2Q0 W. M. DAVIS — GLACIAL SANO PLAINS. 



/. *cal and Temporary Growth of Sand Plains. — A corollary of the rapid 



iwth of the delta plains compared to the retreat of the ice is, thai the 

 growth of delta plain.- was a local, temporary, and spasmodic operation ; for 

 if it had been general, persistent, and continuous, the plain- must have been 

 of vastly greater extent than we find them. It is true that, in front of the 



at terminal moraines, there is a wide-spreading; sand plain; hnt here, 

 however intermittent its growth may have been, its locus of deposition was 

 maintained within narrow limits for a long time. Such was not the case 

 with the sand plains that are dotted over New England; they were formed 

 as the i<e w;i- on the whole retreating; and yet. in spite of their rapid 

 advance compared to its retreat, they occupy hut a -mall part of the 

 country — not more than a twentieth and probably much less. The stand- 

 ing water in which they were built was seldom completely tilled tip, lor their 

 frontal .-lope- commonly descend into meadow-, often of large extent. 



In searching for the cause of the local character and brief duration of 

 their growth, we can hardly expect to find it in the cessation of outward 

 drainage from the retreating ice-sheet, or in the discharge of saud-laden 

 stream- at one time and clear-water streams at another. A more probable 

 explanation looks to the variation in the point of discharge of the Bub-glacial 

 Streams. The larger rivers were presumably fixed, hut the smaller ones 

 must have frequently changed their courses. Winn they discharged into 

 valleys sloping away from the ice front, the valleys became clogged with 



ivel and sand, stretching far down stream, to lie terraced later on; hut 

 when they discharged into valleys of northward slope they were ponded 

 hack, ami their deposits were concentrated in their deltas, until a change in 

 the point of escape was made, when similar processes went on elsewhere. 



It is not uncommon to find the frontal lobes of a .-and plain lying on 

 kame mound- and e-ker ridges of earlier origin, as at the BOUthwest front of 

 plate •'!. The same relation must often occur within the plain. It finds 

 illustration in a valuable section on the Belt Line of the Boston and Albany 

 railroad, a mile BOUth of A n In inula I e, which -hows I lie triangular outline of 

 a Btony e-ker buried in the fun Bel -and lied- of a plain. Tin- edge of the 



must have been south of this point when the esker was formed, and th 



of it w hen the sand plain was built; and between these two date- there must 

 have I.e. n a time of very -mall deposition hereabouts, for kick set beds are 

 wantin 



The coarse, ■_ ravelly character of the top-set beds of saud plains is a natural 

 result of the continued selective process tli.it must have gone on over the 

 :ace during their deposition. The gravelly beds represent the residual 

 material left in the beds of shifting and branching delta stream-, the _ 



pari of the material of liner texture having gone forward to build out the 



front ..f the ddia. In the same way the co irs< . water-worn material of the 



with it- frequently loosi arrangement and very imperfect stratifica- 



