I was recently forwarded a journal article dealing with tektite shapes (Stauffer & Butler, 2010) in which it is postulated that toroid-like (doughnut-like) forms are a logical end-point of radial flow and consequent central-thinning of a tektite disk spinning like a frisbie. A torus is to a revolving spheroid what teardrops are to rods and dumbbells. These authors do not picture a complete tektite torus, but do suggest flat-ended curved cylinders as possible torus fragments. I sorted through our collection looking for examples. Figure 1 shows the prime candidates.

I also located two good examples of what would be the logical previous stage in the evolution of a torus. These are very deeplydished disk fragments. It is easy to imagine central concavities forming by radial flow in a spinning patty. Once this happens, frontal flight pressures
would accentuate the dishing, like blowing soap-bubbles. (I have often wondered if this is how big oblate sphereoids with large central bubbles form?). Figures 2 and 3 illustrate these intermediate stage specimens. From the form in figure 3 to the rupture of the centrally-thinned area forming a doughnut-shape is a short hop. I’ve never seen a complete unbroken Indochinite ring (if you’ve got one I’d love to hear from you—) but McColl (1997) describes a flanged Australite torus (which does not seem to be a detached button flange).


References
McColl, D.H., 1997, A Flanged Toroidal Tektite from Australia: Meteoritics & Planetary Science, v. 32 No. 6, pp 981-982.
Stauffer, M.R., and Butler, S.L., 2010, The Shapes of Splashform Tektites: Their Geometrical Analysis, Classification, and Mechanics of Formation: Earth, Moon, and Planets, V. 107, pp. 169-196.