1 8 CONNECTICUT GEOL. AND NAT. HIST. SURVEY [Bull. 



and sillinianite irregularly distributed. The hills that rise above 

 the Housatonic valley floor between these two nearly connect the 

 types petrographically as well as areally, and there seems to be 

 little question that they belong to the same formation. 



The Canaan Mountain type of Berkshire is the one that laps 

 over the Green Mountain anticlinorium and that comes into contact 

 with the Hartland schist both north and south of Bantam Lake. 

 The latter two rocks are quite dissimilar and are believed to have 

 suffered a different amount of metamorphism. In fact, if the cor- 

 relation be based upon this region alone and the western series, 

 discontinuous with the Berkshire to the North, can be proved with- 

 out question to be the Berkshire, then the Berkshire must be called 

 the older rock. 



The Brookfield diorite, a medium to coarse-grained greenish 

 diorite and quartz diorite with gneissoid and even schistose phases, 

 cut by later dikes of diorite porphyry, granite, pegmatite and dark 

 green biotite hornblendite, and including stringers of the Hart- 

 land schist, forms the central part of the area. It stretches from 

 the outskirts of Litchfield village on the northeast to Mount Tom 

 on the southwest and underlies most of Bantam Lake. At its 

 northwestern extremity on Mount Prospect it is intersected by a 

 comolex set of intrusives, gabbros and related rocks correlated by 

 W. H. Hobbs^ with the Cortland Series near Peekskill, New York. 

 That area is being studied in detail by E. McKnight and is not 

 described in any detail in this bulletin. 



At the southwest extremity of the diorite there lies an irregular 

 area of green and white hornblende gneiss. It forms Mount Tom, 

 Little Mount Tom, and Mount Rat, and projects about one mile 

 to the southwest of the latter hill. This rock, known as the Mount 

 Tom hornblende gneiss, forms dikes in the Hartland schist along 

 the northwest slope of Mount Rat and has a much more perfectly 

 developed foliation in the dikes and along its edges than near the 

 center of the mass. It represents an intrusion that suffered 

 deformation before its solidification was completed. 



It will be seen that this surface distribution of rocks eliminates 

 the area of Rough quag quartzite featured on the geological map of 

 Connecticut of 1907^ as underlying Bantam Lake and covering 

 considerable areas both to the east and west, south of the lake. 

 The reason for this change will be discussed in detail further 

 along. Briefly, it is that the ledges that outcrop over this area are 

 quartz schist identical with many of the layers forming an impor- 

 tant part of the Hartland schist series as exposed in the tunnel. 



The northwest section of this region is underlain by the 

 Thomaston granite gneiss. It occurs throughout the neighboring 



1 On Two New Occurrences of the "Cortland Series" of Rocks within the State of 

 Connecticut. W. H. Hobbs, Festschrift zum Siebzigste Geburstage H. Rosenbusch, 

 ^S-48 Stuttgart 1906. 



2 Gregory and Robinson, Bull. No. 7. 



