Error in COMPUTE/ST --> affects COMPUTE/AIRMASS

["Dr William Tobin, in Vannes, France." <tobin@club-internet.fr>]

I've found an error in compute/st, which I shall also be reporting to
the midas team.

   compute/st can produce sidereal times greater than 24:00:00.  

Example:

> Midas 003> compute/st 1995,09,11 14,56,59 170,27,54
> Julian date:                 2.449972122905E+06
> Local mean sidereal time:    25 : 39 : 28.58

which is wrong: the correct LMST is 01:39:28.58

The error arises because in .../prim/general/src/compxy.for it has not
been appreciated that the quantity STR can exceed not only 24 but 48.
The statement
     IF(STR.GE.24)STR=STR-24.
needs to be given *twice*.

The error will occur less frequently for western observatories, such
as La Silla, than for far eastern ones, such as the University of
Canterbury's Mount John University Observatory.

This error in compute/st has a knock-on effect on compute/airmass, and
in fact that's how I doscovered it.  If the sidereal time is greater
than 25, compute/airmass calculates the meridian airmass, not the actual
one.

The error is present in 97NOVpl2.0

Best wishes,
William Tobin.

-- 
**********************************************************************
* Dr William TOBIN, Department of Physics & Astronomy, University of *
*                   Canterbury, Private Bag 4800, Christchurch, NZ.  *
*    at present at  6 rue Saint Louis, 56000 VANNES, France          *
*                   Tel & fax: +33-2-97-40-71-18                     *
* PLEASE FAX IN WAKING HOURS ONLY - 06:00-21:00 UTC  18:00-9:00 NZST *
*                   E-mail: tobin@club-internet.fr                   *
*   I'm only checking e-mail weekly: fax anything urgent             *
**********************************************************************