Into the Bushveld #1: holy hunks of magnetite!

As a palaeomagician, there’s an intrinsic interest in having a huge hunk of magnetite (well done to those who eventually worked it out) on one’s desk, even if I’m going to have to be careful not to put it near any of my samples (or my credit cards, for that matter). But how has such a thing come to form? Normally magnetite is a fairly minor component of igneous rocks, and of sediments that are derived from them; what process has produced such a concentrated mass of it? Perhaps taking a closer look of the original sampling locality will gives us a few clues:

Main Magnetite Layer, Bushveld


Clue 1: The light rock above and below the magnetite layer is anorthosite – a coarse grained igneous rock composed almost entirely of plagioclase feldspar. Coarse-grained usually means slow cooling, so we’re probably in some sort of intrusive rock.
Clue 2: The upper and lower boundaries of the magnetite layer look very different: the lower boundary is sharp, whilst the upper boundary is quite fuzzy. In fact, there is a very clear mineralogical grading occurring within this horizon: as you move upwards through the layer, you start to see more and more plagioclase crystals within the magnetite matrix, until it takes over as the principal consituent and we are back into anorthosite.

Magnetite grading into anorthosite, Bushveld

The simplest explanation for this is that this whole sequence formed from a magma which took a much longer time than usual to cool, allowing mineral phases which would normally be jumbled together as the rock solidified around them to be separated out by physical processes like gravitational settling.
Clue 3: The lower boundary, whilst compositionally sharp, is not completely flat. In several places you can see undulations where the dense magnetite appears to have pushed down into the anorthosite below, strongly implying that the anorthosite was not completely solid:

flame-esque structures at base of magnetite layer, Bushveld

Clue 4: How’s this for a xenolith?

anorthosite xenolite in magnetite, Bushveld

This again shows that the formation of the magnetite layer appears to have deformed the material below it.
What’s fascinating about all of this is that while we are definitely dealing with igneous rocks solidified from a melt, many of the features of this outcrop – grading, deformation of apparently unlithified rocks – are more reminiscent of sedimentary environments. This sort of thing can only happened when a large pool of melt is injected into the crust in one go. Such a large intrusive body takes a long time to completely cool and solidify, which allows melt to persist for long periods within the system. This means that newly formed crystals are in suspension in the magma, just as sediment particles are suspended in river, lake, and ocean water; hence they can be carried around, and sorted into layers, and deformed before they are completely rigid. How big does the intrusion have to be for such processes to have a significant effect? You’ll have to wait until tomorrow’s post to find out, but I’ll tell you right now that the answer in this case is ‘pretty big’.

Categories: fieldwork, geology, volcanoes

Comments (11)

  1. Kim says:

    It also needs to have a pretty interesting starting chemistry, to end up with magnetite and plagioclase (instead of olivine, chromite, pyroxene, and plag) as its major layers, doesn’t it? (—> not an igneous petrologist, just someone who has taught petrology here…)
    The most magnetite that I’ve ever seen in one place was in a magnetite-ilmenite mine in the Adirondack Mountains of New York. There is also anorthosite (a gorgeous rock, with lots of labradorite plag) there. But it’s been deformed and metamorphosed and has a lot of glacial till and trees on top of it, so its origin wasn’t obvious. (Well, at least to me, and I was there on a field trip with other people who taught mineralogy, and we were discussing small-scale mineralogical things rather than big-picture why-is-this-here kinds of questions.) I wonder if the Adirondack body had a similar origin to this part of the Bushveld?
    Very cool.

  2. Mathias says:

    Hmm…interesting! Very! Is there any similarity with the Tellness Magnetite-Ilmenit deposit in Norway? It has a pretty extensive Anorthosite body, too. Though I do not recall the magnetite to be so massiv there.

  3. Ron Schott says:

    You’ve done a fine job demonstrating that the mineral in question is a cumulate in a large LMI, Chris, but I was already convinced of that. On what basis do you identify it as magnetite rather than chromite? Can you offer a photo of a simple hand magnet sticking to it? Both chromite and magnetite will deflect a compass needle, but only magnetite will pick up a magnet off a wooden table.
    I’ve got samples of chromite from the Stillwater Intrusion in Montana, which are much finer grained than your Bushveld samples, but show similar field relations. I’ve also got quite a number of samples of the ilmenite-magnetite segregation from Tahawus in the Adirondacks – definitely not cumulate there, but certainly consanguineous with the anorthosite. And I have a third sample of nearly pure magnetite ore from the mine dump at Champion in the UP of Michigan (as well as more specular hematite than you can shake a stick at). The magnetite from latter two localities will pick up a magnet, but the chromite from Stillwater won’t.
    So I guess I’m asking for either visual evidence of the magnet test or a cation recount.

  4. Mathias says:

    I was wondering if it is pure magnetite or also includes some ilmenite.
    The chromite samples we have in our university collection are all non-magnetic. I had my diploma exam in december and spent 1 month in the collection practising and tried every sample. All chromite samples, around 5 different ones, are non-magnetic. if they are its so weak u cannot see it with a magnet.

  5. Lab Lemming says:

    Ron, see his previous post for an example of what it does to his compass.
    And like Kim, I’d like to know why the magma was so oxidized, and if the plag shows a reduced Eu anomaly as a result.

  6. Ron Schott says:

    I stand corrected. I thought chromite would deflect a compass but hadn’t tested that when I made my assertion above. Now that I have tested it I see that I was wrong. So I guess I’m buying in to the magnetite ID now.
    Can you point to a field guide or journal article that discusses this locality, Chris? I suppose it’d have to be higher up in the Bushveld stratigraphy after the iron enrichment of the magma peaked.

  7. Andrew says:

    Can I just say that the second photo is a thing of beauty?

  8. kc says:

    This post should be converted into part of an undergrad powerpoint lecture! *directs igneous prof to HA*

  9. Chris Rowan says:

    Andrew – thanks. It took many attempts to get a good one – South African sunshine makes it difficult sometimes! I can e-mail you a high-res version if you want.
    Ron – Will this do?
    Kim – see the next post. Although I should stress that I’m not an expert in this stuff by any means…

  10. Ron Schott says:

    Quite nicely! Thanks.

  11. josul hvan says:

    the abrubt increase in oxygen fugacity needed to precipitate the magnetite horizon says something happened…
    multiple horizons says it was a cyclic event…