Ethan and other interested parties,

Thanks for the comments. This whole business seems to me to be a "can of worms".
A number of things strike me, but I don't want to hog this discussion so
I'll limit this reply to one specific point regarding particle size. In a
paper published in 1966 Emiliani (J. Geology, 74, 109-126) reported some
roasting experiments. He suggested that the effects of roasting on samples
of Recent limestone (0-425,000 yrs BP) were particle size dependent - with
the finer fractions giving the most negative isotopic values - and decided to
use only a particular size fraction (I can't remember the exact size fraction,
but I think it was about 100 microns). This tends to suggest that roasting
is a problem and that particle size is a contributing factor. Interestingly,
I don't think Epstein et al raised this in there original papers and,
of course, it's always intrigued me to know why the Epstein et al
palaeotemperature equation agrees so well with the Friedman & O'Neil version
despite that fact that they used different sources of biogenic carbonate to
derive their equation and roasted the hell out of their samples in the
process.

I don't think we've heard anything from the "roasters" yet. Perhaps they'd
like to chip in with a few comments. Specifically, I'd like to here whether
they feel that roasting at lower temperatures (i.e. 200oC instead of 380oC)
would be beneficial.

Cheers,

Steve Crowley