[NHCOLL-L:2205] RE: Numbering specimens

[Panza, Robin]

From: Raney Morrison [mailto:raneym at yahoo.com]
Would you number everything with the same tri-nomial system and pick an
arbitrary date to start from?  Would you number all the birds under one
system, mammals under another, etc, etc?  Would you try and retrace
specimens in the limited records we do have and match them to a date and
donor?   
 
How you number your specimens depends in part on how you expect to need to
find them, especially if you are planning a computer catalogue.  Most of the
natural history collections with which I'm familiar (I don't claim to have
seen most collections) do not use the pseudo-decimal systems that art and
some anthro collections use (year.accession.item or something akin to that).
They use simple integers, starting at 1.  That's the easiest to use.  No
worry about things like 2-digit years that have to later be expanded to
include another century.  No worry about how many leading zeroes you'll need
to cover a bumper year of acquisitions (so you can sort and 1993.99 comes
before 1993.100).
 
If you want a computer system that will retrieve data with a single query, I
recommend a single sequence of catalogue numbers.  You can always include
fields for taxonomy and nature of specimen (study skin, grouped mount,
free-standing mount, etc.) so that you can retrieve data by these factors
(only birds, only independent mounts).  If you start subdividing the data
into separate catalogue files, you may have to generate multiple queries to
get, say, everything from a given location or donor--one query for each
file.  Also, once you start subdividing, where do you stop?  You suggest
each class in a catalogue sequence, but why not each order, or each genus,
or each style of mounting, or each donor...?  Today's computers have enough
RAM and speed to handle very large files efficiently, so combining the data
into a single sequence is not a problem from that point of view.
 
For the grouped mounts (e.g., Victorian trees), I recommend cataloguing each
specimen and indicating in your catalogue that it's mounted with ##x-z.
That way, anybody who breaks up the group won't be faced with now-multiple
specimens with the same catalogue number, but you will have saved the
information that they were together.
 
There is a factor in favor of subdividing and cataloguing separately,
however.  The more diverse the set of objects, the more fields you need to
include and the more fields will be blank in each record.  For example, eggs
are unlikely to have known sex of the removed embryo; insect records might
include what species of plant (host) they were collected from, but that's
unlikely to be in a bird record; fossils need stratigraphic information
that's irrelevant to things collected alive.  More empty fields means less
efficiency in data entry and wastes computer space (although the latter is
not the problem it used to be--see above).  When the Association of
Systematics Collections (now Natural Science Collections Alliance) tried to
develop a model record that would cover all collections, and the result was
horrendously large and complex.
 
Personally, I think it's valuable to try to match specimens to donor/date,
as best as can be done.  The situation will only continue to deteriorate, so
the longer you wait, the less you'll be able to match up.  And donor
information can be useful.  We have people ask about specimens donated by
grandpa (or something salvaged by themselves) and I can retrieve that data
and show them--good will.  I have also helped with research by biographers
of old-time collectors who deposited material here--community service.  None
of that would have been possible without donor information.
 
Robin K Panza
Section of Birds, Carnegie Museum of Natural History
4400 Forbes Ave., Pittsburgh  PA  15213-4080
ph:  412-622-3255;     fax:  412-622-8837
panzar at carnegiemuseums.org <mailto:panzar at carnegiemuseums.org> 
 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.yale.edu/mailman/private/nhcoll-l/attachments/20040211/7381b6b5/attachment.html