256 PROCEEDINGS : BOSTON SOCIETY NATURAL HISTORY. 
eastern Massachusetts. Topography, especially in the thinner stages 
of the ice sheet, may have had considerable influence on ice move¬ 
ment in this region. It is possible that the somewhat prominent 
elevation south of the railroad junction at Mayflower Park, though 
of little effect when the ice was at its maximum stage, was, during 
the opening and possibly the closing stages, an efficient aid in the 
deflection of the ice to the eastward. 
The principle that the direction of movement in the continental 
ice sheet was in the direction of steepest gradient and normal to the 
ice margin is the foundation on which the more feasible explanations 
of fanning are based. This tendency, if exerted during the succes¬ 
sive stages of a symmetrical retreat of an ice lobe, is, as pointed out 
by Chamberlain (’88, p. 201-202, fig. 25), an efficient cause for 
considerable variations in the directions of movement, and hence 
also for the distribution of the drift. 
The assumption of successive stages is not, however, essential to 
an explanation of the occurrence of fanning in an ice lobe, for the 
thinning and lateral expansion of the ice consequent upon the 
deployment at its mar yin is amply sufficient to account for the 
moderate expansion of the fans of most boulder trains. The degree 
of fanning in the Braintree train, amounting to some 45°, is proba¬ 
bly too great to be explained entirely in this way. 
The line of morainic deposits extending from the junction of 
the Buzzards Bay and Cape Cod moraines northward to the vicinity 
of Massachusetts Bay, marks the line of junction of the main ice 
sheet covering Massachusetts with the lobe occupying the Gulf of 
Maine, Cape Cod Bay, etc. The ledges from which the Braintree 
train of boulders was derived are but a few miles from the north 
and south line of morainic material mentioned, and were, therefore, 
in a location most favorable to fanning by changes of movement 
during the symmetrical advance and retreat of the ice sheet, as 
postulated by Chamberlain. The more easterly movements would 
thus have occurred either in the opening or closing stages when the 
lobes were more or less distinctly separated from each other, while 
the southerly movement characterized the stages of maximum 
glaciation, when the lobes were confluent, or at least closely ap- 
pressed. The curvature of the boundary is the direct result of 
the curvature of the lines of movement of the ice, which approxi¬ 
mated more and more closely the normal to the margin as this 
margin was approached. 
