SNOW 185 
ture, yet suggested a variety of natural objects, 
as feather-mosses, birds’ feathers, and the most 
delicate lace-corals, but the predominant anal- 
ogy was with ferns. Yet they seemed to assume 
a sort of fantastic kindred with the objects to 
which they adhered: thus, on the leaves of 
spruce-trees and on delicate lichens they looked 
like reduplications of the original growth, and 
they made the broad, flat leaves of the arbor- 
vitee fully twice as wide as before. But this 
fringe was always on one side only, except 
when gathered upon dangling fragments of 
spider’s web or bits of stray thread: these 
they entirely encircled, probably because these 
objects had twirled in the light wind while the 
crystals were forming. Singular disguises were 
produced: a bit of ragged rope appeared a 
piece of twisted lace-work; a knot-hole in a 
board was adorned with a deep antechamber of 
snowy wreaths ; and the frozen body of a hairy 
caterpillar became its own well-plumed hearse. 
The most peculiar circumstance was the fact 
that single flakes never showed any regular 
crystallization: the magic was in the combina- 
tion ; the under sides of rails and boards exhib- 
ited it as unequivocally as the upper sides, indi- 
cating that the phenomenon was created in the 
lower atmosphere, and was more akin to frost 
than snow; and yet the largest snow-banks 
