146 
ON THE GOTHAM MARBLE 
lines of least resistance. Thus began the “ trees.” On 
the “ trees ” the “ branches ” were laid down by a process 
analogous to “hedge-structure.” The “branches” differ 
in no respect from the “ hedge.” Owing to the greater 
tendency to contraction above, the “ trees ” spread out 
at the top. 
Wherever in the silt there was a nodule of chemically- 
formed calcite, the precipitate, seeking it out, formed a 
tiny concretion round it, leading to the production of 
“ islands.” 
There is no difficulty in understanding the precipitation 
of the carbonate at the bottom of the silt if we remember 
that that was probabty semi-fluid. Similarly oxalate will 
deposit on calculi through a layer of mucus. 
When “ summer ” returned, it found a set of hard 
polygonal walls, arranged like a honeycomb, with soft silt 
between, resting on the basement beds below. Now as the 
silt in the Landscape Stone was very calcareous, very fine, 
surrounded by calcite crystals, and moist, it was inevitable 
that it should more or less recrystallize. In doing so it 
must contract. The polygonal walls kept it from con- 
tracting laterally, so it had to contract chiefly downwards, 
rising however to the “ trees ” or walls at their tops. This 
formed the “ atmosphere,” and explains why the rising to 
the “ trees ” is chiefly at the tops, and also why the “ atmo- 
sphere” is always the least typical in texture of the constitu- 
ents of the Landscape Stone. 
When, therefore, chemical precipitation took place in 
the new “ summer,” it found a surface like the top of a 
honeycomb, the septa being very concave. To this the 
now abundant precipitate had to fit itself. It did so, in 
parallel layers, and could only partially obliterate it. There- 
fore it presented the characteristic hummocky upper surface 
of the Landscape Marble. It will be remembered that we 
pointed out that these hummocks were not rounded bosses, 
