In conclusion, the analysis of three years of CFHT seeing records
shows that mirror seeing is the only significant local seeing
effect generated inside the observatory. It is found when the mirror
is warmer than ambient air and is correlated
with the surface-air temperature difference at a rate of 0.38
arcsec/K
, which confirms the outcome of
the analysis by [Racine 92] of a smaller set of HRCam data.
[Racine 92] noted also a correlation between seeing and
a positive temperature difference 
between the dome
interior and the outside air with a factor of 0.1 arcsec/K
.
The HRCam data analyzed here
are not incompatible with a trend of this kind, but too few data
points exist for 
to draw a positive conclusion.
The range 
K is better represented in the
FOCam data but in that case no apparent trend is detectable
(fig. 
). Thus we do not have conclusive evidence
for the existence of this seeing contribution.