Lots of resources exist for salary issues that are much more data-thorough, so take the following numbers with a grain of salt. For more in-depth info, a few resources include HigherEdJobs, The Chronicle of Higher Education, or a web search.
I took a (somewhat) random single institution (Bowling Green State University, Ohio) of medium size (ca. 15,000 undergraduates) that also offers graduate degrees (in chemistry, an M.S. in chemistry and a Ph.D in photochemical sciences are available). Ohio was chosen as an example state simply because of the ready availability of data.
Salary information (for a few years back) is available for all Ohio's higher ed institutions through the Buckeye Institute. I grabbed the 2010 info for the university and present below the averages for four standard academic ranks (lecturer/instructor, assistant professor, associate professor, and full professor) across 14 broad but somewhat arbitrarily chosen disciplines. Standard deviations aren't included and sample sizes were small in some cases, so caveat emptor and all that. (NB: click any chart for a larger view)
First, a graph of the disciplines ranked by associate professor salaries. It's quite interesting to me that chemistry is near the top--ahead of biology but also physics and geology. Moreover, associate professor salaries in chemistry rank a little short of computer science (by about $7,000) but above economics (by about $8,000). As would probably be expected, two business disciplines (management and accounting/M&IS) are way ahead. I don't know if that's endemic to the particular school or a general trend. Regardless (and probably again to no one's surprise), it looks like chemistry and the other sciences are pretty far ahead of the humanities by as much as business-related fields are ahead of science.
Ranking by full professor puts chemistry more in the middle of the pack:
Why the giant divide? Market demand certainly plays a large role. Folks with graduate-level business and computer science skills are very, very employable, and generally aren't in markets plagued by the oversupply that science (and especially the humanities) face.
Lastly, check out the ranked salaries for instructors/lecturers. These are the teaching-only positions; for some disciplines this doesn't require a PhD. (For chemistry, I've seen very few lecturers without PhDs; many have postdoc or industrial experience).
Management here has the highest salary by far, but that's incidentally an n = 1 type scenario (there's only one lecturer in the management department, and they appear have an 'executive' position). Here the business gap disappears; average salaries range from $38,000 to $52,000. Interestingly, computer science ranks in at $61,000, which is probably indicative of its very high employability--you have to pay someone a lot to draw them away from an attractive industry job.
For fans of seeing-it-all, here's a ranked-by-associate graph including all four ranks.
Lastly, here's the average salaries for most of the public university chemistry departments in that particular state (Ohio) [note--data was not easily harvestable for Ohio University, Shawnee State, Central State, or Youngstown State].
This itself is somewhat interesting, as there's a wide distribution (assistant ranges from an average of $55,000 to $78,000; associate from $70,000 to $97,000; full professor from $101,000 to $131,000). Moreover, salary averages don't appear to correlate to institutional prestige (cf. Ohio State and University of Akron, for instance) nor to cost-of-living.
Again, take all this data with a grain of salt; I just think it's interesting stuff.