How can I make a data frame of the unique values in an existing column?

I need to make a new data frame (col.3) using only the occurrences in a previous column (col.1) that correspond to unique values in another column (col.2) in an existing data frame.

I need this:

df1
col.1   col.2     
    1    1             
    1    3             
    1    7             
    1    7            
    2    12                
    2    14   
    2    14
    2    14

 df2
 col.3
     1
     1
     1
     2
     2 

I have tried this:

new.col <- cbind(df$col.1[unique(df$col.2)])

But it gives me a column that is both too long, and which does not include the complete set of col.1 values

I suspect that plyr has a simple solution to this, but I have not figured that (or any other solution) out.

How can I achieve my desired result? Preferably using plyr, but base is fine too.