Basic Question on error(_(...

I'm sorry to bother this list with such trivial questions, but I'm
trying to take Prof. Ripley's advice in porting some Lapack wrappers
into my own code, because they are not public.  I'm specifically
choosing programs from R-2.5.0/src/modules/lapack/Lapack.c.  

I can't seem to understand the context in functions such as:

<snip>
        error(_("the leading minor of order %d is not positive
        definite"),
              i);
</snip> 

in functions such as modLa_chol.  Is there some place I can look to
understand the usage of the underscore '_', or could someone explain
this to me?

Thanks for the help!