PLATE 4. 
Figure 1— Wipr Banpine In Cuay anp FINE SAND, Two Minus Norrn oF Hanover, N. H., 
ON THE VERMONT SiDE OF THE CoNnNECTICUT RIVER. 
Till lies from about fifteen to twenty feet below the section here shown. Above the till comes 
sand of coarse texture, alternating in no markedly regular manner with finer sand. For about ten feet 
these alternations can be traced. Then comes an interval of from ten to twelve feet which shows no 
good exposures. The first clay appears about twenty-five feet above the till in a layer about six inches 
thick. In the view this layer appears as a dark band at the bottom of the cut made to the left of the 
little girl. Then come three feet of very fine sand, above which another dark layer of clay appears, about 
five inches thick. Above this clay layers of fine sand about ten inches thick may be seen, and then 
another layer of clay about seven inches thick. The next layer of clay may be seen at the extreme 
upper left hand corner of the plate. From this point upward layers of silt and clay alternate in ever . 
decreasing widths and finally at the top of the deposit the seasonal layers are not over # inch thick. 
The wonderful uniformity of the change from thick to thin bands which this remarkable deposit shows 
from bottom to top may be explained most satisfactorily, according to present knowledge, by a 
retreat of the glacier. In this case it would appear that the retreat was relatively rapid. See Plate 5, 
fig. 1, for the upper part of this exposure. 
Figure 2.— ALTERNATING COARSE AND FINE, WIDE BANDs aT SQUANTUM Heap, Mass. 
Lying on the tillite at Squantum Head come transition beds of alternating slate and conglomerate 
beds which show. considerable regularity of interval as compared with the transition beds on the tillite 
in other places. The conglomerate in the first beds is fairly coarse, containing pebbles with an average 
diameter of about one inch and these first beds are about one foot thick. Alternating with these con- 
glomerates are layers of slate about four inches thick. About ten feet above these layers, come the 
bands shown in this view. The hammer near the middle of the picture indicates the scale. The head of 
the hammer rests on a bed of very fine conglomerate about six inches thick. The shadow of the handle 
of the hammer covers the overlying layer of slate. Several layers of alternating slate and fine con- 
glomerate may be seen above. From a study of the slate layers here shown (see Plate 6, fig. 2 for a 
detail view of the slate) it would appear that the winter seasons of this time were not continuously cold, 
but broken by thaws somewhat as at present. Warmer conditions must have set in, for the glacier 
was retreating. In this case, the conglomerate layer would mean a short and very warm summer with 
longer falls and springs than at the present time. 
