Most universities will not allow X to get a second bachelor's. Fortunately for X, he can get into a grad program even if his first bachelor's is not in physics (although something sciency is preferable, and not a B.A. in underwater basket weaving). Unfortunately for X, grad programs are fairly selective and it's difficult to get in. There's also a significant financial factor in all this. Even if he were admitted, X would have to trade in his Porsche and spend 4 to 8 years in the university on starvation wages, and end up getting a job that pays substantially less than law practice (if he's lucky: most Ph.D.s end up working outside academia, we just had a thread about this).
A reasonable route would sound like this.
- Get a bunch of undergraduate physics textbooks. Repeat everything that X studied back when he was an undergrad student getting his bachelor's, learn things that weren't part of his curriculum. Work on solving problems.
- Ace GRE general and GRE subject physics.
- Apply for Ph.D. program at a local university. Offer to pay his way through while working part time. If X has connections in the physics department, that's a big bonus, especially if connections are at faculty or administrative level.
- Spend the first 1-2 years taking prerequisite courses while working and saving up money.
- Find an advisor ASAP. That's a big potential pitfall for someone who pays his way through. Most students get advisors right away because they need financing. If X pays for his own education, he runs the risk of getting stranded with no advisor and nothing to do once his prerequisites are complete. Should that happen, there's a fair chance that X will drop out. The advisor should be sufficiently interested in the student to take over the tuition check, because his funding is limited. That's the biggest obstacle. Getting in and paying out of his own pocket let's X establish relationships with professors and demonstrate his capabilities. That is assuming he is in fact capable of research work. He'll capitalize on that by getting an advisor.
- Once X is done with classes, he'll spend increasing amounts of time at the university. If he's in theory, it may be possible to go on working part-time for a while, though that is unadvisable. An experimentalist would have to be present at the university full-time.
If X were to start doing all of the above now, he might still have time to get into a graduate program starting fall of 2010, assuming his knowledge of physics is fresh and complete ... otherwise, it's best to start studying and apply next winter.