freda, thanks for the reply, it was helpful, here are some of my observations since then;
In every case, I've seen monarch caterpillar leave the milkweed to molt, and even seem to avoid any milkweed plant altogether when choosing a spot to J-hook on. I've also noticed of the larger caterpillars (and not just with monarchs), twitch when they feel a breeze... like they are trying to protect themselves somehow from flying insects.
The first year I had monarch caterpillars, they were on some first year butterfly weed (asclepias tuberosa) and would regularly leave the milkweed for the surrounding weeds and grass. At first I thought this meant they were quickly being eaten. Caterpillars that I hadn't seen in two days would reappear. Then I thought they wandered a couple feet from the milkweed when they did leave the milkweed. I've found that they don't tend to wander far, and it seems as though as long as they find a spot that is relatively shaded and feel secure, the distance is rarely more than 12 inches from a host plant, to which they probably always backtrack to the milkweed. From this I began being careful where I stepped when checking out the milkweed/caterpillars. Since I don't keep a very large buffer zone of mulched area around the milkweed, I tend not to mow the grass 6 inches from the planted area often, and only trim after confirming there aren't any hiding out. Not all of them would hide during the day, but at least they were usually underneath a milkweed leaf unless they were actively eating.
I have witnessed on multiple occasions: caterpillar finds non-milkweed foliage with in reach... takes the route... for whatever reason the return route is obstructed... the caterpillar can later be found wandering looking for any milkweed to continue feeding on.
If a wandering caterpillar looks to be a late instar, and you return it to milkweed: If it is looking for milkweed to eat, it will stay still for a short while, investigate it's location, wander a bit (usually upwards) then rest, content that it is at least on milkweed -- there is no guarantee it will remain there, though I think this has more to do with whether they approve of the milkweed they've been moved to; the health of the plant, or the type of milkweed. If it is looking for a molting place, it usually is curled up in your hand, generally refuses to grab onto the foliage of whatever plant you try to stick it on, investigates it's location, perturbed its on milkweed finds the nearest way off, often downward or laterally onto other foliage. Generally, the more you handle a caterpillar the more it plays dead, and the longer you'll have to wait to determine what its going to do.
Yesterday, I was wondering if maybe the caterpillars sense the warmth of sunlight and that gets them to seek some cover. Though I doubt it, since they'd all hide if that were the case.
Last year I started (8) tropical milkweed (asclepias curassavica) in a flower pot, which obviously didn't have any weeds/grass nearby for them to hide under. Only twice did I find caterpillars on the side of the flower pot.. and I don't think they were trying to hide, rather I think they were trying to move to another plant. The monarchs kept laying eggs on them and the first year swamp milkweeds (asclepias incarnata), that they kept defoliating the plants.
This year, the potted tropical milkweed (in its second year) haven't gotten any caterpillars. The swamp milkweed, the majority in their second year, have gotten 11 of the 13 caterpillars I've counted so far. I can say that on the upright ones, where the foliage doesn't touch any grass or weeds the caterpillars haven't moved off of the plant. However, I did notice twice where caterpillars had wandered off the tall plants and found there way to some of the first year swamp milkweeds. Of these two, one I think had been in the grass before I trimmed it, probably had to find another path back to the milkweed. The other I'm quite sure I saw hanging out low on a second year swamp milkweed and made the mistake of migrating downward after finishing off the leaf it was eating instead of upward where more foliage was. -- This was an interesting observation for me, because last year, on my bluevine milkweed (cynanchum laeve) if I saw a caterpillar low on the vine I could count on it being in the grass/weeds, as long as it hadn't been eaten. Instead if it was last seen high on a vine then it almost always was on the underside of a leaf, though on two occasions I found them on the underside of the fence the, which the vine was climbing.
I'll make seperate post of my conclusions.