That is up to your implementation, but you can assume that the size of an int is the natural word size of your system. A 32 bit system will most likely have 32 bit integers, so the largest value is 4,294,967,296. You can only assume the range of -32,767 to 32,767 for int if you want your code to be portable though. If you want at least a 32 bit data type then long int is the way to go.
I don't understand what you are asking.

There is a file called LIMITS.H that has all the values you are looking for.

well if you have a float value that is too large for a normal float you can us a double to give it twice the normal memory usage...

i was wondering if there was an int type of variable to do the same.

i am basically working with addition of two hexadecimal numbers each of which can be up to 20 characters long, when i add two of these i am storing it in an int variable, and i am thinking the reason i am having problems with the output is becuase the answer it returns is too large to be stored in the int...

A 20 character long hexadecimal value would exceed any of C++'s types. Try an arbitrary precision library such as GMP.

well the project is for a intro level c++ class , so it seems like the prof wouldnt require us to do soemthin that required libraries other than the ones in code warrior to start with. im thinkin im gonna have to store the value as a character array and only convert it to actual numbers when i need to manipulate them, that way i dont have to deal w/ the whole thing as one huge number, jus go one slot at a time thru the array. thanks alot for speedy replies by the way

Exactly. I've never found the need to do something like this with binary numbers, but before co-processors I coded this kind of thing on the Z80 to give me 128 digit precision in BCD (Binary Coded Decimals).

The largest int ist 2^31 = 2147483647.
One bit is used to store the sign.

Yep, that's why we make assignments like that.