began to intcrfon' witli oaoli other. The line of contaet, which 
became a phine as the growth eontiiuied. is marked by the iiiofe 
Hjr less distinct i)lane of separation, the position of which can be 
seen in tiie surface lim^s extending from these channels upward, 
or in the direction of increase of tlie mass. Following these fibers 
upward to their summits, tiiey are seen to spread out like some 
fi'ondescent veget.al)l(^ growths, thus illustrating the tendency of 
inorganic mattiu' to simulate the forms seen in organic. This is 
a familiar fact, and is witnessed in frost on win<low-panes in 
winter and in dendritt:s in the crevices of rocks, as well as in 
■some agates and chalcedony. Some ludicrous mistakcis havt^ 
l)eeri made l>y microscopists in referring such shapes to organic 
H'auses. l)oth in the crystalline rocks an«l in meteorites. 
The principal problem, however, touching thes*' lu'uiatite 
masses, consists in the existence of thest' channels which main- 
tain their forms and in the main also their direction through the 
mass. The}' were supposed by .Mr. <}resley possilWy to owe 
their existc^nce to the removal of some object round which the 
liematite had grown, but it appears to the writer that they date 
from the time of the development of the crystals, and are normal 
and natural. Their courses can l»e seen in plate ill. and their re- 
lations to the growing fibers can be seen in plate ri. While they 
run, in general, altout perpendicular to the fibrous structure, they 
vary from that direction, and their shapes are also various. They 
are always placed in such positions that they lie in or across the 
jiiaiies of contact of two opposite-spreading frondescent growths, 
.ii.nd in .souje cases they are at the lowt^r extremity of such contact 
planers. It appears as if in the first instance they mark the 
vacancies left by the lirst contacts of over-arching growths from 
opposite^ dinMrtions, in the manner of aisles among thickly-set 
drees. The contacting l»ranches then interfered with the fn^e 
circulation of air, or whatever gases there may have be«Mi present, 
and int(n-ru|)ted :iiid permaiu'ntly stopped the development of 
rthose fibers which had what might be called their foliage and 
rtorescence Ixmeath the ov(U--spreading canopy. Once ciiecked, 
::ind air currents established unfavorable to development, tiie 
<;rystalll//mg fontes wi^re powerless to fill the passage-way, and 
they were hc^rmetically sealed ov(!r in' the sub-metallic, nearly 
black hematit(! scale in which all the natural surfaces ex(!ept the 
terminal fronds are encased. 
