WNMPOLMESTS OF “CAUSE OF GLACIAL PERIODS 549 
easier part of the problem. Whatever may be thought of the 
supposed signs of cold periods at other early periods, the evi- 
dences of glaciation in India, Australia and South Africa near 
the close of the Paleozoic era are so abundant, so specific, and 
so well attested, that they cannot be ignored, and any hypothe- 
sis that assumes to account for glacial periods must take serious 
cognizance of these and must meet the strenuous issues that 
spring from their early age and from the mildness of the periods 
following them. It seems to the writer almost equally necessary 
to take cognizance of the salt and gypsum deposits of various 
periods, which imply degrees of aridity in relatively high lati- 
tudes scarcely equaled at the present day. If the atmospheric 
line is followed, it seems necessary to postulate a reduction of 
carbon dioxide near the close of the Paleozoic era—to say 
nothing of other early times — so effectual as to produce glacia- 
tion between 20° and 35° latitude on both sides of the equator, 
a glaciation the deposits of which aggregate a greater thickness 
than those of Pleistocene times, and whose oscillations, marked by 
thick accumulations of coal, were even more remarkable than 
those of the Pleistocene glaciation. If a depletion of the carbon 
dioxide of the atmosphere sufficient to produce this glaciation 
at this relatively early stage in geological history is postulated, 
it is necessary to assign agencies for the reénrichment of the 
atmosphere in carbon dioxide to account for the mild climates 
in high latitudes in Jurassic, Cretaceous, and Tertiary times. In 
short, it is necessary to assign competent operative geological 
agencies which shall produce effective depletion alternating with 
effective reénrichment of the atmosphere from an early period 
in its history down to the present time. If the salt and gypsum 
deposits, and the prevailing red beds, with arkose elements, be 
regarded as the products of exceptional aridity, and if this be 
assigned to the localization and intensification of heat and mois- 
ture due to the removal of carbon dioxide, as subsequently set 
forth, it is necessary to multiply the oscillations from enrich- 
ment to depletion very notably, and to extend the alternating 
action at least as far back as the close of the Silurian period 
