My logic is that my g-g-grandmother had a daughter, was widowed, remarried then had another family - explaining the gap in the ages of the children. That would put the photo at c 1902. G-g-grandmother would be 29, the older girl 10.
On the 1901 census the family are living near Newport in Shropshire - it looks like Newport on the card to me.
If the photo is early 1890s then I'd be looking at a generation back, but there's no convenient gaps between children ... not sure whether there's any basis in fact for this but it doesn't strike me as likely that a woman would have a photo taken with two randomly chosen children...
D