字符的utf16在转换为字节时有2个负数

import java.io.UnsupportedEncodingException;
import java.util.Arrays;

public class Main {
 public static void main(String[] args)
 {
  try 
  {
   String s = "s";
   System.out.println( Arrays.toString( s.getBytes("utf8") ) );
   System.out.println( Arrays.toString( s.getBytes("utf16") ) );
   System.out.println( Arrays.toString( s.getBytes("utf32") ) );
  }  
  catch (UnsupportedEncodingException e) 
  {
   e.printStackTrace();
  }
 }
}

慰问:

[115]
[-2, -1, 0, 115]
[0, 0, 0, 115]

这是怎么一回事?

[-2, -1] - ???

另外,我注意到,如果我这样做:

String s = new String(new char[]{'\u1251'});
System.out.println( Arrays.toString( s.getBytes("utf8") ) );
System.out.println( Arrays.toString( s.getBytes("utf16") ) );
System.out.println( Arrays.toString( s.getBytes("utf32") ) );

慰问:

[-31, -119, -111]
[-2, -1, 18, 81]
[0, 0, 18, 81]