THE FOURIER-SPACE STATISTICS OF SEED-LIKE COSMOLOGICAL PERTURBATIONS

We propose a new test for distinguishing observationally between cosmo logical models based on seed-like primordial perturbations (such as co smic strings or textures) and models based on Gaussian fluctuations. W e investigate analytically the Fourier-space statistical properties of temperature or density fluctuation patterns generated by seed-like ob jects and compare these properties with those of Gaussian fluctuations generated during inflation. We show that the proposed statistical tes t can easily identify temperature fluctuations produced by a superposi tion of a small number of seeds per horizon scale for any observationa l angular resolution and any seed geometry. Due to the central limit t heorem, however, the distinction becomes more difficult as the number of seeds in the fluctuation pattern increases.