Most journals will ask you to write an abstract of your paper, normally about 100-200 words long, and put it at the beginning of the paper so that people can read through it and see whether they want to read the whole paper. Writing a good abstract can be as difficult, if not more so, than writing the paper. I've got all this brilliant and well-presented research! How am I going to sum it up in 200 words??
There are various techniques you can use to write an effective abstract. One of them is to go through your article and summarise each paragraph in one line. Then use all these lines to form the basis of your abstract (and if you find a paragraph that doesn't say anything that you want to include, perhaps you should think about removing it!). Don't use the lines word for word and just run them together, but use them to shape your abstract. This technique means that there is a logical chronology to your abstract, which matches your main paper, and ensures that your abstract doesn't suddenly contain new information not included in your article. Alternatively you could use the questions listed in our Structuring your Article section to form the basis of your abstract. Of course these are just two techniques, and you'll probably want to find your own.
One important thing to remember about an abstract is that it shouldn't be a 'teaser' like the blurb on the back of a book. If you find yourself writing something like 'What were the exciting results of my work? Read on to find out …' then you probably want to re-think! Your abstract should be a summary of your whole paper – what you did, why you did it, how you did it, what your results were and where your research is going from here.
You will probably also be asked to come up with some keywords if you are writing an academic article. These are words that people will be able to use on internet or database searches to find your article and they may be used to allocate reviewers to your paper. Again, the trick is to be concise and informative. Try to think of the sort of things people would search for in an index or if they were searching for your work on Google – for example, if they're interested in colloid stability then 'colloid' is going to be a better word than 'chemistry'.
Keywords (and abstracts and titles) are becoming more and more important as readers access papers more on-line than in hard copy. Keywords, titles and abstracts are the window to your research and must be accurate, interesting and relevant in order to ensure that (the right) people read your work. Make every word count and remember, this may be the only part of your research people read (particularly if you get it wrong!).