The second method of producing light is through line emission. To understand line emission, we must consider the structure of an atom. In the Bohr model of the atom, positively charged protons (and neutral neutrons ) live in the nucleus of an atom, surrounded by orbiting negatively charged electrons. Thus, our model atom is much like the solar system, except for one critical difference. In this model, electrons can only go in very special orbits (called orbitals) at discrete distances from the nucleus.

Now electrons always prefer to be as close to the protons as possible. (They are negative, the protons are positive, so they attract.) So they usually reside in the ground state, i.e., the orbital that is closest to the center. However, you can kick an electron up into a higher energy state (called an excited state) by giving it some energy. If this happens, then a very short time later (say 0.000001 seconds later), the electron will fall back down to the ground state, and give back the energy it absorbed in going up to the excited state. It will therefore emit a photon.

The key here is that light emitted by the electron in falling down to the ground state can only have a very particular wavelength, since the distance (energy) separation between the ground state and excited state is discrete. For example, consider a tube of hydrogen gas, with an electron current going through it. The electrons in the electric current will collide with the electrons going around the hydrogen atoms. The electrons will become excited and go up to a higher energy state. They will then fall down and emit light at wavelengths corresponding to the energy jumps they took in falling back down. If multiple levels are involved, the electron can have a choice: it can jump from top to bottom all at once, or it can jump into and out of intermediate levels. Either way, the total energy the electron gives off always matches the energy it took to excite the electron in the first place. Someone observing the light will not see a continuous spectrum; instead they'll see a series of emission lines, corresponding to the differences between orbital energies. This is an emission spectrum.

In the above example, collisions with free electrons caused the electrons in the hydrogen atom to become excited. However, another way for electrons to become excited is to absorb light. Suppose you shine light of all wavelengths at a collection of hydrogen atoms in the ground state. Those photons whose energies (wavelengths) are exactly equal to the distance between orbits will be absorbed. The rest of the photons will be ignored. A person staring at the light will then see the continuous spectrum minus the absorption lines of hydrogen. This is called an absorption spectrum.

A variant of the above processes can occur if you kick the atomic electrons extremely hard. Normally, electrons can only absorb discrete amounts of energy: either the photon has exactly the energy needed to cause a jump to a higher energy level, or it doesn't. However, if you give an electron enough energy, via either a violent collision, or an extremely high energy photon, the electron can absorbe the energy and leave the atom completely. The atom will become ionized. In other words, if the energy of the photon is great enough to cause an ionization, then no rules apply: the electron can absorb it if wishes. Of course, eventually, another electron will meet up with the atom that is missing its electron and recombine. This recombination will most likely be into an excited level; the electron will then cascade down. If you observing this system, the extremely high energy photon that started the process would be gone, but an emission line spectrum from the falling electrons would be detected.

Each type of atom (i.e., each element) has its own, unique separations between electron orbitals. Thus, the emission spectrum of each element is as unique as a fingerprint. By observing the pattern of emission (or absorption) lines, one can tell what element is doing the emission (or absorption). (For a demonstration, click here.)

Note that emission and absorption lines are really two sides of the same coin. Sometimes, what you see depends on where you are looking from. If you are looking at a continuum source, and cool gas in the way, you will see an absorption spectrum. On the other hand, if you are just looking at the gas, then you may see an emission spectrum. Or, if the absorption is being produced by extremely high energy photons (i.e., through ionization), then again, you'll see emission. The figure below demonstrates the differences.

The first way to do this is with lenses. When light passes through anything, it gets bent. So, by making a thick lens of the proper shape, you can make the light bend, or refract, so that it comes to a focus. You have a refracting telescope. Many small telescopes are refracting telescopes; binoculars are refracting telescopes.

Refracting telescopes have a few limitations. First, since light only gets bent a little while passing through glass, the focal length, (i.e., the distance from the lens to the focus) is usually very long. Long focal lengths give you a large amount of magnification, but that's usually not what an astronomer wants. When light passes through the atmophere on its way to the telescope, it gets refracted. Because the atmophere is turbulent, the refraction is different every second, These constant changes in refraction, called seeing blur our view of celestial objects. (It's like looking through the ``heat waves'' over a barbeque.) Refracting telescopes give highly magnified images, but they're usually just a big fuzz-ball. In general, astronomers prefer to have a larger field-of-view.

The second problem with refracting telescopes is that the refraction for each wavelength is different. Blue light gets bent more than red light, so the focus for the blue light is different from the focus of the red light. This is called chromatic aberration. Depending on where you put your instrument (or your eye), you'll see different colors.

Finally, refractive lenses can only be made so big. It's pretty difficult to put a multi-ton piece of glass at the front of a telescope, support it by its edges with what essentially is an eye-glass frame, and expect it to be rock solid while the telescope moves to all areas of the sky. The biggest refracting telescope in the world is has a 40-inch lens. Since telescopes are characterized by how much light they collect, which, in this case, is defined by the size of the lens, this telescope is called a 40-inch telescope.

The other type of telescope is called a reflecting telescope. For these instruments, you use a big mirror to bring the light to a focus. This design has many advantages. Since light isn't passing through anything, there's no chromatic aberration, and you can shape the mirror so that the focal length is short. Best of all, you don't have to support the mirror by its edges: you have the entire backsize of the mirror to play with. As a result, reflecting telescopes can be very large. The largest optical reflecting telescope has a mirror that is 10-meters across, and in Puerto Rico, the Arecibo radio telescope has a reflecting dish that is 300-meters in diameter, the length of 3 football fields!

Of course, reflecting telescopes do have one disadvantage. After the light hits the mirror, it comes to a focus high above and in front of the mirror. If you want to look through it, you might block out the light! There are several ways astronomers deal with this. First, for very big telescopes, they can just exist with it. At the prime focus of a big, reflecting telescopes there may be a small cage. Astronomers (or their instruments) can sit inside the cage. Some of the light coming down from the sky is blocked by the cage, but that's the price you pay.

A second way of solving the problem was first thought of by Isaac Newton. Near the focus of the primary mirror, you put a small, flat mirror, and angle it so that the light reflects off this mirror and comes to a focus at the side of the telescope. This is Newtonian Focus design. Since the pick-off mirror is small, not too much of the incoming light is blocked. Of course, the focus of the telescope is still high above the primary mirror , and there's an extra reflection. Since even the best mirrors are only 75% efficient, the design throws away 25% of the light.

Another way of designing a reflecting telescope is to drill a small hole in the middle of the primary mirror, and put a small, curved mirror up near the prime focus. Light comes down, hits the primary mirror, reflects up, hits the secondary mirror , then comes back down, passes through the hole in the mirror and comes to a focus. This Cassegrain design has the advantage of having the focus behind the primary, near the ground. The disadvantage is the extra reflection, and the longer focal length, which gives you more magnification, but a smaller field-of-view.

Occasionally, if an astronomer wants to study just one object and doesn't mind (or he actually wants) a large amount of magnification, then a Coude Focus is in order. As in the Cassegrain design, light comes down, hits the primary, bounces up, hits the secondary, then comes back down again. However, in this case, rather than coming out a hole in the primary mirror, the light is reflected again out the side of the telescope. It then may be reflected into a different room, into a different floor of the building, or even into a different building.

There are few large optical telescopes in the eastern part of the United States. Major optical observatories are located where there is a) clear weather most of the time, b) high mountains, to minimize absorption by the atmosphere, and c) dark skies, and d) stable air, to minimize seeing. Radio telescopes don't have these limitations, since radio waves are unaffected by clouds and atmospheric turbulence. They are, however, affected by radio noise, i.e., radio and TV broadcasts, cellular phone transmissions, etc. One the major radio observatories in the United State is located in a radio-free zone in Greenbank, West Virginia.