This paper proves at first that any two non correlative random variables are still non correlative through linear transformation and the independent random variables of normal distribution are still independent each other through this kind of transformation.