A floating-point number. If length is between 1 and 24 inclusive, the format is that of a single precision floating-point. If length is between 25 and 53 inclusive, the format is that of a double precision floating-point. If Length is not specified, the default is 53.