I'll try to post it later. Meanwhile -- so far as I can remember -- it did exist, it just wasn't nearly as extensive as in the Northern Hemisphere. I'm not positive why this was, but I can speculate.

Normally (whatever that means), Arctic ice is directly connected to more-or-less flat-lying northern landmasses in Canada and Russia by a winter pack. Therefore, when that ice became very much thicker and started to flow, it was able to move for thousands of miles. This continental flow encountered and merged with the outward/southward flow of ice from the Greenland continental ice sheet (which still exists), and with alpine-type glaciers flowing out from the Rockies, Alps, Urals etc.

In contrast, the nearest significant landmass to Antarctica is the long, narrow southern tip of South America -- highly mountainous. So there was alpine glaciation in high latitudes in South America, but not (so far as I know) much in the way of a continental ice sheet. I imagine that at least the South Island of New Zealand -- the mountains of which have ice fields even today -- experienced some glaciation, and there were even alpine-type glaciers in southern Australia which grew significantly. Undoubtedly the Kilimanjaro glacier did as well.

But in general I'd say that glaciation in the Southern Hemisphere was an alpine, not a continental phenomenon, for geographic/topographic rather than strictly climatic reasons.

Edit:

My mistake. The Laurentide ice sheet (aka the "big-ass" glacier that covered most of present-day Canada and the northern US) did not originate with the Arctic pack, but on land in what's now Nunavut. Scroll down to the last figure on this page.

Edit the Second:

Here's an abstract of an article about Pleistocene (i.e., "recent") glaciation on the Australian mainland.

[ 24 February 2005: Message edited by: 'lance ]

quote:

---

With all that ice in one hemisphere, did it introduce a wobble? Or shift the axis of rotation?

---

Warning: possibly long-winded post follows.

I don't know, but I don't think so, for two reasons. In the first place, to introduce a wobble (rather, an additional wobble: see below) would have taken much more ice in the Western Hemisphere, say, than in the Eastern (or vice versa). The maximum southern advance, in other words, would have had to be very much greater on one side of the earth.

Instead, at the last maximum of the Wisconsin glaciation, around 18,000 years ago, ice got to around 40 degrees north latitude in North America, and around 45 degrees in Eurasia (though as this page points out, while alpine and continental sheets merged in North America, European and Asian sheets did not merge). That's probably close enough for rough symmetry.

Second, even had there been great asymmetry, I still think the mass of ice would have been so very much less than the mass of the Earth as not to matter. (I could probably do a back-of-the-envelope calculation using the figures on this page).

But about wobbles: the Earth's axis already wobbles in two senses. The equator shifts from a minimum of 21.5 degrees above the plane of the Earth's movement around the Sun (or the "ecliptic), to a maximum of 24.5 degrees, then back again, over a period of about 41,000 years. (Or, looking at it the other way, the axis shifts from a maximum of 68.5 degrees above the ecliptic to a minimum of 65.5 degrees).

And the axis also wobbles in the sense that it describes a circle, shifting from pointing at Polaris (the North Star) to pointing at Vega, and back again, over a period of about 23,000 years.

According to an astronomer named Milankovitch, who described these cycles around 1910 or so, the interaction of these two motions with the eccentricity of the earth's orbit (its shift from being more to being less elliptical and back again; perfectly circular orbits are unknown in astronomy) is responsible for fluctuations in the amount of solar radiation the earth receives, and therefore for the ice ages.

It's explained and illustrated on this page (where I got the figures, because I can never remember them).

I vaguely remember from undergrad courses a few years ago that some are beginning to doubt or at least refine the Milankovitch theory, but I think it'll likely remain an important part of the explanation for ice ages.

Edit:

I'm curious, so I did the calculation.

According to the page I linked to above, about Florida, at the last glacial maximum the global volume of ice was about 97.08 x 10⁶ km³.

The density of ice is about 0.931 g/cm³, or 9.31 x 10¹¹ kg/km³.

So the mass of ice would have been about (97.08 x 10⁶ km³)(9.31 x 10¹¹kg/km³), or 9.04 x 10¹⁷ kg.

If the mass of the earth is about 5.98 x 10²⁴ kg, oceans included, the ice would have outweighed it by a factor of 5.5 x 10⁴, or 55000, give or take.

But stop the presses! What am I saying? All that ice didn't drop onto the earth from outer space. It's merely frozen water, which is part of the water cycle; when the continents and mountains were covered with sheets of ice, sea levels were correspondingly lower. The total mass of the earth wasn't different; it was just distributed differently. But I think these figures show that even if it were new mass, it couldn't have affected the earth's rotation much, if at all.

[ 24 February 2005: Message edited by: 'lance ]

As every reader of Terry Pratchett knows, yes there is a Great Turtle, a'Tuin (gesundheit!), but the earth itself is supported by four elephants standing on a'Tuin's back.

Notwithstanding all this "turtles all the way down" nonsense, a'Tuin isn't standing on anything, but swims endlessly through space, looking for a cosmic beach on which to lay its eggs.