the prevailing direction of magnetisation as they cooled. But, and
this was the crucial point, if every now and then the polarity of
the Earth's magnetic field flipped over, then the material being
extruded at that time would acquire a reversed magnetisation. Over
millions of years, the sea-floor basalts would record the orienta-
tion of the Earth's magnetic field like a bar code. Although this
was an extraordinary concept, once it was proposed, evidence for
reversals of the Earth's magnetic field throughout geological
time was rapidly found in thick onshore lava flows, and the 'flip-
ping' of the Earth's magnetic field quickly became an accepted
fact. Furthermore, it was argued, the bar code could be used as
evidence in favour of the new theory of 'sea-floor spreading' that
had been put forward a few years earlier.
Supporters of sea-floor spreading proposed that as molten
magma welled up from the depths along the mid-ocean fissures,
it pushed the previous flow apart so that half would move to either
side of the ridge, slightly widening the ocean each time and even-
tually creating new sea floor. After this had occurred many, many
times the original flow could finally end up thousands of miles
away from the ridge, and the two continents, originally part of
the same landmass, would be thousands of miles apart.
As each new lava flow took on the orientation of the prevail-
ing magnetic field, a new 'stripe' in the bar code occurred every
time the field flipped. But although magnetic reversals certain-
ly seemed to support the theory of sea-floor spreading, without
'evidence' the theory remained, like continental drift, locked in
the realms of speculation.
The evidence was found when they learnt how to 'read' the
bar code - and this was done by dating the stripes. The isotopic
system used for this dating was the decay of potassium to argon.
In 1932 Holmes had hoped that one day the decay of potassium
to calcium would revolutionise his work, but it was not until
1948 that Alfred Nier demonstrated the dual decay of potassi-
um to both calcium and argon, and it was the latter which would
be used for dating purposes.