I think science writers were a bit more... creatively imaginative back in the 1800s
(e.g. I don't think they had any way of knowing how much water vapour was in Venus' atmosphere back then, or of authoritatively stating that it had "oceans" or "high mountains").
There is certainly something funny going on. And the logic that water vapour implies liquid water overlooks the existence of boiling point. But I'm pretty sure they had spectrometers in 1890, and I think that that would have let them detect water vapour provided that they were careful to correct for the absorbtion spectrum of Earth's atmosphere.