Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watnot!watmath!clyde!mcdchg!wucs1!wucs2!slustl!slu70!guy
From: guy@slu70.UUCP
Newsgroups: sci.bio,sci.astro,sci.misc
Subject: Re: Mass extinctions
Message-ID: <12@slu70.UUCP>
Date: Thu, 2-Apr-87 10:09:09 EST
Article-I.D.: slu70.12
Posted: Thu Apr  2 10:09:09 1987
Date-Received: Sun, 5-Apr-87 07:23:45 EST
References: <784@scicom.AlphaCDC.COM> <11235@teknowledge-vaxc.ARPA> <3497@cbosgd.ATT.COM>
Organization: St. Louis Univ., St. Louis, MO
Lines: 43
Xref: utgpu sci.bio:225 sci.astro:803 sci.misc:240
Summary: volcanic magmas

In article <3497@cbosgd.ATT.COM>, gwe@cbosgd.ATT.COM (George Erhart) writes:
> Just how certain are we of the composition of the earth's magma ? Given the
> conditions existing beneath the crust, it seems reasonable that significant
> elemental segregation would occur; i.e., there would be local variations in 
> the composition. I believe such variations are credited for the creation of 
The variations are not that large. the segregation of gold is due to processes
occurring after the creation of magmas (mainly hydrothermal) not variations
in magma source. We can measure compositions of lots of volcanic rocks, both
recent and older, so statements about the composition and homogeneity of
the mantle are tolerably well supported.
> gold veins, etc. Further, it seems not unlikely that osmium might preferentially
> segregate to the melt, rather than the lithosphere, and that the ratio of 
> isotopes might be different, as well. 
This is probably the case (none of my references discuss osmium specifically
but I suspect that it is a large ion lithophile). Most of the ratios used
involve ratios of isotopes which are chemically identical and segregate
in the same fashion. They also reflect the ratio (not the absolute abundance)
in the source region.
> 
> Perhaps the (comparatively) small amount of volcanic eruption in "recent" 
> history is sufficiently low as to draw magma only from the region nearest
Magmas which have to pass through continental crust often have a crustal
component. Isotope ratios often allow this to be determined, at least
qualititatively. Certain kinds of rocks (e.g., basalts) can only have a
mantle source. There is simply no way to produce them from crustal rocks
without completely melting the rock which is highly unlikely. There is
also seismic evidence (e.g., for hawaii) which places the source well below
the base of the crust.
> the crust; this magma would more closely approximate the crust's composition.
> Under periods of greater volcanic activity, magma would be tapped from deeper
> within the earth, and might have heavy metal contents and isotopic ratios 
> consistent with the Alverez layer. 
There is no evidence for this happening. There is evidence for different
geochemical reservoirs in the mantle but they involve different kinds of
volcanoes (e.g., oceanic island vs mid-ocean ridge) rather than changes in
volcanic activity with time.
This should be discussed in any decent book on igneous petrology as well
as a number of papers (some quite accessible to non-geochemists) by
De Paolo and Wasserburg. I can provide exact references if you like.
> I dunno; this is just off the top of my head. Critiques are welcome.
Check out the article by Officer et al. in the Mar. 12 issue of Nature
for the most recent exchange in this ongoing battle. I haven't read it
yet so I don't know if I believe them or not.