Starting with your cylinders around arbitrary minor axes, you're definitely doing a good job of checking your ellipses' alignments with their 'true' minor axes, and that is coming along well. I'm also pleased to see that there is a mixture of different kinds of foreshortening, between those that are more shallow and those that are more dramatic.

In looking at the more dramatic foreshortening, I did notice one mild issue that is worth pointing out. With cylinders - or really any form, but it's easiest to discuss in relation to cylinders - when you look at the two opposite ends, there are two kinds of shifts that occur. One is the basic shift in scale, where the far end is understandably smaller in scale than the end closer to the viewer, and the other is the shift in degree, where the far end gets proportionally wider. Both of these shifts equally signal to the viewer the amount of foreshortening, and therefore the amount of relative distance between the opposite ends upon which we base our estimation of the cylinder's length.

If you take a look at cylinders like 106 and 107, you apply the shift in scale, but did not apply the shift in degree to the same extent. That means that if you look at the shift in scale, you'll understand that the cylinders are quite long, but if you look at the shift in degree, they'll appear to be shorter. This contradiction can be noticeable to the viewer, even if they aren't sure what precisely looks off. Always make sure that these two kinds of shifts are consistent with one another, either both shallower or both more dramatic.

Moving onto your cylinders in boxes, you have definitely been thorough once again in applying your error-checking, this time with the variety of line extensions. This exercise in particular is all about developing a student's ability to eyeball the construction of a box that features two opposite sides which are roughly square in proportion. We can develop this instinct by checking each time how far off the ellipse's lines were from converging towards the vanishing points of the box upon whose faces they sit. As they get closer to aligning, that ellipse gets closer to representing a circle that sits on the box, meaning that the plane enclosing it gets closer to being square.

You definitely have room to grow in this area, but as you reached towards the end, your convergences really started to become much more consistent, resulting in more reliably square faces to your boxes. This is something that will help you quite a bit as you move into the next lesson.

As such, I'll go ahead and mark this lesson as complete. Keep the points I've raised here - particularly about foreshortening - in mind, but otherwise keep up the good work.