This paper proposes a model of evolution of networks of cooperation where networks form and evolve through repeated pair-wise interactions between players in a finite population. While the baseline incentive structure of these interactions is that of a coordination game, these incentives are transformed by additional network effects based on the assumption that the benefit to any player from a mutually cooperative relationship is enhanced, over its base value, by a factor of their trustworthiness as perceived by the interacting partner, through network-based channels. In the presence of a set of altruists, the evolution of networks of cooperation is analyzed through repeated interactions between myopically best-responding players who have varied costs of cooperation.