By far, permineralization is the most common type of preservation for most fossil bone, and even when petrification has occurred, there is almost invariably evidence that permineralization occurred first (otherwise, there would be no preservation of the original cavities in the bone!). So, if you are wondering what petrified bone looks like, imagine the bone material being replaced by other minerals, sometimes preserving the fine structure of the bone, sometimes not, and the open pore spaces infilled as seen here. I plan to eventually present some truly petrified bone eventually.
In either case, the boundary between the original, open pore space and the replaced material is quite obvious, because of variations in the shape and orientation of the crystals infilling the pores. In the case of the Haversian canals of bone, this is usually indicated by concentric growth of crystals from the inner surface of the canal towards the interior, often with clear radially-arranged crystals and/or layers of different minerals at early infilling versus later stages.
For more information on bone fossilization processes, including illustrations, see Reid (1996) and Hubert et al. (1996).
PPL
XN -- compact (Haversian) bone,
transverse section. 20x magnification. Note the more laminar
arrangement of the Haversian canals towards the margin (arrowed).
The phosphatic bone material is generally dark in crossed nichols. Most of
the canals are infilled with blocky calcite crystals, but a few are
still open.
PPL
XN -- compact (Haversian) bone,
transverse section, 50.4x. The laminar structure of the bone around
the Haversian canals is visible. Black dots in this area are
the lacuna formerly occupied by bone cells (osteocytes). In crossed
nichols, Haversian canals infilled by calcite have a bright, mottled
"rainbow" of colours, wheras unfilled canals are a uniform dark
blue grey.
PPL
XN -- compact (Haversian) bone,
transverse section, 128x. At this scale, the type and shape of the
crystals infilling the Haversian canals can be clearly seen.
In PPL, the crystals are clear and colourless, with a "rough",
high-contrast appearance ("high relief") compared to the surrounding
hydroxyapatite of the bone. Arrows point to the infilled Haversian
canals.
In XN, the bone material is dark, with some concentric lighter zones arranged on either side of each canal in an "X" pattern. This is caused by the radial orientation of very tiny hydroxyapatite crystals. When parallel to the polarizers, they are dark (i.e. left and right, above and below the Haversian canal). When at an intermediate orientation, the crystals cause some light to get through, hence the light zones. This can be confirmed by rotating the specimen -- the "light" and "dark" zones keep the same orientation while the specimen rotates, because they are fixed by the optics of the microscope. Eventhough the individual hydroxyapatite crystals are too small to observe (SEM can resolve them), their presence and preferred orientation is indicated by this "extinction" pattern. Better examples of this effect are shown below for quartz.
The calcite shows up well in XN with bright "rainbows" of interference colours known as interference fringes. Although not resolvable in the digitized images of this specimen (see below for better), several "orders" of "rainbows" are visible on many of the crystals, indicating a large difference between the refractive indices at different orientations in calcite's crystal lattice. This difference is known as "birefringence", and calcite has one of the highest of the common rock forming minerals, making it easy to recognize. Note the blocky shape to the crystals, and slightly smaller grains along the inner surface of the Haversian canal. This is due to the initial random nucleation of many crystals on the open surface, and dominance of only some crystals as crystal growth continued.
XN -- compact (Haversian) bone, transverse section, 200x. At
this scale, the interference fringes of the calcite crystals are
clearly visible (lower arrow), although the calcite is fine enough
at the upper arrow that it looks like a blur of random colour.
PPL
XN -- trabeculae, transverse section, 20x. Trabecular pores
completly infilled by minerals. The bone material (arrowed in the XN image)
is similar to that seen in the compact bone, but it is laminated
around much larger cavities. An initial, irregular layer
of opaque (black) pyrite is present, followed by concentric
growths of chalcedonic quartz. There are only faint outlines
of these in PPL because of the low relief of quartz. Also visible
in PPL are minor amounts of high relief calcite (grey,
rough appearance, arrowed in the PPL image).
In XN, the radial nature of the quartz is much more
obvious. See below for more detail.
XN -- trabeculae infill, transverse section, 50.4. Spectacular
"cross extinction" is visible in the quartz crystals, where the
needle-like elongated crystals are parallel (dark) or not parallel
(light grey) one or the other microscope polarizer. Also note the concentric
light and dark variations, caused by very slight changes in the
orientation and size of the quartz crystals as they grew. Note
that these patterns match between each of the radial quartz growths.
This occurs because the variations in growth were likely caused
by chemical changes in the solutions precipitating the quartz,
which would simultaneously affect the growth of the crystals
around the entire cavity. Arrows point to the black, opaque
pyrite rim. Note the way the quartz growths preferentially nucleated
on projections in the pyrite rim.
XN -- trabeculae, transverse section, 128x. Closer view of
one of the radial growths of quartz, resolving individual needle-like
quartz crystals.
PPL
XN -- trabeculae transitional to Haversian canal, transverse section,
50.4x. This cavity is completely infilled with blocky calcite
crystals. In
PPL, the high relief of calcite makes the crystal boundaries obvious.
In XN, the crystals are brighter or darker depending upon their
orientation with respect to the polarizers. Note the way the
intial nucleation of calcite growth began with small crystals (arrowed), which
were replaced by fewer, larger crystals as infilling progressed.The colours of this image are a bit off. Most of the calcite crystals have a very light pink or light brown colour in XN, but the brightness of the image has partially saturated the camera and muted the colours.
XN -- detail of calcite infilling, 128x, showing the interference
fringes (arrowed) characteristic of calcite. At least five orders
of interference spectra are visible, versus the one order of
interference colours typical of quartz (see quartz examples
above and below).
PPL
XN -- trabeculae, transverse section, 20x. The open space
of the trabeculae has been infilled with an initial layer
of opaque (black in PPL) pyrite, calcite (grey and high
relief in PPL), and concentric growth of
the chalcedony variety of quartz (low relief, clear in PPL,
radial crystals of first-order interference colours in XN). In
XN, the formation of a concentric layer of chalcedonic quartz around the inner
surface of most of the pores is obvious (arrows). The lower arrowed
example has the remaining space infilled by a large calcite crystal,
whereas the upper arrowed cavity is still partially open.
PPL
XN -- trabeculae, transverse section, 50.4x.
The laminar bone structure
and lacunae are visible in the bone in PPL, as are the fine, radial growths
of quartz on the inner surface of the pore space. In XN, the
fine needle-like radial growths of the quartz are more obvious, and
the interference colours range from first-order grey, through white,
to a maximum of yellow (this section is a bit thick, so the colours
reach slightly higher level than normal). This type of radial, elongated,
concentric growth of quartz is known as the variety chalcedony. This
is the type of quartz that forms many types of agate.
The smaller cavities (arrowed) have been completely infilled by quartz, whereas the the larger ones have only developed a layer, with remaining pore spaces open or infilled with calcite.
XN -- trabeculae, transverse section, 128x. A closer view
of the smaller infilled pore, similar in size and structure to the Haversian
canals of compact bone.
XN -- trabecalae, transverse section, 128x. A closer view
of the incomplete infilling of radial crystals of chalcedonic quartz on
the interior of a larger pore.
XN -- compact (Haversian) bone, transverse sectin, 20x.
The Haversian canals of this specimen have been infilled with a layer
of opaque pyrite, followed by blocky calcite crystals.
Hubert, J.F.; Panish, P.T.; Chure, D.J.; and Prostak, K.S., 1996. Chemisty, microstructure, petrology, and diagenetic model of Jurassic dinosaur bones, Dinosaur National Monument, Utah. Journal of Sedimentary Research, v.66, no.3, p.531-547.