How a NanoDrop works

We often use spectroscopy (such as with a NanoDrop spectrophotometer, which is really quick and only requires tiny volumes) to measure concentrations of molecules like nucleic acids (DNA or RNA) or proteins - by shining light through solutions containing them and seeing how much of the light gets stolen along the way, we can calculate how stuffed full of those molecules the solution was.  This is possible because different molecules have different tendencies to absorb different wavelengths of light to different extents (more in a sec). If you shine a molecule’s “favorite” wavelength at it, the more molecules there are, the more that will get stolen. You could just measure the absorbance of this favorite wavelength and, if you have a pure sample and you know how much your molecule likes that wavelength (the extinction coefficient) you can calculate the concentration (using a law called Beer’s law) just with using that one wavelength of light.  

But is your sample really pure? If other molecules like that wavelength too they can be artificially inflating your concentration - not to mention other problems with having contaminating molecules around that can interfere with things.  
 
So (if possible) you shouldn’t just look at a single wavelength absorption value - instead, you can learn a lot more if you measure over the whole ultraviolet (UV) and visible light portions of the electromagnetic radiation (EMR) spectrum. Each wavelength that gets absorbed will show up as a peak an absorbance spectrograph, which shows how much light of a range of wavelengths was prevented from making it all the way through the sample to the detector on the other side.  
 
for a longer version: http://bit.ly/dnauvbeer

The height of the peak you usually focus on for a particular molecule is where it absorbs best. This corresponds to how much stuff is there (concentration) (taking into account that molecule’s willingness to absorb that wavelength (its extinction coefficient). BUT a lot of information about purity can be gained by looking at where they shouldn’t be absorbing much light - and ratios between peaks can tell you about how pure that stuff is 
 
You can see this if you look at a UV-Vis absorbance spectrum, which you can can get if you use a spectrophotometer like a NanoDrop. To get a full absorbance spectrum, you shine light of all wavelengths (well, at least all visible & some ultraviolet (UV), which is where biological molecules tend to absorb) & measure what goes through (is transmitted). Whatever doesn’t go through is assumed to be absorbed (or abducted by aliens…). So absorbance = 1-transmission. 
 
So how does it work? We take whatever solution we want to measure and put it in a plastic or glass holder called a cuvette which has a window for light to shine through & stick it in a spectrophotometer which actually shines the light through one side & measures what comes out the other. Different cuvettes can have different path distances the light has to travel, and the Beer Lambert law takes this into account - the longer the path, the more molecules light’s likely to hit (regardless of the concentration) - and the more molecules it hits, the more chances there are to be absorbed. Our calculation of concentration is based on how much light gets absorbed so we need to account for this distance. 
 
It’s also important to have a blank - this is just the liquid you have your samples in and all the “constant” components (e.g. salts, etc. that are in each reaction) - its made up of molecules too so it’ll have a characteristic absorbance that will always be there, whether or not the reaction we’re looking for actually occurs. And some of its absorbance spectrum might overlap with our products’. So we want to subtract it out so we don’t confuse it for our signal. 
 
Cuvettes are great for things where you have “large” samples. But if you don’t have much sample though, you’ll want something smaller-scale. 
 
The NanoDrop spectrophotometer has a little pedestal you put a drop of liquid on (a really tiny drop, like 1-2μL “μL” stands for microliter and it’s a millionth-of a liter, or a thousandth of a milliliter. Then you lower the arm → it contacts the liquid then pulls up a little bit and, when it does, it pulls on the liquid. It does this thanks to surface tension. Surface tension occurs because the molecules of the liquid like each other more than they like the air - so they try to stay together & maximize the liquid-liquid interactions while minimizing their combined air exposure. more here: http://bit.ly/surfacetensionbubbles  
 
When you put the drop on, surface tension causes it to remain drop-like. But when you lower the arm & squish it down, some of the water molecules stick to the top surface. And when the arm pulls back up, these molecules get lifted - and the other water molecules don’t want to leave their friends behind → as a result a column of liquid forms 

finished in comments