A NOTE ON THE NEAREST-NEIGHBOR ANALYSIS

In the nearest neighbour method of the random distribution of N object s on a sphere surface the variable (N - 1)sin(2)(alpha/2) is proposed to be used instead of the angular distance alpha. Then both the expect ed mean and variance equal to one for large N, and both the Student's test and F-test are especially simple. An illustrative application sho ws that the superclusters - collected in the All-Sky Catalogues of Sup erclusters of Abell-ACO Clusters - need not be distributed randomly.