Grigorij Kuzmin was born on 08.04.1917 in Vyborg (a town in Finland at that time, now in Leningrad district). In 1924 the family moved to Tallinn, Estonia. In 1935 he graduated the Russian High School in Tallinn and became a student of the Faculty of Mathematics and Natural Sciences of the Tartu University, which he graduated

*cum laude* in 1940.

Influenced by Professors Ernst Opik and Taavet Rootsmae, founders of modern Estonian astronomy, Kuzmin became an astronomer. Opik considered him as his most talented disciple. His first scientific papers were published in 1938. Since 1940 he worked in Tartu Observatory, first as an assistant fellow, then as senior follow and head of the division of galactic studies. He lectured in Tartu University, teaching astronomy, geophysics and stellar dynamics. Kuzmin was the supervisor of Academician Einasto, Professor Kutuzov, Professor Malasidze and many Tartu astronomers.

Works by Ogorodnikov, Agekian, Idlis, Genkin, Tsitsin, Ossipkov and others were influenced by Kuzmin.

In 1961 Kuzmin was elected Corresponding Member of Estonian Academy of Sciences. In 1971 he was awarded by Bredikhin Award of Academy of Sciences of the USSR.

Kuzmin passed away on 22.04.1988 after a hard kidney disease.

Kuzmin was interested in many branches of astronomy, but his most prominent achievements lie in Galactic Astronomy and Galactic Dynamics. Kuzmin was the second (after Ogorodnikov)researcher in Soviet Union who started systematical studies in Dynamics of Stellar Systems.

In 1943 he developed a method how to construct galactic models using ellipsoids of non-homogeneous density and varying axial ratio. He applied this method first to the Andromeda galaxy (1943) and then to our Galaxy (1952).

An important parameter of the Galactic model is the local density in Solar vicinity. He developed a method how to derive this quantity using vertical motions and positions of stars of flat Galactic populations. His results show that there is no evidence for the presence of large amounts of dark matter in the plane of the Galaxy, contrary to results obtained by Oort (1933) and Parenago (1951). His students Eelsalu and Joeveer repeated this analysis using different data and methods, and confirmed Kuzmin results. This discrepancy was solved only recently: modern data have confirmed results by Kuzmin and his students.

The triaxial shape of the velocity ellipsoid shows that there must be at least 3 integrals of motions, and Kuzmin developed the theory of the third integral of motions (1952). Kuzmin showed that only single-valued (isolating) integrals of motion should be used as arguments of a steady distribution function for stellar systems and a number of such integrals is equal or less than three, in general (1953).

He found that Stackel type potentials can be used for constructing reasonable Galactic models, and developed a general theory of axisymmetric Stackel type gravitating systems (1956).

Kuzmin was the first to construct Stackel type triaxial gravitating models (1973).

To study effects of stellar encounters Kuzmin deduced a Fokker-Planck type kinetic equation for gravitating systems (1957), practically at the same time as Rosenbluth and other physicists for plasmas.

Applying his theory to galactic discs containing massive perturbers (such as giant molecular clouds) Kuzmin found a quasi-steady relation between velocity dispersions that is confirmed by observational data (1962).

Kuzmin and his disciples Kutuzov and Veltmann were among the first to construct steady phase-space models for stellar systems. Kuzmin and Kutuzov found a two-integral distribution function for a reasonable mass distribution model (1962). Veltmann developed in 1960--1968 a general theory of steady self-gravitating spherical systems, and then Kuzmin and Veltmann discovered new classes of spherical models that were compared with observational data for globular clusters.

Kuzmin and co-authors (1986) discovered very general family of models of mass dustribution in stellar systems that can be used for approximating the de Vaucouleurs' profile for a surface density and include the Hernquist model and the gamma-model found later.

Kuzmin shown that Lichtenstein's theorem on the existence of a plane of symmetry is valid for wide classes of collisionless gravitating systems (1964).

Kuzmin's articles include many other results of fundamental significance.

Unfortunately Kuzmin did not like to publish his works, and many of his results are unpublished till now. But even papers he published are sources of new studies.