Many speculations have been done concerning hints for physics beyond
the standard model.
One of the promising areas, where such physics could be studied, is the
decay of charmed mesons.
D-mixing or CP-violations are predicted within the standard model to be very
low (r 
10 
and 
10 
, respectively)
[9] and will most likely
remain unobserved.
However, this drawback could be turned into a virtue since no standard
model background is expected for measurable unexpected effects.
Table 2.1 shows some of the best current limits on rare decays or other phenomena.
Table 2.1: Current limits on some rare decays channels
For COMPASS, we estimate the limits attainable
on the basis of 3 10 
reconstructed D 
-decays and about 1-2
10 D 
decays (the same number of CP conjugate states should be
reconstructed).
It should be noted that the limits on the rare decays decrease with 1/N only
in the case of
no background. When background becomes important upper limits only decrease
with 1/ 
. At this stage, possible backgrounds are difficult to
estimate. Most limits may therefore be regarded as lower bounds on
those limits.
- 
mixing. Mixing can be observed in
different ways. Certainly the cleanest way is flavour tagging by
means of full reconstruction of the associated D-meson. We estimate that
we may detect mixing if r 
is 
10 
.
Higher statistics may be obtained if c-flavour tagging is
performed by tagging of the decay 
.
However, this method is limited by backgrounds from double Cabibbo
suppressed decays (DCSD) which may in part be reduced by observing the
time evolution of the D-decays. Limits which may be achieved are of the order
10 (1/ behaviour) - 10 
(1/N behaviour).
Competing experiments are E831 with about 1/5-1/10 of the expected COMPASS
statistics but with the advantage of cleaner events ( 
-production).
CLEO III and BaBar will have similar discovery potentials with BaBar
expecting about 10 
D's of each kind produced in total.
Their estimated sensitivities to mixing are in the order of
10 -10 
.
Again, a 
/charm factory would probably do better than COMPASS or any
B-factory.
for direct
CP-violation (assuming a favourably large FSI in Cabibbo suppressed decays)
the D-system is the most unfavourable for standard model tests.
Samples of 10 -10 
reconstructed charmed D-mesons are needed to
possibly observe a signal.
According to various models, possible FSI giving raise to CP-violation might be
enhanced.
In particular the Penguin diagram, responsible for
a possible effect in Cabibbo suppressed D-decays, might get contributions from
exotic meson states like the 
(1800) [8].
Promising decay channels are D 
K 
K and
D 
K 
, K 
' [9].
If no background is present, sensitivity in 
of about
0.005-0.01 may be achieved where
.
where 
(D) is the decay rate for a given decay channel.
Competition again comes from E831 and B-factories. E831 expects a sensitivity of of about 2-3%, B-factories may be sensitive to asymmetries below 1% [10].
Indirect CP-violation is not discussed here since its observation is
even less probable due to the low -
mixing expected.
.
This decay channel has very good acceptance and trigger efficiency in COMPASS.
We should thus be sensitive to branching ratios of the order of
10 - 10 
, one to two orders of magnitudes below current limits.
A possible feed-through of D 
has to be taken into
account and good / 
separation has to be achieved.
+ cc (lepton number violation)
is less favourable due to bremsstrahlung
effects and the more difficult task of clean electron identification.
Still, one can hope to obtain a similar improvement on upper limits as in
D .
Competition again comes from E831 which may reach limits of order
10 , about a factor
10 above COMPASS, and from the B-factories, which, however, do not quote
any numbers. CLEO II has reported upper limits of 3 
10 and
2 10 on D and
D 
e 
, respectively.
The tenfold increase in statistics expected by CLEO III for the year 2001
will decrease these limits to between
5 10 and 10 , assuming the limits scale as 1/
[11].
Other interesting decays which only can be studied with high statistics
charm samples are 
S=1 transitions in charmed strange systems.
Such decays (e.g. 
) should have branching
ratios of about 10 . The topology of such a cascade decay resembles the decay
of doubly charmed baryons with the exception that only one additional D-meson
is present in the event.
Again, excellent vertex resolution will be needed to sort out this decay
and distinguish it from 
which has a
mass only 16 MeV/c 
lower than 
. About 25-50 such events might be
observed.
The observation of this decay gives information complementary to S=1
decays from hyperons and may shed more light on the I=1/2 rule
[12].