THE ROXBURY CONGLOMERATE SERIES. 45 
formation, beds of sandstone and slate alternate, becoming thinner and thinner. 
At first the alternations show some approach to regularity but they are not 
regular enough to be certainly classed as seasonal. About thirty feet above 
the tillite a pronounced regularity begins and from this point upward the 
banding is so regular that the seasonal theory of deposition is very plausible. 
(See Plates 10 and 13). Not until the sandstone layers become very fine does 
the banding show few irregularities, but wherever currents were strong enough 
to transport sand, irregularities must be expected. About 100 feet above the 
tillite the banding is extremely regular with fine layers of sand sometimes no 
thicker than the diameters of the sand grains and rarely over one eighth of an 
inch in thickness. Where these sandy layers are very thin, they sometimes 
appear to come to an end, but it would be strange if this were not so. . It is 
inconceivable that currents of water should be so constant through different 
seasons as to always deposit sand where such extremely thin layers are in ques- 
tion. The wonder is, rather, that there was enough constancy in the condi- 
tions of deposition to deposit such thin layers of sand in the remarkably regular 
manner displayed at this place. 
The seasonal units are here from half an inch to an inch in thickness, but 
rarely an inch. (See Plate 13, fig. 1). This fine banding is the uppermost 
outcrop at the southern Squantum exposure. 
At this point one should revert to page 6 and compare the description 
there given with de Geer’s description of the transition from coarse to fine 
beds in the Pleistocene deposits of Sweden. It is difficult to avoid the conclu- 
sion that the conditions of deposition in the Pleistocene deposits and these 
Permo-Carboniferous deposits at Squantum were the same. 
After a thorough examination of the southern exposure a similar study of 
the northern exposure revealed the same gradual disappearance of the sandy 
layers in the transition from tillite to slate. Here, however, deeper water pre- 
vailed, for there are only thin beds of conglomerate alternating with slate visible, 
and the change from coarse to fine is more rapid. I found beds of very fine 
conglomerate, about fifteen or twenty feet above the tillite, alternating with . 
beds of fine laminae of slate. The conglomerate layers, with rounded and sub- 
angular fragments averaging one sixteenth of an inch in diameter, were about 
six inches thick, and the series of slate laminae of about the same thickness 
(See Plates 4 and 6). These comparatively thick deposits came to an end 
after a dozen or so alternations and a series of banded slate deposits with occa- 
sional sandy layers came in. About twelve feet above the thick alternating 
