Using numerical simulations, we investigate the large-scale gravitational clustering in a flat universe dominated by cold plus hot dark matter (i.e. OMEGA0 = OMEGA(CDM) + OMEGA(HDM) + OMEGA(bayron) = 1). Primordial density fluctuation spectrum is taken to have the Zel'dovich-Harrison form. Three models are studied, with Model I having OMEGA(CDM) = 0.69, OMEGA(baryon) = 0.01, and OMEGA(HDM) = 0.30 in one flavor of neutrinos; Model II having OMEGA(CDM) = 0.60, OMEGA(baryon) = 0.10, and OMEGA(HDM) = 0.30 in one flavor of neutrinos; Model III having OMEGA(HDM) = 0.69, OMEGA(bayron) = 0.01, and OMEGA(HDM) = 0.30 in three flavors of neutrinos. The initial density spectra are normalized by the COBE quadrupole measurement, and galaxies are identified from the peaks of initial density fields above a certain threshold chosen, to match the observed two-point correlation on scales less than or similar to 10h-1 Mpc. Thus the clustering properties of both the mass and the galaxies are completely specified. The biasing parameter (for the ''galaxies'') determined in this way is bg almost-equal-to 1.2 for Model 1, 1.5 for Model II and 1.6 for Model III. The clustering and motions of the simulated ''galaxies'' are compared with recent observations. The spatial distributions of galaxies in the hybrid models are very frothy; filaments, sheets, voids etc. of sizes 10-50h-1 Mpc are frequently seen in the simulations. All three models are in good agreement with the observed local bulk motions and with the count-in-cell statistics sigma2(l) in redshift surveys of IRAS galaxies. One exception is the sigma2(l) of the QDOT survey at l = 40h-1 Mpc: the value is too high to expect in the models. But its statistical significance was recently questioned with an analysis of the 1.2 Jy IRAS survey. Model I does not have sufficient large-scale power to explain the two-point angular correlation function of the APM survey, the two-point correlation function of Abell clusters. Furthermore, its galaxy pairwise velocity dispersion around 1h-1Mpc is too high to reconcile with the observation. The other two models can be adjusted, within the observational errors, to fit all observations on scales from approximately 1h-1Mpc to approximately 50h-1Mpc, showing that the power spectrum of the initial density fluctuation is close to the predictions of these two models, and indicating that the observational results of galaxy clustering and motions on large scales are consistent under some reasonable theoretical assumptions.
