#128 was fixed by 245eb88 in 2012, but then 4218a07 reverted it with no explanation given. So I'm reopening it again.

Old bug description:

These two methods are using 0xFF as the bit mask when they should be using 0xFFFF. Each method returns a WORD, which is a 16-bit value. By using 0xFF only the lower byte is preserved.

New Code:

public WORD getLow() {
    return new WORD(longValue() & 0xFFFF);
}

public WORD getHigh() {
    return new WORD((longValue() >> 16) & 0xFFFF);
}

Old Code:

public WORD getLow() {
    return new WORD(longValue() & 0xFF);
}

public WORD getHigh() {
    return new WORD((longValue() >> 16) & 0xFF);
}