change read.table by scan

Hi all,

Before the amount of data given has grown i was initially using 
read.table to load the values inside R.

It was feeding my needs because i could tell read.table h=T, then use 
attach to access the values by columns names.

Now it takes 20 seconds to load the data's and the first enhancement i 
could do is to win some time on the load of the data's...

How could i use the scan function (which is faster) to build the data as 
if i've used read.table("file",h=T) then attach?

/* the code i use actually */
tab<-read.table("blob/data.dat",h=T)
attach(tab)
/* is it possible to produce the same using scan? */


thks all for understanding my question.

guillaume.