理解构造函数和析构函数的调用顺序

#include <fstream>

using namespace std;

ofstream out ("order.txt");
#define CLASS(ID) class ID{\
public: \
 ID(int) {out << #ID " constructor\n";}\
 ~ID() {out << #ID " destructor\n";}\
};

CLASS(base1);
CLASS(member1);
CLASS(member2);
CLASS(member3);
CLASS(member4);

class derived1 : public base1
{
 member1 m1;
 member2 m2;
public:
 derived1(int) : m2(1), m1(2), base1(3)
 {
  out << "derived1 constructor\n";
 }

 ~derived1()
 {
  out << "derived1 destructor\n";
 }
};

class derived2 : public derived1
{
 member3 m3;
 member4 m4;
public:
 derived2():m3(1),derived1(2),m4(3)
 {
  out << "derived2 constructor\n";
 }
 ~derived2()
 {
  out << "derived2 destructor\n";
 }
};


int _tmain(int argc, _TCHAR* argv[])
{
 derived2 d2;
 return 0;
}

order.txt的输出内容:
base1 constructor
member1 constructor
member2 constructor
derived1 constructor
member3 constructor
member4 constructor
derived2 constructor
derived2 destructor
member4 destructor
member3 destructor
derived1 destructor
member2 destructor
member1 destructor
base1 destructor