The paper establishes the impossibility of determining a single individual with confidence but nevertheless confirms certain statistical trends. That is significant, as the butterflies are not singletons. If you find one
aurelia, for example, there will be others - this is a meadow butterfly, found in relatively sedentary colonies. If you examine a number of individuals closely you are likely to gain a much higher degree of confidence. Also, there are asymmetries in the overlaps. For example,
athalia commonly shows a contrasting marginal line, making this feature useless for confirming
aurelia or
britomartis. But
britomartis in particular very rarely shows a line the same colour as the lunules. So this character is more useful for eliminating
britomartis, especially if you check it on several individuals.
The book I ordered before leaving for the UK is
http://www.libreriadellanatura.com/the- ... -1807.html (it can be bought from many bookshops, but that is the cheapest for me). I'll let you know if it has any useful insights. My Swiss guide, also in Switzerland, has early stage photos of all species, which I will take a look at when I return.
Aurelia, where I see it at least (confirmed by violence, not committed by me), has a much less rapid, gliding flight than
athalia. It appears dark and rather fluttery on the wing.
Finally, in the interests of accuracy, I'm sure Jack won't mind my pointing out he copied the link from a little higher up this same page, in Pieter's post!
Guy