We present optical spectroscopy of galaxies in clusters detected through the Sunyaev-Zeldovich (SZ) effect with the South Pole Telescope (SPT). We report our own measurements of $61$ spectroscopic cluster redshifts, and $48$ velocity dispersions each
calculated with more than $15$ member galaxies. This catalog also includes $19$ dispersions of SPT-observed clusters previously reported in the literature. The majority of the clusters in this paper are SPT-discovered; of these, most have been previously reported in other SPT cluster catalogs, and five are reported here as SPT discoveries for the first time. By performing a resampling analysis of galaxy velocities, we find that unbiased velocity dispersions can be obtained from a relatively small number of member galaxies ($lesssim 30$), but with increased systematic scatter. We use this analysis to determine statistical confidence intervals that include the effect of membership selection. We fit scaling relations between the observed cluster velocity dispersions and mass estimates from SZ and X-ray observables. In both cases, the results are consistent with the scaling relation between velocity dispersion and mass expected from dark-matter simulations. We measure a $sim$30% log-normal scatter in dispersion at fixed mass, and a $sim$10% offset in the normalization of the dispersion-mass relation when compared to the expectation from simulations, which is within the expected level of systematic uncertainty.
We use three years of data from the Supernova Legacy Survey (SNLS) to study the general properties of core-collapse and type Ia supernovae. This is the first such study using the rolling search technique which guarantees well-sampled SNLS light curve
s and good efficiency for supernovae brighter than $i^primesim24$. Using host photometric redshifts, we measure the supernova absolute magnitude distribution down to luminosities $4.5 {rm mag}$ fainter than normal SNIa. Using spectroscopy and light-curve fitting to discriminate against SNIa, we find a sample of 117 core-collapse supernova candidates with redshifts $z<0.4$ (median redshift of 0.29) and measure their rate to be larger than the type Ia supernova rate by a factor $4.5pm0.8(stat.) pm0.6 (sys.)$. This corresponds to a core-collapse rate at $z=0.3$ of $[1.42pm 0.3(stat.) pm0.3(sys.)]times10^{-4}yr^{-1}(h_{70}^{-1}Mpc)^{-3}$.
