Richard writes:
"Documents within a volume tend to be identified by <unitloc> rather
than <unitid>, so the <unitloc> becomes the prime means for referring
to an item."

Is this tendency essential?  Here in the Bodleian we are - according to
ancient Library practice - only cataloguing to file or volume level in
EAD, and using the index as a finding aid for individual folios or pages
(in general).  We are not using <unitloc> at all.  (Partly because our
shelfmarks *are* the location code, I admit - though this is not what
most UK repositories do!)
I confess I find it difficult to get used to the concept of <unitloc>;
document locations often aren't included in UK catalogues for security
reasons, but are recorded elsewhere (in the location index, arranged
by reference code, with a code for shelf, bay, box and/or folder; or
location inventory, arranged by location code; ideally both!).
Obviously if you can hide this information from the public, it is worth
including it in an EAD record, but I find the intricacy of recent
discussion of <unitloc> quite suprising!
I would also welcome any further suggestions on Richard's message.
Sarah

**********

Sarah J A Flynn
Department of Western Manuscripts
Bodleian Library
Broad Street
Oxford OX1 3BG
Direct line: -1865 277164
email: [log in to unmask]

My opinions are not necessarily those of the Library.

**********