If you had not heated the sample long enough to remove all the water of hydration, how would your subsequent calculations have been affected?