SOME GLACIAL WASH-PLAINS. 91 



restoring the glacial conditions, we should imagine the 

 ice-front in contact with the head of the sand-plain, and 

 the northern part of the island covered with ice rising as 

 a gently slo})ing [)lain to the northward. From the front 

 of the ice, rivers emerge laden with gravel, sand, and 

 mud, as is the case with the plains confronting the Mala- 

 spina glacier to-day. 



From the form of the plain on the east, it is thought 

 that the ice-front turned southeastward and ran out over 

 the Nantucket shoals. This interpi-etation is expressed 

 in the accompanying maj) (Fig. 3), and on the general 

 map of southeastern Massachusetts. (See 18, Fig. 7.) 



Martha's Vinef/ard plain. — 'j'he Marlha-sVincyard })lain 

 appears to have arisen in the angular space between the 

 two lobes of the ice-front previously named. From Vine- 

 yard Haven harbor, the ice edge extended southeastward 

 across Chappaquiddick Island in the direction of Nantucket 

 as shown by the ice-contact delineated on the general 

 map (Fig. 7 ) . Topographic signs of this ice -contact exist 

 on either side of Edgartovvii harbor. From Vineyard 

 Haven, the ice front also extended southwestward lying 

 for the greater part of its extent on the highlands of the 

 island. At an earlier period than the time of sand-plain 

 building, it is probable that the ice extended southward 

 of the island ; at least, as Professor Shaler has pointed out,^ 

 the southernmost part of this island and the neighboring 

 island of No Man's Land are till covered. The position 

 of the ice-front in the highlands of Martha's Vineyard is 

 clearl}^ indicated by houlder-belts (16), a type of frontal 

 moraine accumulated on southward slopes where the fine 

 materials were readily washed to the lower grounds. The 

 position of the principal belt is shown on the accompany- 

 ing map (16, Fig. 7). 



I See his report on Geology of Martha's Vine3ar(1, 1888. 

 ESSEX INST. BULLETIN, VOL. XXIX 7 



