Float Questions

I have 2 questions about floating point numbers:
1) What is the largest value you can have using 32 bit?
2) How are the number of decimal points determined?

These questions are in relation to Blitz3D and the way it handles them.

1) 340282346600000000000000000000000000000.0
2) floats use a variable number of decimal points

Blitz3D prints floats using the 7 most significant digits which refelcts the accuracy of a 32 bit float (you can add 1.0 successfully to 10,000,000.0 but you can not add 1.0 to 100,000,000.0 as it is smaller than 1/2^23 of the number you are adding to).