To add to the issue of the use of protoconchs, I offer the following:

Among the Neritidae there are species with a standard veliger stage (all of
the marine Nerita and as well as most freshwater species) and those with
crawl away juveniles (plus some very interesting variations).  The direct
development or crawl away juveniles are limited (as far as I know) to the
freshwater species of Theodoxus - the small, stream and lake dwellers from
Europe, the Mid-East, and northern Africa; and the American species Vittina
usnea (Neritina reclivata of authors) and Clithon meleagris (Theodoxus
meleagris of authors).  According to Klaus Bandel a telling physical
characteristic that marks this difference is the protoconch.  The protoconch
of nerites with a veliger stage is smooth while the protoconch of the crawl
away types is axially ridged.  This axial ridging is also found in the
related Pisulina - an ocean cave dweller that most likely uses direct
development.  Why the difference?  Perhaps it has to do with diet.  The
species with a veliger stage build an embryonic shell from yolk and add most
of the protoconch as active, feeding veligers after hatching.  On the other
hand, the direct development types use yolk for the embryonic shell but then
continue to build the protoconch while still in the egg case - nourished by
"food eggs," (unfertilized eggs there expressly for nourishing the
fertilized egg) or more yolk.

I offer this in the argument of "what physical features can be used to
determine a species," because I believe you have to take each family/genus
individually to decide what features are determinant as well as use a
mixture of physical traits.  For years people have (overly) relied upon
radular characteristics - sometimes to the exclusion of all other
characteristics.  In fact the 1866-1893 work by Troschel and Thiele has 50+
pages devoted to classifying the Neritidae on just radular characteristics.
The problem is, the radula can vary within a single species of nerite so
much that it could be linked with a number of other, closely related
species!

I am not saying we need a mystical "gestalt" process for species
identification - but we do need to look at all that is available to us.
This includes the supposed magic bullet of DNA.  As this becomes more
available it should be considered just another useful marker and not the
"final " determinant.

A species is an artificial box we have constructed to aid our understanding
of the world around us.  I believe Tucker Abbott once said (grossly
paraphrased here) that a species is what a local inhabitant recognizes as
different and set apart from other local animals and plants - it looks
different.  There is no true, always 100% correct test.  There seems to be
an exception to every rule used to determine species.  Sorry to sound a bit
preachy here - but this subject hits a nerve.  I have spent the last few
years working on the Neritidae - a family that experts list has having 50
species while other experts say 250!

And with that, I'll go back to my corner.

Tom Eichhorst in New Mexico, USA