Thanks for the advice. I was hoping to add a video co-processor and make it a 16-bit computer.

Ah ok. Something like the TI video chip used in the MSX? https://en.wikipedia.org/wiki/Texas_Instruments_TMS9918   I won't say it's impossible to do on a breadboard but it'd certainly be a lot of wiring.  At 16Mhz clock its possible you'd have some issues but you'll probably be ok if you keep the wires short and have decoupling caps and reasonable power/gnd lines.  A good way to understand what you're getting yourself into is just look at the schematics of similar computers; e.g. the TI99/4A which used a different (but not wildly different; still 16-bit data bus) CPU; look at this PCB; http://www.mainbyte.com/ti99/hardware/big.html you could probably lose about 25% of those chips and still have something that worked.  What you're proposing is doable but fairly ambitious; ironically the fact that the chips are old makes it more complicated than it would be using modern stuff. 

oh also check out this https://www.ist-schlau.de/

also there's a 68k breadboarded project from 2014 on HAD; http://hackaday.com/2014/11/08/breadboarding-a-68000-computer-in-under-a-week/

You'll be wanting an EEPROM for code (well, two, because the 68k has a 16-bit data bus; you could use a 68008 instead) and some RAM (SRAM, again two) plus a couple of TTL chips for address decoding as the absolute bare minimum. Erasing and reprogamming two EEPROMs every time you want to change any of your code gets old real quick (and you need a UV eraser and a programmer). Are you really attached to using a 68k? It'd be much, much easier to use a modern microcontroller; no ram/eeprom required, simpler/quicker/cheaper to program and debug. You could use e.g. a dirt-cheap STM32F103 or something with an ST-Link clone; together will cost you about $8, or just get an ST discovery board. If you specifically want to run 68k code you can get an STM with enough ram and emulate it.  Another simple approach is to run a 68k emulator on (say) a Raspberry Pi or a C.H.I.P; you can hook your sound chips up to it with some simple level shifters (and may even be able to get away without those for the address bus).  What are your goals and what are your current skills; I guess that's what I'm asking.  If you want to drive those sound chips there's a hundred ways to do it and most of them are easier than getting a 68K working.   However, I understand sometimes a fella's just got a 68000-shaped itch to scratch, so if that's you, that's fine too.  LMK what you're looking to achieve and I'm sure I can give some input.

Note you'll also need to generate a clock for each of the sound chips - you could use 4Mhz (use a TTL divide-by-four chip on the 16Mhz crystal). If you used a modern microcontroller you could generate that internally.  It really depends what you're trying to achieve (and that could be specifically 'learning how to hook up a 68k to some retro sound chips').  Ironically if someone gave me your list of components and said "get this working, in a box, connected to some buttons and a speaker, as simply as possible", I'd reach for a Raspberry Pi and the MAME source code, which emulates all of them; could be done in a couple of hours. :-)

EEPROM is electrically erasable, you don't need UV for that. That's EPROM.

With EEPROM all you need is the programmer. The good news there is that there are modern ones for around $50 that also program most flash chips.

Even if you're using recycled EPROMs you can get an eraser for $15. If you're rapidly reprogramming it and the amount of time it takes seems inefficient, just slow down and spend more time checking the code for correctness before programming. Often that is actually quicker than guess-reprogram-check. I find it to be quicker than a emulator, too, but I'm a minority on that; I don't use a C debugger either.

In many cases you can just use a discrete flash chip too, if you're just reading they're not really that different. Most cheap printers have at least 4M of flash, most people don't even bother to grab it if they part out a printer.