how to transform string to variable name in a fuction?

there is a data frame, like this:

df

 aa       bb
1  a     20.27802
2  b     22.10664
3  c     21.33470
4  a     22.32898
5  b     19.73760
6  c     20.38979
.....(suppressed)
what I want to do is to copy the data frame's rows into different data frames according to the levels of 'aa' column,

df.a <- df[df[,1]=='a',] ; df.b <- df[df[,1]=='b',] ; ....
df.a

 aa       bb
1  a 20.27802
4  a 22.32898
...

So, when completed, there should be df.a, df.b,df.c, etc.
If we could do this by hand, it is pretty fine.  But could I write a loop to do this ?
when I tried this using a funciton, there is a problem.

for ( i in levels(df[,1])) {

+  name = paste('df',i,sep='')
+  name <- df[df[,1]==i,]
+ }

name

 aa       bb
3  c 21.33470
6  c 20.38979

ls()

[1] "df"   "i"    "name"

i

[1] "c"
there is not data frames df.a, df.b,etc.

Could you please give me some suggestion?
I have found that write a function in R for a beginner is difficult. Is there  any tutorial on writing the functions in R?
Furthermore, someone also said that loop is not used as frequently as in other script language (e.g. bash, perl). So, If you have any other smart means do this more efficiently, please let me know, I would appreciate your kindness.

       [[alternative HTML version deleted]]