# Why do we see spectral LINES, not circles/triangles etc?

1. Sep 22, 2014

### girlinphysics

We were looking at the spectral lines of sodium in class and I was wondering, why do we see lines, and not any other shape like circles?

2. Sep 22, 2014

### vanhees71

We can't answer your question very well, if you don't give us some details. Without the exeprimental setup, we can not figure out about the shape of the spectra you observe.

3. Sep 22, 2014

### girlinphysics

We had a sodium lamp set up in the lab with a spectrometer and were told to measure the diffraction angle of the doublet components in the sodium spectrum. We saw lines like this: http://hyperphysics.phy-astr.gsu.edu/hbase/quantum/modpic/picklespectrum.jpg
I was just curious as to why spectra occur as lines, not other shapes.

4. Sep 22, 2014

### bhillyard

Was the light beam passed through a straight line slit? It usually is so the straight lines are effectively images of the slit.

5. Sep 22, 2014

### girlinphysics

The light coming from the sodium lamp was incident on a slit of adjustable width (we made it very narrow until the spectrum could be seen). We also used a diffraction grating.

6. Sep 22, 2014

### ehild

The spectrograph maps the image of the slit onto a screen. If it was a circular diaphragm instead of the slit you would see a circular spot with monochromatic light. If you used Mercury vapour lamp you would see circles of different colour. If the slit was made of the form of a cat you would see a cat spectrum. ;)

ehild

7. Sep 22, 2014

Staff Emeritus
A rainbow is an example of a geometry where you get circles.

8. Sep 22, 2014

### Staff: Mentor

To see this, make the slit wider, and see what happens to the spectrum lines. They should widen out into "spectrum rectangles" which overlap if they're wide enough.

9. Sep 22, 2014

### Khashishi

Some theory of atomic line transitions should have been explained in your class. You should look up some resources on spectral line shape. It's not really a line, but a Voigt profile.

The spectrum itself has the wavelengths laid out in a line because the diffraction grating is made up of linear slits arranged in parallel. If you had a different grating (say, circular bands), the shape of your spectrum would change.

10. Sep 22, 2014

### sophiecentaur

If you want to be purist, it might be better to call spectral lines 'Spectral Maxima' and 'Spectral Minima'. That would take care of the particular method you use for analysing a spectrum.
If you were to display the spectrum of a radio signal as a set of 'lines', people would complain because a two dimensional graph can show the value of amplitude on the y axis with the frequency (or wavelength) plotted on the x axis. But that's also the form in which data of optical spectra is displayed for serious analysis.

11. Sep 23, 2014

### FactChecker

If a vertical slit of light goes through a prism, the colors get sorted out nicely. A place for each frequency and each frequency in it's place. Martha Stewart would say "It's a good thing". But if the light goes through another shape (like the outline of a cat, as a previous post said), then the light spectrum of the tail would get smeared over the light spectrum of the head, and there would be a jumbled mess of light frequencies at any one spot, all from different parts of the cat. That would be useless.

12. Sep 23, 2014

### ehild

It depends on the grating constant and the size of the aperture, also on the optical arrangement. The diffracted light should be focused to a screen. My students did that experiment with a grating , using a Hydrogen lamp and apertures of different shapes. So the spectral "lines" became well separated "spectral dots", letters and traffic signs.
If you use a laser as light source, the diffraction pattern consists of spots of about circular shapes.

ehild