R help - Jun 2003 - error terms in mixed ANOVA's

Hi,
	Newbie here, just joined the list.
	I discovered something interesting when performing a mixed between and 
within (repeated measure) ANOVA in R.  The error terms (SS actually I 
guess) are inflated by the number of factors of the within measure.  
This just seems wrong.  I think I'm doing it right

e. g.

subj	grp	test	x
1	1	1	...
1	1	1
1	1	2
1	1	2
2	2	1
2	2	1
2	2	2
2	2	2


So, the anova is done with

aov (x~ grp*test + Error(subj/test))

	Is there something I am doing wrong here?  On a related note can I 
properly analyze a subset of between measures by adding something to my 
aov command instead of aggregating the data?

thanks
John