CI is (meanJ - meanS) +/- t-critical * sqrt( sdJ^2/n1 + sdS^2/n2)
where means and standard deviations (sd) are computed from sample data
n1 and n2 are the sample sizes both equal to 8.
t-critical is the t-score for 0.025 tail alpha and 7 DOF = 2.365
If the CI does not contain zero, then you can claim that there is a significant difference between he two samples at alpha = 0.05.