1010 
heavier, and climate more oceanic. The shifts of the 
wind zones may have been associated to some extent 
with fluctuations in the ice-covering of the arctic seas. 
References in the works of classical geographers suggest 
that near the beginning of the Christian era Iceland 
was icebound. During the period of the Norse colonisa- 
tion of Greenland there was little ice, but glacial con- 
ditions returned about 1200 and remained predominant 
throughout the ‘Little Ice Age”; they are now im- 
proving. The parallelism is not exact, however, and 
in the past 2000 years it is more likely that the varia- 
tions of atmospheric circulation have governed the 
variations of arctic ice than that the reverse is the case. 
THEORIES OF CLIMATIC CHANGE DEPEND- 
ENT ON EXTERNAL INFLUENCES 
Variations of Solar Radiation. The most obvious 
reason for climatic changes is that the radiation which 
the earth receives from the sun varies with time. It 
has frequently been suggested that ice ages were due 
to decreased solar radiation, either because the sun 
emitted less radiation, or because some obstacle such 
as a cloud of cosmic matter was interposed between 
the sun and the earth. Conversely, the warm periods 
were attributed to increased solar radiation. In 1929, 
however, Simpson [30] argued that glaciation should 
result from an zmcrease of solar radiation. 
Simpson’s theory is, briefly, that glaciation depends 
not so much on temperature as on an excess of snow- 
fallover melting. The total precipitation over the earth 
as a whole must equal the total evaporation, which in 
turn depends very largely on the solar radiation. In 
high latitudes and at high elevations most of the pre- 
cipitation falls as snow. Now consider the effect of two 
cycles of solar radiation (Fig. 1). Starting with a 
minimum of radiation most of the precipitation falls as 
snow, but the total amount is small. As the radiation 
increases, the proportion of snow to rain decreases, but 
at first this is more than counterbalanced by the in- 
crease of total precipitation, while summer melting is 
still unimportant. At this stage there is an accumulation 
of snow, resulting in glaciation. Eventually, however, 
the rise of temperature due to increased radiation 
reaches a stage at which the actual amount of snowfall 
begins to decrease and is exceeded by the summer 
melting. The glaciers and ice sheets break up and a 
mild rainy interglacial sets in. As the radiation passes 
its peak and decreases, the process is repeated in the 
reverse direction, bringing another glaciation, but 
eventually the decrease of precipitation starves the 
glaciers and ice sheets, causing a second interglacial, but 
this time with a cold dry climate, which persists until 
another increase of solar radiation begins a new cycle. 
Simpson considered that the stage of glaciation is 
rather near the peaks of the solar cycle, so that the 
warm moist interglacials were short and the cold dry 
interglacial was long, giving the well-known Alpine 
succession (p. 1005). In low latitudes each solar cycle 
is represented by a single cycle of rainfall, giving 
two pluvial periods covering respectively Gunz plus 
Mindel and Riss plus Wurm. 
CLIMATOLOGY 
The facts do not entirely support this theory. The 
Mindel-Riss interglacial was not cold; at its height 
Kurope and North America were warmer than at pres- 
SOLAR RADIATION 
HEIGHT OF END OF MORAINES 
INTERGLACIAL 
PERIOD 
INTERGLACIAL PERIOD 
a 
© 
Ca 
Wi 
a 
= 
<q 
6 
< 
a) 
oO 
fia 
w 
= 
2 
Fic. 1.—Effect of two cycles of solar radiation on glaciation. 
(Reproduced by courtesy of Sir George Simpson and the Man- 
chester Literary and Philosophical Society.) 
ent. In East Africa the Riss and Wurm were not com- 
bined and Wayland [83] thinks that the first pluvial 
also was double. A third objection, that there is no 
physical reason for the existence of a cycle of solar 
radiation of the order of 500,000 years, has been partly 
met by Hoyle and Lyttleton [13], who showed that the 
passage of the sun through a cloud of interstellar matter 
would cause an increase of solar radiation. 
Willett [34] has pointed out that some of the objec- 
tions to Simpson’s theory are removed if we suppose 
that the Quaternary represents four solar maxima in- 
stead of two, the radiation never reaching the level 
necessary for a warm interglacial, and also that the 
cycle of radiation takes place mainly in the ultraviolet. 
Another theory of solar control was put forward by 
Huntington and Visher [15]. There is some evidence 
that when sunspots are numerous and large, the storm 
tracks are displaced towards the equator and the cli- 
mate of the earth as a whole is stormier, rainier, and 
cooler than at times of sunspot minimum. Since ob- 
servations began, sunspots have varied, not only in 
the well-known li-year cycle, but also over much 
longer periods, reaching a high maximum about 1372 
and being almost absent from 1676 to 1725. Hunting- 
ton and Visher suggested that there have been sunspot 
cycles of geological length, associated with changes in 
the distances of the nearest fixed stars, and that these 
have caused the geological changes of climate. This 
