The .js also looks at the captcha, just to make sure that something, anything was written in before the page is allowed to be submitted. The second is where the captcha input is actually validated ...

When a web application using reCAPTCHA challenges a visitor, Google provides an image set through JavaScript code. When a visitor solves the image-based puzzle, an HTTP request is triggered.

Those online tests where you click boxes to prove you're not a robot are going away. A new test is here, it's always running in the background, trying to figure out if you're really a human being.