Results (fun pictures)

Figure 1: Shows the initial setup. The ``source'' xterm is vertically maximized. There are no waves on the backdrop. The three grey rectangles are my hidden GNOME panels.

Figure 2: After a rapid unmaximize/maximize toggle. This has produced a pulse of plane waves travelling to the right. Notice that the wavelength is not constant across the pulse. The shorter-wavelength waves travel faster.

Figure 3: A similar single pulse (but not the same one) which has interacted with the slits. Most of the pulse reflected back to the left, except for what was transmitted through the slits. The most distant part of the transmitted pulse has a circular profile, indicating that the slits are in fact point sources of spherical (err, circular) waves. But notice that the outgoing waves are divided into five trumpet-shaped bright sections, well-seperated by four dark bands. According to our model the dark bands should be straight, not trumpet-shaped. But our model contains several approximations which may not be satisfied here.

Figure 4: A long train of pulses, made by furiously hitting the maximize-toggle key. Now the wavelength is much shorter, so the dark bands are closer together and straighter. They become apparent where $L>d$ and only extend out to $x<L$, where our approximations were valid. There is still a hint of the trumpet shape at the front of the pulse (which is much more obvious when the pulse is moving). It would be interesting to measure the wavelengths and fringe spacing and see if it is what it has to be.