A cosmologia é um domínio da física no qual o emprego da teoria da relatividade geral é indispensável. No entanto, uma cosmologia baseada na teoria newtoniana da gravitação é possível em certas circunstâncias. O campo de aplicabilidade da teoria newtoniana pode ser substancialmente estendido caso ela seja modificada de forma a dar um papel mais ativo à pressão como fonte do campo gravitacional. Isto foi feito no âmbito da teoria neo-newtoniana. As limitações à construção de uma cosmologia newtoniana, e a consequente necessidade de uma teoria relativista em cosmologia, são revistas. A proposta de uma teoria neo-newtoniana é apresentada, e suas consequências para a cosmologia são abordadas.

Cosmology is a field of physics in which the use of general relativity theory is indispensable. However, a cosmology based on newtonian theory for gravity is possible in certain circumstances. The applicability of newtonian theory can be substantially extended if it is modified in such way that pressure has a more active role as source of the gravitational field. This was done in the neo-newtonian cosmology. The limitation on the construction of a newtonian cosmology, and the need for a relativistic theory in cosmology are reviewed. The neo-newtonian proposal is presented, and its consequences for cosmology are discussed.

Large values for the mass-to-light ratio (Ò) in self-gravitating systems is one of the most important evidences of dark matter. We propose a expression for the mass-to-light ratio in spherical systems using MOND. Results for the COMA cluster reveal that a modification of the gravity, as proposed by MOND, can reduce significantly this value.

The synthesis of helium in the early Universe depends on many input parameters, including the value of the gravitational coupling during the period when the nucleosynthesis takes place. We compute the primordial abundance of helium as function of the gravitational coupling, using a semi-analytical method, in order to track the influence of G in the primordial nucleosynthesis. To be specific, we construct a cosmological model with varying G, using the Brans-Dicke theory. The greater the value of G at nucleosynthesis period, the greater the predicted abundance of helium. Using the observational data for the abundance of primordial helium, constraints for the time variation of G are established.