How to find references from publications in the physics literature? Coming from Computer Science, I'm a bit puzzled about the format of references in physics publications. They usually lack a title, which I'm usually using to locate a publication online (for obtaining pdf and bibtex entry).
A typical reference entry appears to be:


*

*Zhou H J and Wang C 2012 J. Stat. Phys. 148 513-547


What I'm asking for is an efficient method to locate the referenced paper online. This usually includes a pdf (e.g. preprint from arxiv.org) and a valid bibtex entry.
 A: Most physics publications nowadays (at least in my experience) include a link to the paper's DOI, which is the easiest way to get to the reference's abstract page. If I am reading a printed-out paper and care about the references, I'll usually have the references section in a browser window and use that to go to each reference.
Absent that, the best bet is usually to google (or duck-duck-go, or whatever) the paper title by itself. This tends to maximize the chance that you'll get a page with the paper title in big bold letters in a heading, and that is usually the paper abstract page itself. It is indeed fairly rare for journals to include paper titles in bibliographies, but the general trend seems to be in that direction. (For example, APS now allows and encourages the practice. This is something we should all be doing, really. Electronic ink is cheap!) The paper title is also usually the only way to locate freely-available eprints, either on the arXiv or on university repositories. If you want a free eprint and don't have the paper title, concentrate on finding the journal version first and then use that.
With a citation like the one you gave, it can be a bit of a hassle. Direct googling of the citation can yield only papers that cite the one you're interested in, in which case things do seem to be more troublesome than they ought to be. In these cases:


*

*Google scholar is generally better at locating references with such limited information. The advanced search (currently accessible via the drop-down triangle symbol at the right end of the search bar) allows for very fine-tuned searches, though it has its peculiarities in terms of what works best and what almost works but doesn't. (In particular, for authors, author:"C Wang", inside quote marks and using initials.)

*Google Scholar has a companion browser extension, the Google Scholar Button, available for Chrome and Firefox (and, reportedly, Safari). It's in general quite sturdy, but it is obviously limited by the capabilities of Google scholar itself. In my experience, it does a pretty good job of finding most normal references and pointing you directly to where you want to go.

*Paper Locator is a pretty new, open source webpage and browser extension that helps do exactly this. It's currently a bit buggy (when it works it's really nice, but sometimes it gets tripped up by trivial stuff) but it looks quite promising.

*Direct browsing of the journal website will generally do the trick, but it is a bit of a hassle. If you don't have an issue number, though, this can be a long process of guessing which link has the paper you want.

*NASA's ADS database is useful for locating papers. I also find it quite useful as an indicator of when to stop looking for an arXiv-hosted version (though it won't list institutional eprints).

*The standard academic search engines, including CiteSeerX, Web of Knowledge, ScienceDirect and Scopus can be quite effective. They tend to require institutional subscriptions, though, which makes them less useful.

*If you're looking for a book, WorldCat can be very effective in locating a copy in a library near you; in particular, it can find libraries that have the volumes you want and sort them by distance from any specified location, which is quite handy.
The short version, though, is that there's usually no silver bullet if you don't have the paper title. I'm leaving this open as a community wiki, though, so people can add in resources that also work. (If you do know a silver bullet, though, give it its own answer!)
A: The line

J. Stat. Phys. 148 513-547

contains all the information needed to locate the article. It appeared in the Journal for Statistical Physics, issue 148 on pages 513-517. In fact, if you just type the stuff in my quote into Google, the third hit (at least on my search) is a .pdf version of the paper.
A: The service authorea.com has a very good built in system to search for citations across many disciplines.
This is what your particular example returns:

If instead of using the full line presented by the OP with all the information on the article (ie: Zhou H J and Wang C 2012 J. Stat. Phys. 148 513-547) you give this service the smaller line: J. Stat. Phys. 148 513 (as presented in one of the comments above), the correct article is still recovered, this time in the third position.
I've personally used Authorea to write the most part of an article and I found that its citation resolving/inserting tool is probably the best thing about it. 
A: *

*Use Google Scholar search! It turns up the paper when searching for
2012 J. Stat. Phys. 148 513-547

or even
J. Stat. Phys. 148 513-547


*Barring a good academic search engine, a longer route is using something like Web Of Knowledge or Scopus, but those both need subscriptions which, in light of Google's excellent search engine, seems a bit daft.

*A third and final alternative is to search on the journal's web archive. Searching for J. Stat. Phys. turns up the Springer.com web page of the journal, from which you can search the archive ("All Volumes & Issues").
In your case the 148 is the volume, and 513-547 are the pages in that volume.
A: If you are interested in references you might also have a look at Scientillion. It is a search engine, mainly for physics publications and shows the references for many articles like Sample Article which saves you the time looking them up in the PDF.
