Science is messy. Sometimes it's literally messy, as anyone can see from the current state of my kitchen, where my 12-year-old daughter is conducting an experiment for her middle-school science fair.
And sometimes the mess stretches across the galaxy. In my last post I discussed an astrophysics discovery that appears to support the inflation hypothesis, which posits that the creation of our universe was kickstarted by a phase of extraordinarily rapid expansion. Using a microwave telescope at the South Pole called BICEP2, a team of astronomers detected a distinctive pattern of polarization in the cosmic radiation streaming across the universe. Because theorists had predicted that the process of inflation would produce this pattern, its observation was widely hailed as one of the most important scientific milestones of the 21st century. In the weeks since then, though, another team of researchers has cast some doubt on the discovery.
Led by Hao Liu of the University of Copenhagen, the researchers proposed an alternative explanation for the polarization observed by BICEP2. A galactic structure called a radio loop -- a wave of gas and dust spewed by a stellar explosion -- passes right through the patch of sky that was surveyed by the microwave telescope. Metallic dust grains in this loop might align with galactic magnetic fields, polarizing the microwave radiation they produce. Did this radiation skew BICEP2's results? If so, the observed polarization may not have been caused by cosmic inflation after all. We won't know for sure until later this year, when researchers announce the latest results from the Planck space telescope, which is also measuring the polarization of the cosmic radiation.
This development shows once again that science moves in fits and starts. Nothing is proven until it's proven. Good researchers are always skeptical, always questioning their own results and those of their colleagues. It can be a frustrating process. Just ask my daughter.
She started her science-fair experiment with a simple plan: Measure the buoyant force of liquids of various densities. She decided to do this by dropping a marble into a graduated cylinder full of liquid and timing how long it takes to reach the bottom. She used a foot-tall cylinder that holds about 350 milliliters of liquid. If dropped into an empty cylinder, the marble took about a quarter-second to hit bottom, which roughly matched the theoretical prediction: Distance (one foot) = ½ × gravitational acceleration (32 feet per second per second) × the square of the elapsed time (one quarter × one quarter = one sixteenth).
When the cylinder was full of water, though, the marble took about half a second to reach the bottom. My daughter was very excited to see this result. (To improve the accuracy of her measurements, she repeated the marble drop 20 times, then threw out the highest and lowest results and averaged the remaining eight. It's easy to make mistakes when you're measuring such a brief event with a stopwatch.) As she started to calculate the buoyant force of water on the marble, I looked over her shoulder and pointed out that she also had to take into account the force of friction against the marble passing through the water, which I believe is related to the liquid's viscosity. This upset her because she hadn't mentioned friction or viscosity in the experiment proposal she'd submitted to her science teacher. I told her, "That's OK. You can revise the proposal." She replied that this was her experiment, not mine, and I should mind my own business.
(In retrospect, it seems clear that she should've used a spring-loaded scale to measure the weight of the marble as it's suspended in the liquid. That would've revealed the liquid's buoyant force without the complication of friction. But I didn't mention this to my daughter. It would've upset her even more.)
She proceeded to the next step of her experiment, which was to create liquids of various densities by mixing liquid soap into the water. But when she weighed her solutions, she made a disconcerting discovery: Even the highest concentrations of soap didn't change the density of the liquid. Soap, as it turns out, is about as dense as water, which explains why those bars of Ivory floated in the bathtub in those long-ago TV commercials.
Now she was really flummoxed. Her desperation was so great that she actually came to my desk, where I was staring at the computer, minding my own business. At her request I looked up the densities of various liquids, trying to find a substitute for soapy water. Unfortunately, most household liquids have densities that are pretty close to water's. Milk, for example, is only a few percentage points denser. Olive oil is about 10-percent less dense, but that wouldn't be enough to significantly change the marble's drop time. Mercury is 13 times denser than water, but the fact that it's phenomenally toxic made me reluctant to start breaking thermometers. Finally, we found a promising candidate: corn syrup, which is about 40-percent denser than water. So we went to the supermarket and bought three pints.
When my daughter put the marble into the cylinder full of corn syrup, she expected the drop time to increase by a few seconds. Instead, the marble sank oh-so-slowly through the thick, brown liquid, taking 23 seconds to reach the bottom. But the bigger surprise came when she stirred some water into the corn syrup, creating a mixture that was four parts syrup and one part water. When she dropped the marble into this mixture, it took only two seconds to reach the bottom. Her simple science experiment had suddenly become complex: Something nonlinear was going on! Was there a threshold concentration at which the viscosity of the corn syrup substantially increased? The obvious next step was try to try a thicker mixture -- maybe nine parts syrup and one part water.
The experiment is continuing. (It's due next week.) In the meantime, I'm spending a lot of time wiping the sticky clots of corn syrup from our kitchen countertops. It's a messy job, but I'm willing to do it for science.