108 J. F. CAMPBELL ON GLACIAL PERIODS. 
igneous rocks and by the dip of metamorphic gneiss. Since the 
caves of Elephanta were hewn out of trap, the level of the floor has 
changed; so that water poured on an idol flows in the opposite 
direction from that which the sculptors intended. Mr. Judd has 
shown that great voleanic cones once stood on spots which now are 
in “ the Hebrides.” Their age is proved by fragments of sedimentary 
rocks preserved in sheets of trap and basalt. Where trap rocks cover 
avast area to a great depth in India, it is possible that volcanic cones 
in proportion may have risen high enough to reach the cold climate 
which now breeds glaciers in lat. 27°-28° N., only 10° further north. 
That is one possible explanation for these old Talchir boulders which 
occurs to me. It has been found that water colder than the freezing- 
point of fresh water is at the bottom of the Indian Ocean now under 
the line and north of it. It has been supposed that this cold layer 
of sea-water is flowing northwards from the Antarctic ice, and dis- 
placing surface-water warmed to 80° or even 90° by the sun in the 
Indian Ocean. Surface-currents which flow southwards along the 
African coast prove marine circulation and a water supply equal to 
the waste. It is well known to fishermen and sealers that icebergs 
founder, and that ‘“‘ anchor-ice” grows at the bottom of the sea off 
Newfoundland and the Labrador. Though ice floats, it may be sunk 
by stones or anchored to them. It seems possible that an antarctic 
current of cold water moving northwards at the bottom of the sea 
may roll foundered icebergs, and move sheets of anchor-ice loaded 
with stones, even beneath warm tropical seas. It is possible, but 
very improbable, that sunken ice should travel 73°. Mr. Fedden 
sees nothing in the Talchir boulder-deposit to indicate the action 
of a continental ice-sheet or of a local glacier. He suggests the 
action of ground-ice in shallow water. But these supposed records 
of a cold climate coincide in position with tropical coal-plants. The 
two records do not agree if the events recorded are supposed to haye 
happened near the same sea-level. The case is a problem; but ex- 
isting conditions of climate that best agree with the record are the 
deep-sea cold of the Indian Ocean and the hot tropical climate at the 
surface and on shore. These give a difference vertically equal to the 
difference between the climates of Bombay and the North Cape of 
Europe at the sea-level, or between the existing climates of the 
plains of India and of hills near them which rise to the level of per- 
petual snow in Nepal. (Mr. Fedden’s paper and a rubbing taken 
from a boulder described in it accompanied this paper,with a moraine- 
stone from Ivrea, for contrast.) This exceptional find does not concern 
the recent “ glacial period” or superficial geology ; it belongs to the 
age of the Indian coal. If these boulders are in fact glaciated, they 
prove that the world’s cimate there was somehow cold enough for 
ice to exist within the tropics, while it was also hot enough at the 
same place for the growth of coal-plants—a very long time ago. 
Even a change in the position of the earth’s axis would not explain 
this double record of great heat and great cold. 
25. 18° to 31° N. lat. In the end of September 1876 I landed at 
Bombay and travelled by rail to Allahabad. The rocks seen were 
