s interesting region. Going one day in a southwest direction from
Minnekahta, to look for fossil cycads, I stumbled on a block of
sandstone with a rather smooth surface on which were some peculiar
markings, such as I had never seen figured or described. The rock was
evidently a block from the Dakota sandstone. Its smooth upper surface,
which represented a bedding plane, was covered with a thin coating of
silt or fine clay which adhered to the block. The markings were in this
clay. They were straight, shallow grooves from one-half to two inches in
length, and from one-sixteenth to one-eighth inch in width. They were
joined into patterns in which some sprang out from the sides of others
and again themselves sent out other branches. Some crossed each other. I
noticed that there was a quite uniform angle of divergence in these
branches, and I was able to make out that this usual angle was about
sixty degrees. I also noted that the grooves narrowed to sharp points.
Somehow, immediately I concluded that the cracks were the result of ice
crystals, and I at once saw the propriety of frozen water having existed
in an age during which deciduous trees began to appear. This was theory.
We have since that time learned to know that cold climates far antedate
the coming of the dicotyledons.

As I had no suitable photographic equipment, I took pains to make
accurate drawings of a part of the pattern as it appeared on the rock.
My original drawing is shown in Plate I. A brief description of these
markings was later furnished in the Scientific American, of February 19,
1895.

It took me some years to find any similar markings again. In the early
spring of 1903 I had occasion to make a visit to Mexico, when I spent a
half day in Ojinaga, which is a little village south of Presidio, in
Texas, on the Mexican side of the river. Some sidewalks in this little
village are built of flags of limestone belonging to the Eagle Ford
formation. To my great delight I found some of these slabs having
precisely the same kind of markings that I had noted on the sandstone in
Dakota. Naturally I attached some importance to the fact that the Eagle
Ford corresponds quite closely in age to that of the Dakota sandstone.
Both were made at about the beginning of the upper Cretaceous age. I
noticed here a considerable variation in the closeness of the patterns
of the markings. Occasionally they were found as separate single lines,
several inches removed from each