The UNICODE language has two letters "µ" which are written identically but which have different meanings - one with the value 0#00B5 is described as the "MICRO SIGN" and is the Latin-1 page while the one with the value 0#03BC is described as "GREEK SMALL LETTER MU" and is in the Greek page.
There was a case for not using the "µ" symbol some years ago when printers used pure ASCII (ie the character set 0#0000 thru 0#007F) and the British variant was to replace the "#" symbol with the "£" symbol. Those days are now past and apart from the United States (who can still use pure ASCII), all other countries who use the Western European Latin script include "µ" in their character set along with "£", "þ", "Ã", "ã" (having values in the range 0#0080 thru 0#00FF) etc. Certain European countries (for example Germany), have a key marked "µ". ----- Original Message ----- From: "Pierre Abbat" <[EMAIL PROTECTED]> To: "U.S. Metric Association" <[email protected]> Sent: Sunday, December 11, 2005 1:43 AM Subject: [USMA:35377] Re: blood tests > On Saturday 10 December 2005 16:26, Anon Anon wrote: > > That is widespread in US medicine. American medical > > sources say that 'µg' should not be used. They > > recommend 'microgram' or 'mcg'. The stated reasons are > > that it could be mistaken for mg or ng. The same > > reasoning would apply to any use of the prefix. See: > > http://www.med.umich.edu/prmc/star/archive/2004/0107/abbrev.htm > > > > The British medical view is that 'mcg' must not be > > used. See: > > http://www.pjonline.com/pdf/hp/200206/hp_200206_exercises.pdf > > http://www.firstdatabank.co.uk/pdfs/multilex_bulletins/September-2005.pdf > > I don't see how 'µ' could be confused with 'm' or 'n'. 'µ' is concave up and > begins below the line; 'm' and 'n' are concave down and entirely above the > line. 'm' could be confused with 'n' though. > > As to computers miswriting the symbols, they are probably using obsolete > software. 'µ' requires two bytes in UTF-8, and if the computer is set up only > for ASCII, it's not available. There are still lots of applications using the > punch-card mentality, in which all fields are fixed-width and all letters are > uppercase (it is possible to encode a lowercase letter on a punched card, but > it was very rare back then). As databases nowadays allow arbitrary-width > fields in any of several character codes, I consider this inexcusable. > > phma > >
