Especially in academic writing, it is frequently necessary to refer in your text to other work of which you have made use or to which you want to direct the reader's attention. There are several different systems for doing this, and they are not all equally good.
By far the best system is the Harvard system, also called the author date system, and this is the one I recommend. In the Harvard system, you provide a reference in the form of the author's surname and the year of publication; this is enough to direct the reader to the list of full references in your bibliography. Like any brief interruption, the date is enclosed in parentheses, and the surname goes there too, unless it is a structural part of the sentence. Multiple references are separated by commas. Where necessary, a few words of explanation may also be placed inside the parentheses. Here are some examples:
Very occasionally you may need to cite something which somebody else has told you personally, either in conversation or in a personal letter. You do it like this:
A second widely-used system is the number system, which is particularly popular in some scientific circles. Here a reference takes the form merely of a number enclosed in square brackets:
There are several other ways of citing references, but they are all highly objectionable and should never be used. A few writers put complete references into the body of the text, which is both distracting to the reader and absurdly inefficient, especially when the same work is cited several times. Very many writers have the bad habit of putting references into footnotes and flagging them just like ordinary footnotes; not only does this practice clutter the page with pointless footnotes, but it wastes the reader's time by constantly sending her off to consult "footnotes" which are nothing but references. Do not use footnotes for references.
Worst of all is the dreadful hotchpotch used by many scholars in arts subjects, in which references are presented sometimes in footnotes and sometimes in the text and are almost always incomplete and full of cryptic abbreviations which the reader has no hope of deciphering. If you spatter your work with unexplained exotica like DCELC, REW 1317, Schuch. Prim., Urquijo BSP IV, 137 ff., and so on, then no doubt the other eighteen specialists in your field will follow you, all right, but the rest of your readers will be helpless. Do not provide incomplete references, and do not use unexplained abbreviations. If you find that the use of some abbreviation is unavoidable, then explain it clearly, either the first time you use it, or, better still, in a list of abbreviations at the beginning of your work.
The perpetrators of such inexcusable obscurity have the further outrageous habit of citing references with the Latin abbreviations ibid. and op. cit. What do these mean? Well, ibid. means "This is another reference to the last thing I cited; it's back there somewhere, maybe only a page or two, if you're lucky." And op. cit. means "This is another reference to the work by this author which I cited some time ago, and, if you want to know what it is, you can leaf back through twenty-five or fifty pages to find it, you miserable peasant." (Technically, they mean `in the same place' and `in the work cited', but my explanations are far more honest.) Don't use these ghastly things. A writer who uses them is expressing utter contempt for the reader, and should be turned over to the Imperial Chinese Torturer for corrective treatment.
Use the Harvard system. It's vastly superior to everything else.