Question 1190564
Let the first sample be denoted by random variable {{{X_1}}} and the second sample by {{{X_2}}}. 
First few notes about {{{X_1}}}and {{{X_2}}}. 
---1 They are indepdent RVs. Why? Because sampling happens with replacement. Which means, if I sample {{{X_1}}}, then no matter what{{{X_1}}} really was, the distribution of {{{X_2}}} remains the same.

Now consider the random variable {{{X_3=(X_1+X_2)/2}}}. Do you see why {{{X_3}}} is the sample mean?

Now we can simply apply formulas and theorems to calculate each of the quantities that you have asked.
First, what is mean of the sample-mean? Easy. As you may know, "mean" of a random variable is nothing but an "expectation" operator "E" on the random variable. Therefore mean of the sample means would be {{{E(X_3)}}} which is {{{E((X_1+X_2)/2)}}}. Remember {{{E}}} is a linear operator? Which gives us {{{E(X_3)=(E(X_1)+E(X_2))/2}}}. What is {{{E(X_1)}}}, or mean of the population? It is {{{(3+8+10+15)/4}}} which is {{{36/4=9}}}. What is {{{E(X_2)}}}? Well that is also 9 (Remember we are sampling with replacement?)