I have apparently reached the age where I should look before I write. I do seem to recall that Joe moved the mark on the 1896/1898, but it must have been from 1092xx to 1091xx (his book not handy at present) not dipping all the way down under 109xxx. I hope he corrects me.
With all of the above responses noted, I still think the likehood of a blip at that particular changeover is much less likely than at any of the others. Not saying it could not happen, just saying it is unlikely.
Here are a couple of earlier examples with an obvious difference in the numbered part itself. Bear in mind that these examples were made by basically the same work force, on the same machinery, under the same basic system of operation, only a very few years apart.
(1) The trapdoor receiver underwent a significant change in late 1878, somewhere between 96271 and 96309. NO overlaps have ever come forward.
(2) The 100 receivers made in 1888 (from 4155xx to 4156xx) for the positive cam rifles, run (including a handful of overuns) consecutively. That is not random chance, some of the "experimental" models span many thousands of numbers.