How to assign barplot/plot to a variable and recall at later stage

Question

How can I assign plot to a variable and have it printed it at later stage? With lattice xyplot you can do it easily:

s <- xyplot((1:10)^2~1:10)
s

However not with base plot

a <- plot((1:10)^2,1:10)

What I have got is a barplot in a function:

funbar <- function(x) { barplot(x) }
variab <- funbar(1:10) # this plots the data

So I would like it to print at later stage by calling variab

In fact I have additional graphical functions in a function, something like this:

funbar <- function(x) { barplot(x); points(....); text(....) }

I could not find any info on this obvious part. The plot=FALSE doesn't help here.

EDIT: Here are the data for barplot taken from ?barplot

require(grDevices) # for colours
tN <- table(Ni <- stats::rpois(100, lambda = 5))
r <- barplot(tN, col = rainbow(20))

EDIT 2: To answer Josh question as of why I would like to do this. Example:

layout(matrix(1:2, nrow=2, ncol=1))
tN <- table(Ni <- stats::rpois(100, lambda = 5))
r <- barplot(tN)
tN <- table(Ni <- stats::rpois(100, lambda = 5))
r <- barplot(tN)

Answer 1

Perhaps this can be solved with xy.coords:

a <- xy.coords((1:10)^2,1:10) #plots nothing

And can be plotted later:

plot(a)

I came upon this idea by examining the code for plot.default.

As for barplots, perhaps the rect function will accomplish something similar (as this seems to be at core of what barplot.default is doing).