A researcher is comparing the mean income of two different cities. The sample sizes are $n_1=25$ and $n_2=25$. The variances are found to be significantly different. Which test is most appropriate for assessing $H_0: \mu_1 = \mu_2$?