To check a bar spirit level you need a flat nearly level place, like a concrete floor, as well as a nearly vertical flat place, a column or post will do.
Check each spirit vial independently. Lay the level on the plumb or level surface as appropriate to each vial. Look at where the edge of the bubble is in relation to the reference lines on the vial. Flip the level 180[sup]o[/sup] along the vial’s longitudinal axis. If the edge of the bubble is in the same place then that vial is true. To check a 45[sup]o[/sup] vial set the level against a true plumb surface at what reads as a 45[sup]o[/sup] angle in the vial. Mark where the level touches the plumb surface and the floor. Rotate the level around its center point on an axis perpendicular to level. Set level back where it touched the floor and plumb surfaces before. Recheck vial for true.
Coincidentally, I’ve been reading The Dawn of Innovation: The First American Industrial Revolution by Charles R. Morris and this topic was discussed. Morris was writing about the difficulties of getting technology started when you lack the products that are the results of technology. And one example he gave was how do you measure whether or not something is level if you don’t have a known level surface to measure it against?
The method created was to make a couple of surfaces as level as you could to start with. Then you applied a thin coat of ink to one of the surfaces. Then you laid the second surface on top of the first inked surface. Then you lifted off the second surface and looked at the ink pattern on it.
Now if your two surfaces were perfectly level, the ink from the first surface would have contacted every spot on the second surface and you’d have a solid coat of ink. But, of course, this never happened to start with. You’d have places on the second surface that had ink on them and places that were bare - and all of the inked areas represented higher spots on the surface.
So you’d gently sand down all of the places that had been inked and you’d try again. And you’d keep repeating the process until the two surfaces fit together perfectly and the ink from the first surface touched every spot on the second surface.
And this was just the first step. Because you have no reason to think that your first surface was perfectly level. So all you’ve done at this point is make the two surfaces match without knowing if either is level. You might simply have a first surface whose high spots now matched the low spots on the second surface.
So you made a third surface and ran the whole process again using the first and third surface. And when you had those two perfectly matching, you ran the process a third time with the second and third surfaces. Eventually you’d reach a point where all three surfaces matched perfectly and the only way that could happen was if all three were perfectly level (if a high spot on the first was matching low spots on both the second and third then those low spots wouldn’t match when you compared the second and third).
You now had three perfectly level surfaces and you could use them to measure whether other objects were level.
<slight nitpick> I think that you mean that you would now have a (actually, three) perfectly ‘flat’ surfaces, along with a perfectly flat/straight level (or straight edge), don’t you?
Drill a tunnel (or dig a trench) straight from Los Angeles to New York (or Moscow to Beijing), where “straight” means, you know, straight! and lay a railroad track from end to end.
Now, set a train on track, starting at either terminal point, and let it roll freely downhill. Without even applying any power, the train will take you from Los Angeles to New York (or Moscow to Beijing) in just 84.4 minutes!
Oh, BTW, did I forget to mention that your railroad track and train wheel bearings need to be frictionless? Really really frictionless. And no air resistance either, so you need to build an air-evacuated tube around the track too for the train to run through.
I believe they set a Marshmallow Peep on each level, and then see which way the eyes are pointing. If they are pointing directly ahead, then it’s level!
Being ‘level’ generally refers to a gravitational equipotential surface. This is easy to determine because it always lies at right angles, in all directions, to a plumb line. It would also be the way the surface of a liquid would be if left to settle.
Actually, that is not quite right. A plumb line points in the direction of gravity. If the earth were a perfect sphere, that would be the center of the earth. It is a nearly perfect oblate spheroid and, as such, at any point other than the equator (and the poles, actually) it does not point to the center. It points in the direction perpendicular to the local tangent to the surface (assuming that surface can freely deform–think the surface of a windless lake) and in the northern hemisphere that line hits the axis south of the center and in the southern hemisphere it is north of the center.
Alongside Mt. Everest, there is a detectable amount by which a plumb line is deflected towards the mountain. The earth is not even a perfect oblate spheroid.
I once asked a bunch of my mathematician colleagues the question, where does a plumb line point and they all first claimed the center, before realizing the true answer. The true answer comes from a well-known theorem of multi-variable calculus that says that a level line (tangent to the surface) is perpendicular to the gradient.
As several people pointed out, the process I described creates a flat surface but not necessarily a level one. This is true and I was wrong to describe it as level. But building a flat surface is a necessary step in building a level surface.
While we’re picking nits… That is theoretically the plumb line, but in actuality, you will be measuring a line pointing to the center of mass of the earth.
