The FDA is comparing the mean caffeine contents of two brands of cola. Independent random samples of 6-oz. cans of each brand were selected and the caffeine content of each can determined. The study provided the following summary information.

Brand A

Sample Size= 15

Mean= 18

Variance= 1.2

Brand B

Sample Size= 10

Mean= 20

Variance= 1.5

How many cans of each soda would need to be sampled in order to estimate the difference in the mean caffeine content to within .10 with 90% reliability?