R help - Nov 2009 - multiple regression model representation

Dear all,
I have a problem when trying to fit a model to a dataset. The model I am fitting
is: y~var1*var2*var3*var4*var5, with var3 and var4 having many 0´s.
When I construct the model and I simplify it, it remains a model with 8
explicative variables (or interaction of them) and the intercept, but when I try
to plot the model, R gives me the following error:

Error in plot.new() : trying to draw in a null device
Furthermore: missing warning message
Not plotting observations with leverage one:
  62, 78 

What I do is the following:
model1=lm(Error ~ Ciottoli *Ghiaia *Limo *Massi *Sabbia, data=datos)
summary(model1)

model1=lm(y ~ var1 *var2 *var3 *var4 *var5, data=mydata)
summary(model1)

model.error=step(model1)
summary(model.error)

model2=update(model.error,~.-var2:var5)
summary(model2)

model3=update(model2,~.-var1:var5)
summary(model3)

model4=update(model3,~.-var1:var2)
summary(model4)

model5=update(model4,~.-var5)
summary(model5)

model6=update(model5,~.-var4)
summary(model6)

model7=update(model6,~.-var2)
summary(model7)

model8=update(model7,~.-var1)
summary(model8)
plot(model8)

I would be very please if you could throw light on this.
Thank you very much in advance!
Best wishes!

Manuel
	[[alternative HTML version deleted]]