3^4 The. Anierica/i Geologist. May, i898 
land. This latter is to be regarded rather as an outlet of the 
valley of the Taunton river and Mt. Hope bay than of the main 
portion of Narragansett bay, with which it is, in fact, only 
indirectly connected. It is, therefore, set aside as having 
little or no bearing upon the level of the waters of the bay. 
The outlet to the west of Conanicut island, known as the 
Western passage, will first receive attention. According to 
chart No. 353 of the United States Coast and Geodetic Survey, 
the passage is narrowest when opposite Fox hill, having there 
a width of almost exactly a mile. The present average depth 
computed from the same source is thirty-six feet. The area 
of its cross section is, therefore, 190,000 square feet, or nine- 
teen times as great as that of the glacial stream which laid 
down the Barrington plain. In other words, the Western 
passage alone could carry ofi the floods of nineteen such 
streams without appreciable increase in the hight of its waters. 
Considering that of the "eight plains mentioned, not more than 
three at the most can be correlated as belonging to the same 
stage of ice retreat, there certainly seems no cause here for any 
increase in the hight of the waters. 
If this is true of the Western passage, it is even more so 
of the Eastern passage. Referring again to the chart, we 
find the narrowest point of this latter passage is along a line 
running southeast from fort Dumpling, the width being 3,300 
feet and the average depth 120 feet. The area of the cross 
section is, therefore, some 400,000 square feet, or forty times, 
as great as that of the glacial stream mentioned. Both out- 
lets remaining open, an increase in the hight of the waters 
could only take place when sixty or more streams of the size 
of the one laying down the Barrington esker entered the bay 
at one time. There is certainly little evidence that such was 
the case. 
It may be argued, however, that many of the glacial 
streams entering the bay are unrepresented by sand plains. 
Granting this to be so, it is yet possible to show that with the 
increased surface slope of the flood consequent upon any 
increase in the hight of its waters, the discharge would rapidly 
assume proportions exceeding all possibility of supply. 
Enormity of Flood Dcnumdcd. — The distance from Green- 
wich cove to the open sea at the southern end of Conanicut 
