Lag Time in Calling Methods?

Question

I ran the two code examples below and got the same elapsed time for both: 10 ms. I was wondering whether calling a method causes lag: for instance, if you had to add four components to a frame's content pane, would calling getContentPane() four times be less efficient than setting up a reference to the content pane container and adding the four components to the container that that is returned?
The reason why I think that there would be a lag time, however infinitessimal, is because calling a method pushes an EIP onto the stack and returning from a method pops the EIP from the stack. It only makes sense that skipping this step would result in faster execution (once again, however infinitessimally faster that would be).
Here is the code for the two programs I used to test this out. The examples go back to my example above -- I am setting up a reference to the content pane in one example and continually returning it in the other. As I said, both returned the same elapsed time: 10 ms. I'm not sure whether my method of timing the programs was inaccurate or whether there really is no lag time in calling a method. What is the verdict?
1)
package cntrl.testing;
import java.util.Date;
import javax.swing.JFrame;
import javax.swing.JButton;
import java.awt.Container;
public class Test1 extends JFrame
     public Test1()
          long i = new Date().getTime();
          this.setSize(500, 500);
          this.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
          Container c = this.getContentPane();
          c.add(new JButton());
          c.add(new JButton());
          c.add(new JButton());
          c.add(new JButton());
          long f = new Date().getTime();
          long te = f - i;
          System.out.println("Time elapsed: " + te + " ms");
     public static void main(String[] args)
          new Test1();
}2)
package cntrl.testing;
import java.util.Date;
import javax.swing.JFrame;
import javax.swing.JButton;
public class Test2 extends JFrame
     public Test2()
          long i = new Date().getTime();
          this.setSize(500, 500);
          this.setDefaultCloseOperation(JFrame.EXIT_ON_CLOSE);
          this.getContentPane().add(new JButton());
          this.getContentPane().add(new JButton());
          this.getContentPane().add(new JButton());
          this.getContentPane().add(new JButton());
          long f = new Date().getTime();
          long te = f - i;
          System.out.println("Time elapsed: " + te + " ms");
     public static void main(String[] args)
          new Test2();
}

Answer 1

You're probably using some operating system where the smallest time interval you can measure is 10 milliseconds.
I would suggest that pushing something onto the stack would take an interval measured in nanoseconds. You have, let's say, a 1 gigahertz machine? That means it performs one billion operations per second, give or take. That's one per nanosecond.
There are a million nanoseconds in a millisecond. You're going to have to run a LOT of code to take a measurable amount of time.