Maria Brumm has a Master's degree... in Science! She wrote her thesis on 
I was struck by the similarity of these two images. Which one does your dentist want you to eat?
I won't speculate about anyone's dentist's motivations, but the top image comes from The Science of Ice Cream, and the bottom from the USGS. I'm not sure what conditions the ice cream was stored under, but the USGS image is from a limestone that had been stored next to a piping hot intrusion of monzonite, and partially baked into marble.
Frozen desserts left in the freezer too long will undergo a similar process. The edges of each ice crystal are constantly exchanging water with the sugar solution. Small crystals, which have a lot of surface area, will tend to shrink especially fast while the more energetically favorable crystals (in the ice cream's case, it's just big ice crystals that are better than small ones, but in rocks it is usually a different mineral phase) will tend to grow. This process is accelerated at higher temperatures, which is why it's a bad idea to leave a carton of ice cream sitting out on the counter even if it doesn't obviously melt.
Here, the top image is of fresh ice cream, and the bottom is of the same ice cream after it's sat in the freezer for 7 weeks:
Donhowe et al., 1991. Determination of ice crystal size distribution in frozen desserts. J. Dairy Sci. 74 (10).
It's not freezer burn and off flavors from sitting next to fish fillets that make old ice cream taste bad - or it's not just freezer burn and fish, anyway. Our perception of taste is closely linked to texture and mouthfeel, and those giant ice crystals are just yucky.
Physical Science
Comments
This is brilliant. You must turn this into a lab for Geo 1001
Posted by: greg laden | August 21, 2008 10:12 AM
Well, there's another book going on my "must-read" list! (Why is it that all the books I want to write have already been written? Imagine the research for "The Science of Ice Cream!")
Now that you explain it, it does kind of make sense that ice cream and parbaked marble might have similar structures. But it's definitely not a comparison I would ever have thought to make.
Posted by: Dr. Kate | August 21, 2008 12:13 PM
So, I'm guessing that in the case of the image of the partially formed marble the granules are calcium carbonate that has yet to undergo diagenesis and the lattice streaked area is the newly formed marble?
Posted by: Bill | August 21, 2008 4:25 PM
To Bill:
The granules are teeny-tiny garnet crystals. (Green garnet, according to the caption on the USGS link that Maria gave. Not pure calcium carbonate in the starting material.)
It would be a bummer to bite into ice cream and hit a bite of garnet. Ouch.
Also, the possibility that my ice cream might metamorphose is a good excuse to eat it.
Posted by: Kim | August 21, 2008 4:52 PM
It's apparent that if we took a vote on igneous v. metamorphic ice cream, that igneous would win hands down - whereas some of my favorite rocks are metamorphic!
Posted by: Silver Fox | August 21, 2008 6:55 PM
Thanks, Kim. I've afraid I have a dial-up connection that is at the moment, even for dial-up, dead slow, so I thought I'd just ask.
Posted by: Bill | August 21, 2008 7:29 PM
So are we doing ice cream stratigraphy tomorrow?
Posted by: Lab Lemming | August 21, 2008 8:09 PM
I suspect that sedimentary deposit of ice cream might involve a lot of hysterical crying from the five-year-old.
Posted by: Kim | August 21, 2008 9:13 PM
You'll be pleased to know I just pooped my pants.
Thanks.
Posted by: Ibod Catooga | August 28, 2008 1:43 AM