THE RIVER SYSTEM OF CONNECTICUT 479 



over-refinement to introduce a correction of this nature, and they 

 are, therefore, left straight. When all important trough lines 

 had been thus represented, the map was reduced by photogra- 

 phy and the etching of Plate I produced after the important 

 rivers had been made a little heavier in order that their course 

 might be apparent. These details of the study have been noted 

 because the dangers of introducing the personal element while 

 drawing the course of a river with any theory of its orientation 

 in mind are verv great. The map conveys a wrong impression, 

 therefore, chiefly in its slightly exaggerating the volume of cer- 

 tain streams which it was necessary to draw with heavier lines in 

 order that their correspondences might be apparent. 



The first trough lines to impress the observer are those desig- 

 nated a-^, a^, ^3,(^4, upon the map, lines which trend approxi- 

 mately N. 44° W., and which include the lower reaches of the 

 Housatonic, of the Connecticut below Middletown, of the lower 

 Willimantic and the Shetucket, and a stretch of the Quinebaug. A 

 less-marked trough line between a^ and a^, would include impor- 

 tant bends of the Naugatuck and Quinnipiac rivers. The sharp 

 bend of the Connecticut River at Springfield and Patchoug River, 

 tributary to the Quinebaug, may indicate the course of another 

 trough line in this series. Most noteworthy of all lines in this 

 series, however, is a 2, since the lower stretch of the Connecticut 

 (some twenty-five miles long) is continued in the zigzag Sebethe 

 River, so as almost to connect with the stretch of the Farming- 

 ton River above its sharpest bend at Farmington. The direc- 

 tion of the lower stretch of the Connecticut is of especial inter- 

 est because the river at Middletown deserts the softer Newark 

 sediments to flow across the crystalline uplands, a peculiarity 

 which has been explained by Professor Davis' through conform- 

 able superimposition, the stream being supposed powerful enough 

 to maintain an initial course along this direction during the rise 

 of the uplands, at which time the harder gneisses were discov- 

 ered. The same hypothesis has been offered by KiJmmel^ to 



' Davis : The Triassic Formation of Connecticut, Eighteenth Ann. Rep. U. S. 

 Geol. Siirv., Pt. II, 1898, p. 165. 



' KuMMEL : Some Rivers of Connecticut, Jour. Geol., Vol. I, p. 379, 1893. 



