MORAINES OF RECESSION 433 
The general application of this idea to the oscillations that took 
place is plain on a comparison of the course of any two contig- 
uous moraines of the series from their apexes in the center of 
the valley to their turning point in the interlobate. The Fort 
Wayne and Defiance moraines are nearly fifty miles apart at 
their apexes, but they converge as they rise toward the north- 
east until they are only eight or ten miles apart a few miles 
beyond Adrian. On this basis it would be expected, further, 
that the moraines themselves would be comparatively wide and 
flat where the amplitude of oscillation was great and vice versa 
Here again there is some evidence of agreement of fact with 
theory. With the period of oscillation at 10,000 years and the 
amplitude at 100 miles, which appears to have been its approx- 
imate measure after passing Fort Wayne, the five-mile moraine 
belt would be occupied by the ice-front about 1345 years, and 
a ten-mile belt about 1920 years. 
THE CHARACTER OF THE GLACIAL OSCILLATIONS AS REVEALED BY 
EES DRE AGS 
Assuming that the moraine series was produced by a climate 
oscillation it becomes a matter of the highest importance to 
discover if possible what the character of that oscillation was 
Did the ice-front merely retreat and halt in the simplest possi- 
ble rhythmic fashion, or did it follow a more complicated move- 
ment of alternate retreats and readvances with halts between ? 
It would be expected that the way in which it was built would 
make some difference in the form or shape of a moraine. If 
the moraines took any dominant or common form, and if that 
shape corresponded to one that would result theoretically from 
some particular process, it would be fair to presume that that 
had been the method of their building. 
At a first glance it would appear that the moraines show no 
recurrent features that are particularly suggestive in this respect, 
and it must be admitted that many of them, perhaps the major- 
ity, do not. At least they do not show such features clearly 
enough to be readily recognized. But there are some of the 
