I am putting these scans here just to aid discussion.
Leaf blade - most of petiole excluded
V. nagaswai since that has green peduncles with sparse long hairs
V. arcuata since that has green peduncles, and leaves with a wide sinus
V. confusa since that has green peduncles, and blue petals, and too-elongate leaves
V. formosana since that has green peduncles, lateral and top petal tips somewhat square, long white spur, lower petal notched (don't know if that's typical), leaves pale between veins, petals pale mauve
V. obtusa (typical form) (another photo) since that has petals mainly purple-blue, other discrepant characters
V. inconspicua since that has variegated and more-deeply-cordate leaves
V. diffusa since that has different sepals, (at least in some forms) winged petioles, and no trichomes
Vv. mandschurica, grypceras, biflora (nothing like it)
But species of which I have not been able to find an illustration or description yet include:
- V. senzanensis Hayata
- V. shinchikuensis Yamamoto
- V. tenuis Benth.
...so I cannot rule these out yet.
This is the main photo suggesting BSWJ6728 is V. adenothrix. I would be happier about identifying it as that species if the photo showed reddish petioles and sepals, but of course that character could be influenced by ambient light levels. It also agrees quite well with the holotype as photographed in Akiyama and Ohba "Catalogue of the type specimens preserved in the Herbarium, Department of Botany, The University Museum, The University of Tokyo", 2001. plate 8.
I am discussing the matter with Bleddyn & Sue and others, so this is work in progress.