This was an interesting test with interesting results. Firstly, looking at the chain, it appears that there are no links as we see in a more "conventional" necklace or bracelet....this chain was described as a "rope chain", and that is what it looks like in the photo...so, no links, but continuous strands.
My own experience is that "conventional" chains come in on the TID below what you would expect for an item (of silver) which was solid, but having the same weight as the chain...how far below depends on the size of the links. The theories about the detector only picking up one link, or TID depends on whether the links are touching each other, or whether the chain is in a ball or stretched out, in my view, all carry some weight and all have some bearing on the response we receive from our detector. Most of the very small silver chains that I have found have all had, as part of their make-up, one small part that could give off a reasonable signal, eg, the clasp, a small ID plate or a charm of some sort. Chains with large links (that I have found) usually have larger clasps (if the chain is still fully intact), others have clasps of a similar size to the links...these larger chains, while not coming in high, or in the "silver range", still come in high enough to be closes to the lower end of the silver range, or similar to a very small silver ring, on some detectors. On my X Terra 70, a fairly solid chain bracelet (without any charms etc) comes in at the higher end of the gold range. This is using the HF 10 x 5 DD coil or the MF 9" concentric coil.
The responses from this rope chain seem strange to me, but I would be interested to see what responses it gave if detected in the ground, whether coiled up or stretched out. Is it possible that there is not as much silver in the chain as 92.5%. HH