Dear Louise,

I concede that it's not always possible to ensure complete homogenization
of biogenic carbonate samples. The brachiopod and bryozoan materials I
used as a basis for my tests came from environments where measurements
of salinity and isotopic composition of water varied by very small amounts
and the annual temperature range was between 4-6oC. Although this temperature
range is equivalent to between 1-1.5 per mil, it is reasonable to assume
that a satisfactory degree of homogenization was achieved by hand crushing of
bulk samples (about 1g in total). The sample weights used for isotopic
analysis were 2-3mg and each test was replicated up to 10 times so I think the
chances of the differences occuring between tests being due to an inhomogeneous
starting material, when using sub-samples drawn from the same pot of powder,
are quite small. You can draw your own conclusions on whether you think this
is OK.

I should point out that a paper in J Sed Pet a few years back identified
hydrogen peroxide as an agent capable of causing significant calcite
dissolution. I suggest you move to sodium hypochlorite (see Scott Carpenter's
message for more details). Secondly I would suggest that it's inadvisable
to heat any carbonate to red heat despite your suggestion that it doesn't
appear to affect your samples. If this works for you I think you've been
saved by two factors: (1) you're dealing with ancient carbonate which
probably contains little organic matter compared to modern biogenic carbonate;
(2) you're nummulites represent a very large particle size relative to the
component particles in a powdered sample- as has been discussed already (see
earlier messages) particle size and increased surface area may play an
important role in controlling the degree of isotopic exchange which takes
place during any heat treatment of carbonate materials.

Cheers,

Steve Crowley