensington, S.W.7, 
March ro. 

The Combination between Oxygen and 
Hemoglobin, and the Criteria of Adsorption. 
H2#MOGLOBIN combines with oxygen approximately 
in the ratio of 16,670 to 32, by weight, as was shown 
by Peters (Journ. of Physiol., vol. 44, p. 131). It is 
clear, then, that in solution the particle of hemoglobin 
is very much larger than the particle of oxygen which 
combines with it. If one might assume that the 
densities and shapes of the particles were similar, 
then their surfaces would be in the ratio 64 to 1; 
in any case, and whatever the degree of aggregation 
of the particles, probably only a very small part of 
the surface of the haemoglobin particles can be 
actually covered by oxygen when combination ceases 
at the stage of oxy-hemoglobin. 
This shows that the attraction of hemoglobin for 
oxygen is a highly localised property of the hemo- 
globin particles. For if this attraction were more or 
less evenly distributed over the surface, it would be 
satisfied only to a small extent, when a small part 
of the surface was covered, and at higher concentra- 
tions of oxygen than those which are found experi- 
mentally to give saturation with oxygen, more oxygen 
would be taken up. 
Taking the thermal motions of the particles into 
account does not affect this argument, since the 
movements of the particles according to the laws of 
the kinetic theory do not affect their surface areas. 
Now, if the attraction of hemoglobin for oxygen 
is of such a character that it is satisfied when only 
a small portion of the surface is covered, it seems 
impossible to regard this combination as a case of 
adsorption. 
The criteria of adsorption are perhaps not yet so 
well defined as could be wished, if differences of 
opinion as to whether a given process should be 
classified as adsorption or not are to be avoided. 
I am inclined to think that a process is rightly 
classified as adsorption, if the substance taken up 
the surface continues to be taken up until the whole 
surface is uniformly covered, but not otherwise. 
Covering the surface uniformly is of course meant 
in the sense in which a gas or homogeneous solid is. 
said to fill space uniformly; that is, uniformly to a 
being armed with a microscope to which individual 
atoms are small. 
This definition is both definite theoretically, and 
in accordance with common conceptions of adsorption. 
It is difficult, indeed, to see what other definition is 
possible in the present state of knowledge. It is 
perhaps, however, desirable to state the definition 
clearly ; although, as I feel it must have been present, 
whether formulated or accepted as self-evident, to 
the minds of many workers on adsorption, no sort 
of novelty is claimed for it here. 
A definition of adsorption based on the nature or 
quality of the forces attracting the adsorbed substance, 
5 attics 
~~ 
