I would say yes, but I collect slide rules.
Slide Rule Universe is a great site to research slide rules, but they are expensive. A new, in-the-box slide rule is a great collector's item, but if you use it, the value drops immediately. They also charge a fairly high price for used slide rules.
E-bay is the better option. Rod Lovett has been tracking slide rule prices on eBay for several years. You can get a feel for what a reasonable price is for different slide rules by checking his site.
http://sliderules.lovett.com/srsearch.html
For example, Slide Rule Universe charges over $200 for a Post Versalog in near mint condition. On e-Bay, you should pay less than $100, and only over $50 for very good condition with the original leather case. The only difference is you'll probably have to do at least a little cleaning and restoration work (you can find info on how to clean and restore slide rules at Slide Rule Universe).
The basics of using a slide rule is very easy. A Reitz or Mannheim style is pretty easy to learn (with Mannheims being more common in the US), plus it's good to learn the basics before getting the more exotic scales. You'll have a A and D scale on the slide rule body and a B and C scale on one side of the scale. You pull the slide out and turn it over to get trig functions and base 10 log scale. Believe it or not, this is all you need provided you're good with logarithms and are comfortable working with complex numbers.
The more fun slide rules are the duplex rules that start adding special functions, such as a natural logarithm scale that's usually broken up into around 8 segments (for instance, the Post Versalog advertises 8 log log scales, but it's really one long scale broken up into 8 segments to give you a good range. They start including reversed CI scales and scales that run from pi to pi instead of .1 to 1, and so on. The idea being to provide enough flexibility that you can set your problem up to minimize slide rule movements and minimize the number of times you have to read a value from your slide rule (you only have 3 significant digits, so you lose a little accuracy every time you transpose a value from your slide rule onto paper). While the basics are very easy to learn, learning to set the problem up to get the quickest and most accurate answer is tougher. The extra scales do give some extra capabilities, but not many. For example, you can solve quadratic equations on a Post Versalog very quickly using kind of a visual iterative method, but you could also do the same thing just a little slower using a Mannheim or Reitz and the quadratic equation.
You could also solve equations like e^x = x^4 for x, and probably better than a person could using a graphing calculator. Using and understanding your calculator takes as much time and effort as using a slide rule, but most don't realize that. Very few calculator owners could solve something like e^x = x^4 correctly, in spite of the fact that it's so easy it's pathetic. I knew one person who never realized just about all of the constants he'd ever used in college were stored in his calculator until his very last semester before graduating. Learning to use a slide rule well is easier than learning how to use your calculator well.
Generally speaking, a good graphing calculator is a little better than a slide rule, but a good duplex slide rule with 20+ scales is in the same ballpark. A Post Versalog is immensely more capable than your standard scientific calculators like the TI-35.
Of course, the extra scales have their own hazard. You start to use the log log scales so much, you get stuck if you have to deal with a number larger than the highest value on your log log scales. You almost forget you could do any size number using your base 10 logarithmic scale and the fact that the natural log of a number is just 2.30 times the base 10 log (which is why it's good to use the basic Mannheim or Reitz style first instead of jumping right into the fancy duplex rules).
), and will be doubling up on mathematics courses relevant to a physics degree.
