ST. LAWRENCE COUNTY. 339 



Commencing at the southern tier of towns, a large extent of granite and gneiss is spread 

 out, and forms the surface rock from Fowler southwest to Antwerp. The first two or three 

 miles, this mass of primary is gneiss ; but as it passes south or southwest toward Antwerp, 

 it becomes gradually a well characterized granite. This range is barren of interesting mine- 

 rals ; and for the whole distance which it may be traced, it presents only a few imperfect 

 ^crystals of black tourmaline and garnet. It is rather coarse, and appears to be intermediate 

 in kind between the granite of New-England, and the white albitic granite associated with 

 primary limestone. 



Connected with this mass is a very fine bed of graphic granite, one or two miles southeast 

 of the village of Gouverneur. The feldspar of this variety is splendent upon a recent frac- 

 tured surface ; it is a fine handsome rock. The feldspar itself is suitable, without doubt, for 

 porcelain. This mass is also probably a continuation of the granite at the phosphate of lime 

 locality, though it cannot be traced continuously to that place. Upon the western borders, it 

 is penetrated by beds and veins of limestone, if these western beds belong to the same mass 

 of granite which extends from Fowler to Antwerp. The central part of it docs not furnish so 

 many beds and veins of limestone as are found upon its skirls towards the west and south- 

 west. I have brought this locality of granite before the reader, more particularly for the 

 purpose of pointing out one in which he may observe the j^assages of one rock (as they are 

 considered) into another ; the northern wing or prolongation being in parts decidedly gneiss, 

 which, as it ranges west or southwest, passes clearly into granite. 



Pursuing this rock in northwest and northeast directions, we find ridges or beds at nume- 

 rous places in Gouverneur and Dekalb, each of which maintains the same general characters 

 already desci-ibed, presenting variations from granite to gneiss. In Hermon, it fonns, over 

 a wide extent of surface, the principal rock ; and still further east, in Edwards, it becomes a 

 very important one. In the towns which succeed, as Canton, Russel and Pierpont, granite is 

 largely developed. In the latter place, however, there is a greater preponderance of horn- 

 blende than farther south ; and here I observed that it contained hypersthene. The primary 

 of this part of the county, too, is more distinctly gneiss, mingled with hornblende, as in 

 Franklin county. We do not find the same relations to primary limestone ; at least it is far 

 less abundant, and when it occurs, the beds are not so extensive. 



According to my observations, the peculiar interchanges of one rock with another, and the 

 constant connection with primary limestone, is confined mostly to the following towns, namely, 

 Hammond, Rossie, Fowler, Gouverneur, Dekalb, Hermon, Russel, Edwards and Canton. 

 In all which I have given in this list, there are the same associations, first of the rocks, then 

 of the imbedded minerals. The intermixtures diminish towards the northeast ; and hence we 

 find at Pierpont, and the towns towards Duane in Franklin county, a prevalence of gneiss 

 and hornblende : they become, of course, darker colored ; and in the iron mines of Duane, 

 we find some hypersthene, but it is never largely developed. Trap dykes become much more 

 frequent in the region referred to ; while in Rossie, Dekalb, Gouverneur and Hammond, the 

 ordinary greenstone in dykes rarely occurs ; and in saying rarely, I mean to put the expres- 



