I apologize for not getting to the bottom of this sooner: I've been observing the last three nights and also had some social obligations. Although I still feel the Pena et al. paper is best ignored, I think now that the identity of V915 Sgr is no longer in question. As Kato-san pointed out, the GCVS1 shows the position of V915 essentially the same as the current editions up to v4.1. However, the first edition usefully points out a citation to a note by Innes from Union Observatory (South Africa) Circular no. 37 from 1917. This is evidently the true first report for the variable. Luckily, we have these Circulars in the Lowell archive. It is star 126 in this list, given at 1875 position 18 33.6 -29 11, magnitude range (photo-blue) 11.0 to >17; an ephemeris is given showing maximum at epoch 2420337 + 290 days. This position precesses to 18 41 32 -29 04.2 (2000), obviously very close to the RCB-like object described earlier (Tycho-2 position is 18 41 33 -29 04.1 at the same precision). Now at Taichi's prompting I looked again at the Luyten list in AN 261,454 (1936), where the object is given again at _1900_ coords 18 35.3 -29 08, which precesses to 18 41 39 -29 02.4 (2000). A smaller range in magnitude is given, as Mati Morel noted. Still given the large scatter in Luyten's positions (and the roundoff to 1' precision in the two sources), I think it is reasonable to assume this is Innes' star, and that the GCVS (and Herbig, etc.) are correct in making this link. It was my error in giving a rather different position from Luyten's paper earlier: this resulted from his showing both equinox 1875 and 1900 positions in the list (I precessed the 1900 position assuming it was instead 1875). Thus I now conclude that V915 Sgr is indeed the one that's being observed; the MISAO star is a new variable; the mag. 10.5 red Cordoba DM star I mentioned earlier is no longer relevant. \Brian