Three types of always absolutely convergent series are presented for the cdf of convolutions of not identical p-variate scaled gamma distributions (in the sense of Krishnamoorthy and Parthasarathy). Actual computations are comparatively easy, at least for  Moreover, three general integral representations are given. As a corollary, an integral over  is provided for the convolution of univariate differently scaled gamma distributions. For Jensen’s  p-variate gamma distribution, depending on different one-factorial correlation matrices  a further integral representation is derived, which is useful at least for  The convergence of Jensen’s general p-variate series for the cdf is investigated. Finally, the maximal component of Jensen’s bivariate gamma with the associated correlations  is shown to be stochastically less than the maximal component of a bivariate gamma distribution with ndegrees of freedom and a correlation given by the mean square of

multivariate chi-square distribution, convolutions of multivariate gamma distributions, Jensen’s multivariate gamma distribution.