I have used grid lines for years "E" screens on my "F3" bodies and
similar on my FM2 bodies. I now use the "on demand" grids in my D200
viewfinder.
Handy feature, that.
I also learned, young, how to take a screen and scribe my own lines with a carbide scribe or a Lunzer diamond scribe. I like to add a light 8x10 crop line set to the screen.
E screens aren't really rule of thirds, they're squares, 4x6 pattern (3 and 5 lines) so you get ROT only on the horizontal.
I honestly find the rule of thirds "close" but not quite right. The
actual Golden mean isn't quite 1/3 on a 3-2 aspect ratio.
Hey Dale!
The "golden ratio" is different from the "rule of thirds". They both have their origins in the superstitious veneration of mathematics that is similar to human visual processing (optical psychophysics) , but they are two distinct phenomena. Think about the "golden ratio" defining a pleasant aspect ratio for the "box" or "Frame" of a photograph, and the "rule of thirds" determining how we place objects in the box, whether or not the box has a golden ratio shape.
Phi, the golden ratio
============
Phi, (sqrt(5)-1)
2, about 1.618, is often known as the golden ratio, golden section, golden mean, or divine proportion. It has its origins in the shape of the human visual field. We find rectangles of that aspect ratio appealing because that is how the eyes and brain scan a scene. Horizontal photographs in that ratio can be "entered", we can easily process them as a "scene" itself, instead of treating the photograph as merely an "object". Since it's a shape we process well, it also works out wonderfully in architecture, a golden ratio building can become a visual scene in its own right, instead of being an object in a larger scene, so it's wonderful for an important building that should "capture" our attention.
The ancients (especially the Greeks) were very fond of the concept of interesting and important numbers that gave us views into the nature of the universe. They knew how to calculate many numbers, such as phi (around 1.618), e (2.718), pi (3.1416). When one of these "magic" numbers happened to fall close to a psychophysical or biological proportion, they took it as a sign.
The Rule of Thirds
===========
The rule of thirds is also often erroneously called the "golden mean" or "golden section". It also relates to human visual processing. Humans are (or were created to resemble) highly evolved hunter/gatherers. We have two modes of operation, "scan" and "lock". Basically, we are designed to continually scan a scene, analyzing all the interesting details, and taking short "fixes" on interesting features. Walking, we scan for hazards that might trip us, obstacles to avoid, predators, prey, food sources, the other members of our tribe, possible mates, etc.
We do not stop scanning until we see something so important that it needs our complete attention. That's when we lock and center, and prepare a reaction (fight or flight, court a mate, eat, defend or rescue offspring, etc).
That's why the rule of thirds works. A major subject at center has to be something of such import that we would lock onto it. Prey (object of desire) possible mate (object of a different sort of desire), danger, etc. As modern humans, we sublimate the primitive object of desire concepts into concepts such as "object of great beauty", so we can center, for a while, on a really beautiful flower.
It's difficult to make a great photograph with the main subject centered. If the subject isn't so interesting that you're comfortable centering on it every time you see that picture, it's not "habitable", you can't really live with that piece of art.
Enter the rule of thirds. By moving the main subject off center, we fool human visual processing. We say "here's something moderately interesting, but not so overwhelming that it needs to be locked onto and centered". This can make the image more habitable. The rule of thirds points are simply good locations for doing this. Subject off center, but not too close to the edge. Something interesting close to the edge of the scene causes us to recenter the scene, turn head or body to a direction that puts this new interesting object closer to center.
It's not an exact rule, but it's close. go too far inward of the ROT points, and you have a centered composition again. Too far outward, and you're too close to the edge.
So I use the grid primarily for horizontal and vertical alignment. I
tend to place my "almost third" close to the grid line just outside
of where a "real" one third is.
That works...
Like most "rules" you have to use your own judgment for what looks
"right" when you click the shutter.
Exactly. Part of it is determined by the "visual weight" of an object: that combination of brightness, color, shape, resemblance to known a known D4 (differing degrees of desirability or danger) objects.
For landscapes, I have started using an old red cokin filter masked
to a 3-2 frame. I then used a sharpie for 1/3 grid lines. when
judging a scene, before picking up the camera I can look through the
filter and place subject areas on the grid based on light/dark mass
rather than the subject itself. It enables one to view the scene for
its strictly graphic content independent of subject or color.
Excellent. By removing "color", a high priority aspect of visual weight, you're able to visualize other aspects of visual weight separately.
Sort of like the way view camera adherents consider viewing the scene upside-down to be an advantage, they remove the D4 (recognition of familiar objects) aspect.
Some photographers were such adherents of this that they used left-right swapping cameras (like TLRs or MF SLRs without correcting prisms) turned sideways in order to invert the image. I've even heard of people doing this with a Leica VIDOM finder.
--
Ciao! Joseph
http://www.swissarmyfork.com 
