24 
estimating the temperature for his glacial epoch this element 
is omitted altogether. It is plain, however, that it must then 
have been not much less than it is now. The contour of land 
and sea was nearly the same as at present in the Boulder Drift 
period, and the Atlantic basin had nearly its actual outline, 
and reached as far to the north. The strength of the current 
must depend on the contrast between the heat of the southern 
summer and the cold of the northern winter, so far as these 
were directly dependent on the sun. This would be only 8 
per cent, less than it is now. On the other hand the current 
would be greater in the summer half of the year*, and serve 
more fully to blot out the ti’aces of the cold of the previous 
winter. The general result would be aix increase of summer 
heat and winter cold, each about 28° at the most, but probably 
much diminished by the equalizing effects of aerial and ocean 
currents. 
42. Another element has still to be considei'ed. In 
Mr. Croll’s Table, p. 320, vol. iii., the longitude of the pei’i- 
helion at the date B.C. 210,000 is stated to be 144° 55'. From 
the last entiles it seems plain that this amount has reference 
to a fixed and not a movable solstice or equinox, and is the 
change resulting fi’om the progression of the apsides alone. 
The change from precession for this same period, at the present 
rate, would be eight complete circuits and 46° 5G'. Hence the 
true longitude of the perihelion, on this view, would be 144° 55' 
— 40° 50', or just 98°. Thus the northern summer solstice, as 
it is now, would be nearly in aphelion. This is precisely the 
opposite condition to that which forms the basis of Mr. Croll’s 
theox-y. We need to go backward or forward 10,000 years, to 
have the winter solstice in aphelion, when the excenti’icity is 
•0497 or *0569. In the former case the midwinter increase of 
cold would be only five-sixths of Mr. Croll’s estimate, when 
his other data are retained, or the decrease, which has been 
induced from 3 7°* 7 to 28 0, 7, would bo further reduced to 
23 0, 9, or the midwinter temperature by the rulo be 1 5°*1, which 
is higher than the temperature of Canada. 
43. The main principle involved in Mr. Croll’s theory is 
that the cold or hot state of each hemisphere is detennined 
chiefly by its midwinter temperature, and this in turn by the 
simple ratio of the direct solar heat then x’eceived, tho excess 
over the mean temperatui’e of space, or — 239° F., being 
determined by a simple rule-of-three calculation. And sinco 
tho winter northern solstice is now vei’y near the perihelion, 
the present excess abovo the averago value, when combined 
with the deficit at other periods, results in a very considerable 
disproportion. The ratio, accoi’ding to Mr. Ci’oll, 850,000 
