I'm going to assume that the circled numbers are the page numbers. Most students tend to actually number each individual one, but as long as the pages are numbered (which I think they are?) I should be able to go through them despite imgur mangling the album order as it often seems to.

I think you've done a pretty good job with this challenge, and despite your own uncertainty, you've largely done a good job of going through each ellipse and identifying the correct minor axis. I did notice a couple here and there where the minor axis you'd drawn after the fact still wasn't entirely accurate, but they were few and far between.

One thing I did want to mention however has to do with understanding what different aspects of a cylinder and its ellipses tells the viewer, and whether or not all of those assertions remain consistent with one another. To explain this, let's look at the longer cylinder on the bottom left of this page. Here you've got two ellipses, roughly the same degree (both on the wider end), the far end being somewhat smaller in scale, and there being a fair bit of distance between the ellipses. Here's what each of these tells the viewer:

• Minimal degree change between near/far end ellipses tells us that the ellipses are very close together.

• Notable change in scale between the near/far end ellipses tells us that the ellipses are somewhat farther apart.

• Both near/far ellipses are on the wider end of things tells us that the ellipses are oriented to face the viewer.

• There being a fair bit of distance between the ellipses, in combination with the wide degree on both ellipses, tells us that the cylinder is very long (and therefore the ends are quite far apart in 3D space.

As you can see, these assertions are not entirely consistent - specifically, the fact that there isn't much of a shift in degree between the far and near ends isn't consistent with what the other assertions say. As such, you'd want the far ellipse to be wider, or the nearer ellipse to be narrower.

To keep it simple, you can think about foreshortening - where typically the far end gets smaller and the closer end gets bigger - to also incorporate the proportional width (the degree) of the given ends. As the far end gets farther away, it gets wider. This will work hand-in-hand with the far end getting smaller in scale, so you're never going to end up with the degree shifting a lot, but the scale shifting very little, or vice versa.

Throughout your cylinders in boxes you've largely done a good job, though I did notice a few boxes that looked roughly like this. The key issue here is that your far end ellipse gets so wide that it ends up aligning its major axis to the vanishing point its minor axis should have pointed towards. Because of this, the structure of the cylinder falls apart, with the far end no longer representing a circle in 3D space. This of course occurred because the box itself wasn't constructed correctly, with the length-lines converging to pairs rather than a single shared vanishing point, in turn causing that far plane to end up much wider than it ought to have been.

The thing about this exercise is that it is ultimately more about having students learn how to more intuitively draw boxes that have face-pairs that are proportionally square (or approximately so). When the boxes don't quite get close enough to this goal, it throws the ellipses off, which in turn makes those check-lines for the minor axis and contact-points go out of whack.

All things considered, I think you've done a pretty good job of improving on this front. Just remember above all else to think about the tendency to have lines converge in pairs, rather than having all four lines of a given set of parallel lines converge consistently towards their shared vanishing point.

With that, I'll go ahead and mark this challenge as complete.