In the same table, if Levene's test is not significant (p >= 0.05), read the first line of the t-test portion. If the Levene's test is significant (p < 0.05), read the second line of the t-test portion.
The two lines differs in how the variance is estimated. If the two distributions are the same, then the variance is pooled, if the two are different, then the variance would be a combination of two sets of variance. For the formula, you can see here:
http://www.okstate.edu/ag/agedcm4h/acad ... page26.htm