Literal decimal floating-point numbers can not cannot always be represented precisely using the double primitive type because the underlying representation of double is binary. This imprecision becomes apparent when a BigDecimal is constructed from a double.
...