154 TJie American Geologist. March, isfi^ 
zling- parts of the analysis are: Si (>- — 12.946 per cent and 
\rh O — 17.848 per cent. Silica is high and apparently out of 
place, while IL O is low for the sulphates. This substance 
forms abundantly on the exposed shales as greenish yellow 
rather compact and somewhat globular masses out of the 
water which is constantly dripping from the lower beds. 
Whenever these masses become detached or are subjected to 
evaporation the efflorescence is noticeably different in charac- 
ter. It is white and porous or frost-like, and presents the 
usual appearance of the sulphates formed upon exposed mar- 
casite nodules. A bitter taste to the shales at the contact op- 
posite St. Croix Falls was noted by the Wisconsin geologists" 
and ascribed to the formation of sulphates. A complete chem- 
ical analysis of the white substance has not been made. 
Explanation of Plate XII. 
Fig. I. Sectiofi of Volcanic Tuff. 
The figure is from a microphotograph of a section of the volcanic 
tuff from Taylor's Falls. Diabasic characters are shown by the darker 
grains in the figure, and one fragment especially at the right side ex- 
hibits a coarser texture than is usual. Several grains near the lower 
margin of the field are devitrified glasses. In grains of this character 
flowage is sometimes prominent. The light colored fragments through- 
out the field are now chiefly ciuartz. But these almost all show their 
secondary character by the penetration of actinolite needles which pro- 
ject in beautiful clusters. Finer fragments of a more angular outline 
lie between the larger grains. 
Fig. 2. The St. Lawrence Sliales. 
The figure is reproduced from a photograph of a hand specimen 
obtained at Osceola Falls. The darker portions of the figure represent 
quartz sand; the light wavy threads and bands are greenish clay shale. 
A study of this specimen and a comparison with others of different 
localities, especially those representing phases of the dolomites, and 
also a series of chemical tests, completed since writing the paragraph re- 
ferring to this figure in the text, altogether have led me to ascribe not 
a little of the irregularity of banding in the shales to the removal of sol- 
uble constituents subsequent to their original deposition. The St. Law- 
rence formation in some localities is a dolomite. A theory of the origin 
of dolomites as maintained by Hall and Sardeson in their paper on 
"The Magnesian Series of the Northwestern States" (loc. cit.), argues 
the removal of calcium carbonate from rocks at a greater ■ rate than 
magnesian carbonate. The result is a limestone growing by continuous 
*Geology of Wisconsin, vol. Ill, 1880, p. 418. 
