Join 3,374 readers in helping fund MetaFilter (Hide)


Help me find elegance in science and experimental design.
October 31, 2011 7:53 PM   Subscribe

What are the most elegant experiments in all of science?

Today I learned about the Meselson and Stahl experiment that proved semi-conservative replication occurs in DNA. By growing bacteria in a substrate rich with heavier nitrogen isotopes, the scientists ensured that the bacterial DNA (which utilizes nitrogen in nitrogenous bases) would be heavier than similar DNA synthesized in a non-isotopic substrate. After the bacteria were removed from the heavier substrate and allowed to produce another generation in normal substrate, Meselson and Stahl found that the subsequently synthesized DNA was a hybrid of heavy/light. In subsequent generations, only one of the two strands would demonstrate this hybrid light/heavy characteristic, thereby proving that DNA synthesis was semi-conservative and that the parent strand is preserved. My biology teacher described it as one of the most beautiful and elegantly designed experiments in all of biology.

I then asked my Orgo professor whether she knew of any particularly elegant or beautiful experiments or reactions in chemistry and she responded that she thought the Diels-Alder reaction represented a reaction that exemplified elegance in organic chemistry. The reaction itself is remarkably adaptable and requires little energy to produce a valuable substrate (a cyclohexene) Wikipedia describes the Diels-Alder reaction as the "Mona Lisa" of organic chemistry reactions.

Now I want to find more! What are some of the best examples of elegance and beauty in science and experimental design?
posted by ghostpony to Science & Nature (34 answers total) 53 users marked this as a favorite
 
These are all physics experiments, but certainly fit your criteria otherwise.
posted by Gilbert at 7:57 PM on October 31, 2011 [1 favorite]


It's mentioned in Gilberts link, but I always thought the double-slit experiment was beautifully simple. Very simple to demonstrate with a cheapo laser pointer.
posted by Static Vagabond at 8:05 PM on October 31, 2011


The pitch drop experiment is elegant in its way.
posted by jedicus at 8:06 PM on October 31, 2011 [2 favorites]


The Michelson-Morley experiment is another elegant setup, and it's a good illustration that sometimes negative results can be important.
posted by penguinicity at 8:16 PM on October 31, 2011 [4 favorites]


This may not be a useful strategy for finding them, but experiments named for the scientists that devised them tend to be elegant--the only one I can think of (your example reminded me) is the Hershey-Chase experiment, which was a step toward confirming that DNA was the chemical that carried genes.
posted by pullayup at 8:17 PM on October 31, 2011


The Hershey-Chase experiment was mentioned previously, but for some reason it didn't get the same sort of attention even though it identified something so crucial (also, it's the original "will it blend?", except this time with phage coatings!).

The trend so far seems to be pairs of scientists (or at least the pairs they could fit in an experiment name). Maybe being partnered up is the secret!
posted by ghostpony at 8:20 PM on October 31, 2011


I think the Millikan oil drop experiment qualifies.
posted by cosmac at 8:21 PM on October 31, 2011 [2 favorites]


Ok, Robert Millikan and Harvey Fletcher experiment!
posted by cosmac at 8:23 PM on October 31, 2011


It's not an experiment, but I have always found Maxwell's Equations breathtakingly elegant. Four simple equations (together with a fifth, the Lorentz force law) explain basically everything about classical electricity and magnetism.
posted by jonathanweber at 8:29 PM on October 31, 2011 [1 favorite]


Millikan Oil Drop experiment, replicated by young physics students everywhere.

The main thing I remember about doing this experiment is the poor dude who managed to shock the hell out of himself, knocking the entire experimental apparatus onto the floor in the course of his accident.
posted by mikeand1 at 8:46 PM on October 31, 2011 [1 favorite]


Speaking of oil, Benjamin Franklin's experiment (similar to one performed by Pliny) is quite elegant. Take a known quantity of oil and allow it to spread over a large area. You can then estimate the size (certainly establish an upper bound) of the oil molecule from a measurement of the area.

Another of my favorites is Eratosthenes' measurement of Earth's circumference. To be able to measure the whole world based on a measurement of two cities 800km / 500mi apart is the height of turning small details into life-changing conclusions.

Double vs. single slit is really elegant, too. Captures a lot of the spooky nature of quantum mechanics while being historically significant.

It's hard to point to a single definitive experiment, but another along the lines of the double slit experiment are the various "lights" (many not in the visible spectrum) that were found to pry electrons out of metal. This was the photoelectric effect that showed that there was some sort of relationship between photons and electrons, and led to a lot of other conclusions as well. If you could produce a similar elegant experiment showing a link between, say, photons and gravity, a Nobel would be obligatory.
posted by wnissen at 9:03 PM on October 31, 2011 [2 favorites]


In the 1970s the structure of cell membranes was still being worked out. The fluid mosaic model, which hypothesized that proteins float more or less freely in the lipid bilayer, was supported by some lovely experiments showing that fluorescently labeled proteins could diffuse in the plane of the membrane. The loveliest was probably that of Frye and Edidin (J Cell Sci, Sept 1970), in which they labeled one cell red and another green, then fused them together and watched the colors mix. This animation explains it pretty well. Old hat now, but breathtaking in its day.
posted by Quietgal at 9:16 PM on October 31, 2011


The Miller Urey experiment seems to fit.
posted by tomtheblackbear at 9:16 PM on October 31, 2011


Milgram's experiment on obedience has always struck me as having a certain ghastly elegance.
posted by flabdablet at 12:42 AM on November 1, 2011 [2 favorites]


I think Millikan is the apex of experimental elegance, but a couple other ideas:

(Here I define an "elegant experiment" as an experiment which is relatively simple but very far reaching)

1) Madame Wu's observation of parity violation in an experiment proposed be Lee and Yang. For decades no one imagined that parity could possibly be violated. In one simple experiment it was shown that parity violation is an inherent feature of nature (nature does not treat left and right equally).

2) Experimental tests of Bell's Theorem. I believe Bell's theorem is a criminally overlooked piece of our concept of the universe, and the experimental test is basically just an extension of the double slit experiment. These tests show that if causality holds, nature has a fundamental statistical aspect that can't be explained away with hidden variables. The universe plays dice.
posted by auto-correct at 1:01 AM on November 1, 2011 [1 favorite]


Milgram's experiment on obedience has always struck me as having a certain ghastly elegance.

Seconded. As shocking as his work was, it remains a masterpiece of empiricism. On a higher plane than the subject matter, his book Obedience to Authority documents the Art of designing a science experiment.
posted by Blazecock Pileon at 1:44 AM on November 1, 2011 [1 favorite]


Just about anything Galileo did. The man was an alien life form. Unique.
posted by FauxScot at 2:50 AM on November 1, 2011


Arthur Eddington's proof of Einstein's General Theory of Relativity. FTW. Dude moved the stars in the sky.
posted by sexyrobot at 3:38 AM on November 1, 2011


and...lol (srsly, just clicked over to this page, which i usually do right before bed), speaking of Galileo...video of one of his most famous experiments being conducted on the moon!
posted by sexyrobot at 3:45 AM on November 1, 2011


Books have been written on this 1, 2, 3
posted by lalochezia at 4:47 AM on November 1, 2011 [1 favorite]


The Geiger-Marsden experiment, more popularly known as the Rutherford gold foil experiment.
posted by DevilsAdvocate at 4:59 AM on November 1, 2011


Sending a man to the moon to watch a hammer and a feather drop at equal rates is either elegant or the opposite of elegant.
posted by fatllama at 6:22 AM on November 1, 2011 [1 favorite]


The minimal pair test for determining if a sound is phonemically contrastive in any given language.

You could argue that Chomsky's various intuitive/phenomenological proofs of the existence of deep structure, such as the phenomenon of grammatical ambiguity, amount to experimental procedures of a particularly elegant sort.
posted by spitbull at 7:02 AM on November 1, 2011 [1 favorite]


(Sorry, I should have said "if a PAIR of sounds ARE phonemically contrastive" in my prior answer.)
posted by spitbull at 7:03 AM on November 1, 2011


Oooh, also Berlin and Kay's experimental investigation of the so-called "Whorf hypothesis" using basic color terms, although in the end their elegance led to drastic over-simplification of the experiment's implications.
posted by spitbull at 7:06 AM on November 1, 2011 [1 favorite]


Reading The Emperor of All Maladies(a book about cancer), I just learned about the Ames test, which struck me as very elegant. Take bacterial cultures where the bacteria can only flourish if they've mutated in a specific way, and then introduce different chemicals to the cultures. 90% of the time, the substances that cause the bacteria to mutate are also carcinogens. Before this test was discovered, the only way to identify a carcinogen was to wait until a bunch of people got cancer and then try to figure out what they had in common.
posted by showbiz_liz at 7:27 AM on November 1, 2011 [1 favorite]


I second lalocheezia's amazon links and would like to add one more: Shamos's Great Experiments in Physics. It has the advantage of being (edited) reproductions of the actual descriptions of the experiments made by the experimenters themselves, which can help remind you just how crazy science was (in order to be wary to just how crazy science still is).

I'm a particular fan of the Faraday effect experiment described in it.
posted by dis_integration at 8:22 AM on November 1, 2011 [1 favorite]


If we define elegance as simple approaches giving amazing answers, Young's Experiment, aka the double-slit interferometer experiment, is one that combines remarkable simplicity with a truly dramatic effect on physics after it was first performed. It proved that photons acted as waves (thus, the particle-wave duality.)

If we defined elegant as getting truly difficult data out, there's two contenders, both modern -- LIGO and Gravity Probe B.

But really, my favorite is WMAP, if only for the simple results table and the effect those results had.

H0 71.0 +-2.5 kilometers/megaparsec
t0 13.75 +-0.13 gigayears
Ωb 4.56% +- .16%
Ωc 22.7 +- .14%
ΩΛ 72.8% +.15% -.16%

H0 is Hubble's Constant, the measure of the expansion of the universe. t0 is the age of the universe. The Omegas are what's in the universe -- Ωb is ordinary matter, Ωc is dark matter, and ΩΛ is dark energy.
posted by eriko at 8:27 AM on November 1, 2011 [1 favorite]


Read about what people were awarded Nobels for. Those are often the awesome, new way of thinking experiments that change fields.
posted by maryr at 8:45 AM on November 1, 2011


Many of them have already been mentioned, but this AskMe has some nice experiments.
posted by TedW at 10:35 AM on November 1, 2011


For me, it has to be Michelson-Morley, because it shows science at its best: they were trying to detect something that wasn't there (the ether) - that is, they were trying to provide experimental evidence for a hypothesis that turned out to be false. And the result of that "failed" experiment was a profound leap forward in our understanding or reality. Now that's what I call elegant.
posted by Decani at 12:19 PM on November 1, 2011 [2 favorites]


cosmac: "I think the Millikan oil drop experiment qualifies."

Please tell me this is some sort of sick joke. For the uninitiated, introductory High School and College students are subjected to this primitive experiment to measure the charge on an electron. It involves high voltages, and observing tiny droplets of oil through a microscope, and timing their motion over a certain distance (hopefully they'll be traveling in the right direction). It's known for being finicky and virtually impossible to do correctly, to the extent that some have accused Millikan of fraud. It also requires a huge number of (tediously acquired) data points in order to be accurate, given just how tedious and error-prone the data acquisition process is.

I've done it three times. I'd sooner volunteer to be waterboarded than do it again. Just writing this, I'm having traumatic flashbacks to the 12+ hours I spent in that stuffy lab, numerous electric shocks, and complete failure to get any meaningful results before the TA finally sent us all home.

On the other hand, Millikan's experiments that confirmed Einstein's theory on the particle nature of light (as an extension of Max Planck's work postulating the quantization of energy) were quite elegant, even though they were very difficult to construct and execute with the technologies of the day.

Ironically, despite being a great experimentalist, Millikan was a bit of a Luddite, and believed in the existence of "the ether" until very late in life, and doubted the theoretical underpinnings of the existence of the photon, which he himself is widely credited as experimentally confirming beyond a reasonable doubt.
posted by schmod at 12:37 PM on November 1, 2011 [1 favorite]


Oh, and Feynmann had a knack for these experiments. Here, he demonstrates a simple experiment that shows why the Space Shuttle Challenger exploded, in approximately 30 seconds.
posted by schmod at 12:41 PM on November 1, 2011 [1 favorite]


One more: Pasteur's experiments where sterile growth media (petri dishes with yummy nutrients) were placed in various vessels. Some were open at the top; these had mold/fungus/bacteria growth. Some were sealed, and had no growth. Some though, were exposed to air but only air that flowed uphill. There was no growth in these vessels, showing conclusively that there was not just "something in the ether" causing what we now know to be microbiological growth.
posted by wnissen at 12:56 PM on November 1, 2011 [1 favorite]


« Older Recommend some good online mul...   |  No one seems to be allergic to... Newer »
This thread is closed to new comments.