DaneClark - I was just going back over this thread, and it struck me that you might be wondering why in the world I suggested a different approach than Sam's (ie, to finding the edges) when his method works perfectly well. Unfortunately, if the image was slightly different, the "color range" tool might not work as well, so knowing several different approaches to a given problem is always worthwhile.
For example, what if the cartoon you are working on had either no black outlines at all, outlines in a color other than black, or weak or discontinuous outlines so that one color abutted the adjacent color? What if there were other very dark colors in the image besides the outlines, and/or the image you are working on is actually a scan of a comic printed by an offset press consisting of vast numbers of tiny dots of different colors, and to make matters worse, the printing might be faded and there might easily be strong paper texture in the scan? In such cases (and many others), the method of selecting blacks with the color range would likely not work as well.
In addition, selecting objects within an image is, IMHO, one of the most important skills that a Photoshop artist can have, and since there are so many ways to do this, beginning to learn some of these alternative methods is very worthwhile.
Finally, it also occurred to me that it may not be clear to you what the relation is between the white-on-black images that I presented and your specific problem. The relevance is that once the edges have been highlighted in the way that I showed, other selection tools (that are often prone to errors / less powerful) suddenly become much more efffective. For example, suppose you wanted to select one specific area for coloring. Then, you could likely use either the "quick mask" or "magic wand" tool. Click in that area on my B&W image with one of these tools, and the selection would likely instantly run right up to the white border and stop there. Save that selection and make a layer mask out of it. Repeat for each area of interest.
Anyway, I hope this at least partially clarifies why I presented this alternative approach.
Cheers,
Tom M