JULY 28, 1899. ] 
the hornblende-biotite areas and in several 
localities are in apparent proximity to gab- 
bro. (jf) Farther north, on the Mississippi 
River, from Two Rivers past Little Falls to 
the valley of the Elk River, are extensive 
exposures of a fine-grained hornblende- 
biotite schist carrying bosses of gabbro and 
lenses of quartz-diorite [J. H. Kloos, Neu. 
Jahrb. fir. Min., 1877, S. 225] and, also 
locally, thickly studded with staurolite 
crystals and garnets. (g) Finally the in- 
teresting masses of epidote granite and 
associated basic eruptives of Western 
Stearns, Todd and Cass counties. 
I have reached the conclusion that all the 
areas enumerated under (a) to (g) above 
belong to the same geologic time division, 
viz., the Keewatin. The clastics, partially 
altered clastics and thoroughly crystalline 
schists in the areas (a), (6), (¢c), (d) and 
(f) belong to a single rock series and the 
granites and gabbros of areas (d), (e), (f) 
and (g) are eruptive through them. The 
staurolite, garnet, quartz-lenses, etc., essen- 
tially contact minerals, bear circumstantial 
evidence of the proximity of eruptive 
masses of granite or gabbro even where such 
masses are not now seen owing to enormous 
subsequent erosion or the covering of gla- 
cial drift. 
Among the considerations upon which 
the foregoing conclusion was reached are 
the following: 1. The quartzose clastices 
and hornblende-biotite schists, which are 
admittedly one and the same rock series 
[Irving, R. D., Fifth An. Rep. Director 
U.S. Geol. Sur., p. 196], can be traced by 
petrographic and structural characters 
through Mahtowa, Barnum and Moose 
Lake in an almost continuous succession 
of exposures from the Thomson conglom- 
erate to the coarser garnetiferous schists, 
carrying quartz stringers and lenses in con- 
siderable profusion west of Sturgeon Lake; 
and these in turn through reported ex- 
posures [Hopewell Clarke, Land Commis- 
SCIENCE. 
109 
sioner, St. Paul and Duluth R. R.] to the~ 
Snake River valley schists filled with 
dikes of granite. 2. The relation of 
the Snake River granite dikes in T. 42, 
R. 23 W., and the granites of Kanabec, 
Mille Lacs, Benton, Sherburne and Stearns 
counties cannot be traced in the field, yet 
their petrographic characters are essen- 
tially alike, and they have always been 
assumed to be the same. 3. The stauro- 
lite-bearing southern border of the Missis- 
sippi Valley schists disappears beneath the 
glacial drift in striking nearness to the 
granites of Stearns and Morrison counties. 
4. Nowhere in Minnesota has this type of 
granite been found intrusive into or through 
the Animiké [for illustration compare Irv- 
ing, R. D., Seventh An. Rep. Director U.S. 
Geol. Sur., pp. 421, 422]; in several places 
in central North America it is reported as 
penetrating and lying upon the Keewatin 
[e. g-, Lawson, A. C., Geoi. and Nat. Hist. 
Sur. Can., 1885, ce., p. 14]. 
Summarizing: The Keewatin of Minne- 
sota, therefore, odcupies a much greater area 
than has hitherto been assigned to it, since 
it underlies the large central region of the 
State. It here consists of two distinct rock 
groups, one a clastic-crystalline and the 
other an eruptive, partly acid and partly 
basic, breaking into and through the former. 
The two exhibit in places a typical eruptive 
unconformity, yet volcanic activity appar- 
ently ceased before overlying rocks were 
laid down upon the intermingled eruptive 
and clastic material. 
The hornblende-biotite granites of cen- 
tral Minnesota constitute enormous erupted 
masses, probably laccolitic in structure, 
which towards the northeast give place 
to a system of dikes which break through 
the schists and cause the greatest strati- 
graphic confusion. It is in this region that 
the schists become thickly studded with 
contact minerals. 
The succession of characters representing 
