Http loc not a citation

From
Date
David Schwimmer suggests that e-journals will introduce undesirable novelties
in taxonomic citations, e.g., 

         _Xiphactinus audax_ Leidy, 1870, p. 212, fig. 23
         _Portheus molossus_ (Cope, 1871), p. 45, fig XI 
         _ X. audax_, Weasel, 1996, HTTP://www.taxolist.ggg.edu       


An http address should not be included as part of a formal citation.  Would
you include a library address in a traditional citation (e.g., Smith, 1996.
My article, Journal of Paleontology, 234(32):16-45, UC-Berkeley Library, 76
Orange St., Berkeley, CA 00000)?

E-journals should be found many places, as paper journals are. It is not up
to the author to tell the reader where copies can be found.  Rather, it is
the responsibility of the knowledgeable and ingenious reader to discover the
location where the journal resides. It is only because of our fascination and
obsession with this new Web technology that we even pay attention to the http
addresses.

Of course, the reader should expect that the e-journal will be found in
predictable places.  Thus, it will probably become incumbent upon research
libraries to maintain web sites that either contain e-journal files or mirror
the true localities of those files.

With major institutions becoming the repositories for e-journal files (will
"file" replace the term "article"?), it will be more likely that changing
storage formats will not be obstacles to access.  Libraries will periodically
have to update their storage formats, hardware, and software, but the user
should always be able to download the files in whatever format is current.
 Formats may improve with time, making older files seem antiquated, but the
same is true looking back two hundred years when one compares line drawings
with photographs.