Under CMS, floating-point representation (unlike scientific notation ... in order for any floating-point number to be represented exactly, you must be able to express it as the previously mentioned ...