According to the help page, tables may be given captions, but this does not work properly when the viewer is using the Brunel skin - the caption can only be seen when the invisible text is selected (by using the mouse). It does not matter whether the table is created using the Brunel skin - it is the preference of the researcher trying to read the entry which makes the difference. The following is suggested as a simple work-around, and gives results which are very similar to those obtained by using the <CAPTION> tag in the other skins, but which is visible at all times in all skins.

The prototype table, with <CAPTION>, would be have the form
<TABLE BORDER="1">
<CAPTION>
The Caption
</CAPTION>
<TR><TD>
lines of table data ...
</TD></TR>
<TR><TD>
... as entry requires
</TD></TR>
</TABLE>

giving the appearance
 

but this can be approximately emulated, in all skins, by means of

which is displayed as
 

This workaround does not properly enlarge caption text in Alabaster or Goo, but does give it a little prominence. However, the work around uses only GuideML which is approved for use in Edited Guide entries.