FLOAT(b) & INTEGER
   FLOAT(b) specifies a floating-point number
with binary precision b. The precision b can
range from 1 to 126.
    INTEGER :== NUMBER(38)