742 H. M. BANNISTER 
that are known to have been transported distances of from 500 
to 1000 miles could not have traveled faster than the main body 
of the ice, and must, as we know from the evidences they bear 
of retardation and friction, have traveled much more slowly. 
Even allowing the extravagant estimate of two feet per diem for 
the ice movement throughout (and the recent investigations of 
Chamberlin and others on the Greenland ice-cap have demon- 
strated that this is an improbability) we can demonstrate that 
a single invasion competent for the transportation of a single 
erratic from its northern source to the southern limits of the 
drift would have required a period of from 15,000 to 20,0Q0 
years.* If we admit, as is more reasonable, that the average 
ice motion was much less than this,—probably not over a very 
few inches per diem, we will have to more than quadruple this 
estimate. Taking into account, however, the inevitable conclu- 
sion that the duration of a single ice invasion was not limited to 
the conveyance of a single erratic or simultaneous group of 
erratics, and that there were, in all probability, several of these 
invasions with intermediate periods of sufficient length to allow 
the development of extensive forests and the accumulation of 
heavy deposits of vegetable mold, indicating a lapse of proba- 
bly many thousand years, we are compelled to multiply the 
above figures by an indefinite multiplier. The outcome in any 
case is that the brief duration allowed for the glacial period by 
some recent authorities is absolutely incompatible with the evi- 
dence of erratics, according to the land ice or glacier theory of 
the deposition of the drift. 
tT have not in my argument taken account of the slowness of advance of the ice 
border which must also be considered in calculating the total. Opposed to each year’s 
advance there must have been a summer’s melting, and judging from the evidence of 
the Greenland ice-cap, this latter element must have been of considerable importance. 
The Greenland ice-sheet hardly gains at all upon the unoccupied land even in North 
Greenland, the ice surplus all escaping by the few glacial outlets which are much less 
active than those in South Greenland. In the case of the Laurentide glacier trench- 
ing upon a fairly temperate region with a long summer, the estival melting must have 
been quite marked. The additional transportation by flood torrents, etc., can be esti- 
mated by the extent of the till deposits as compared with glacial striz, which even in 
southern Illinois are reported as found on the underlying rocks nearly to the southern 
margin of the drift. 
