Think of integers as a whole string of numbers but with no decimal point. Let's say, for argument, you can have 10 digits.
Now think of having only 9 digits but you can have a decimal point at any position (floating). The other digit says where the decimal point is placed. It's much like exponential notation but oriented to computer storage.
Note that this is just for visiualizing the concept. The actual method of storing floating point numbers is much more involved.