Easy, floating point numbers are an approximation, approximations can change when operated on.
You have assigned the double constant (9.9)to a float variable so it has converted that double to float.
Then you compare the float variable to a double constant (9.9) so the compiler converts the variable from a float to a double.
So note that in the first 2 lines of your code the constant double 9.9 is implicitly cast down to a float and the implicitly cast back up to a double. With-in those casts it is perfectly with-in the relms of possibility that 9.9 becaomes 9.89999999999999 or less so true gets printed.
To avoid it either define a as a double or use floating point constants 9.9f to avoid implicit casts.
Try reading
this too;
