Why pandas group by call a function for a group twice?

Question

a = pd.DataFrame({"a":[1,1,1,1,1,2,2,2,2,2], "b":range(10)})
a.groupby('a').apply(lambda df: print("***\n" + str(df) + "\n***\n"))

I am expecting the lambda function be called exactly once for every group, but it has been call twice for the first group. Any one know why this happen?

***
   a  b
0  1  0
1  1  1
2  1  2
3  1  3
4  1  4
***
***
   a  b
0  1  0
1  1  1
2  1  2
3  1  3
4  1  4
***
***
   a  b
5  2  5
6  2  6
7  2  7
8  2  8
9  2  9
***