Probability of Spurious Correlation

If we have two variables and wish to describe the co-relation between them, it is commonplace to use the correlation coefficient, R. The correlation coefficient is a measure of the linear relation between two variables. However, for a given correlation coefficient, how do we make some estimate of the validity of R in expressing the co-relation? For example, if we have a correlation of 0.9 is that a good or a bad correlation? How much confidence can we place on the correlation? If we compare two variables for which we have just a few samples, say 10, and observe a correlation of 0.8, is that better then a correlation between two variables, of which we have 20 samples, when the correlation is 0.7? To answer this question we can use a two-tailed Student T test, a standard statistical test which can be done in a spreadsheet.