bison info doc - precedence in recursive parsing

[anand]

Hello,
    in info doc of bison, it is mentioned that rule gets its precendence from 
last terminal symbol.
pasted below:
The first effect of the precedence declarations is to assign precedence levels 
to the terminal symbols declared. The second effect is to assign precedence 
levels to certain rules: each rule gets its precedence from the last terminal 
symbol mentioned in the components. 
 
For below example, how the parsing will happen if above is the case:
 
nonterm1 : nonterm2
                 | nonterm1 term1 nonterm3 nonterm4
                 | nonterm1 term2 nonterm5 nonterm6
 
One more example:
nonterm1 : nonterm2 nonterm3
                  | nonterm1 term1 nonterm2 nonterm3
                  | nonterm1 term2 nonterm2 nonterm3
 
Consider first example. First parser will consume nonterm2  then for term1 
rule, nonterm1 will hold nonterm2 value and term2 higher precedence than term1, 
parser have to pass value to nonterm1 in term2 rule but it is invalid if we 
provide term1 precedence to term1 rule because as the statement suggests, last 
terminal token gives rule precedence. Assigning value of "nonterm3 nonterm4" to 
nonterm1 in term2 rule is not possible or is it?
Value is for term or nonterm but not partial set of symbols in rule. is it 
right?
Does the info doc says last terminal symbol as last but one symbol which if it 
is terminal?
 
Regards,
Anand