How to use "Correlation" argument in lme function ??

[Safwan Qasem <sqasem@yahoo.com>]

Hi all,
I need to use the function lme and especially the argument "correlation". This function is defined in the book "Mixed-Effects Models in S and S=Plus"  p.474 as follows:
lme ( fixed, data, random, correlation, weights, subset, method, na.action, control )

But my version of S-Plus 4.0 still tells me that lme does not have "correclation" arguement and that ir is defined as follows :
> library(nlme)
>  lme
function(fixed, random, cluster, data = "" start = list(D = NULL, theta = NULL, alpha = NULL,
        delta = NULL), est.method = c("RML", "ML"), re.block, re.structure = c("unstructured", "diagonal",
      "identity", "compsymm", "ar1"), re.paramtr = c("matrixlog", "cholesky", "spherical", "logcholesky",
     "givens"), serial.structure = c("identity", "ar1", "ar1.continuous", "compsymm", "ar2", "ma1",
  "ma2", "arma11"), serial.covariate = NULL, serial.covariate.transformation = c(
"rank.within.cluster", "none", "round", "global.rank"), var.function = c("identity", "power",
   "expon", "cte.power", "strata", "user.power", "user.expon", "user.cte.power"), var.covariate =
  NULL, var.estimate = T, subset, na.action = "" control)
{
## lme as a generic function
      UseMethod("lme")
}

Did I miss something ???
How can I use lme function as described in the book ??
Any help ot idea are welcome.

Hind Qasem