In 1750, Michell studied artificial magnets and developed a method for magnetization. As a geologist, he constructed a torsion balance for measuring gravitational forces; this device became famous only by its second inventor, Charles-Augustin de Coulomb (1736-1806), and was eventually used by Henry Cavendish (1731-1810), who used the device of his life-long friend, John Michell, in his famous experiment to measure gravity between two test masses; from his measures, Cavendish was able to give a very acurate estimate of the Earth's mass (6*10^18 metric tons). In 1760, Michell constructed a theory of earthquakes as wave motions in the interior of the Earth, and suspected a connection between earthquakes and volcanism.
In 1767 he published an investigation on double stars and clusters, and calculated the probability of finding chance alignments of stars (asterisms). In particular, he investigated the Pleiades cluster, and calculated a probability of 1/496,000 to find such a group as a chance alignment anywhere in the sky. Also, he found that much too many pairs and close groups of stars were visible in the sky to assume that all these were chance alignments. and concluded that many of them should be physical pairs or groups, held together by
"the influence of some general law [..], to whatever cause this may be owing, whether to their mutual gravitation, or some other law or appointment of the Creator."
Later, in a letter to Cavendish of 1784, Michell published thoughts about the effect of gravity on light, including an early concept of Black Holes, masses dense enough to prevent light from escaping.
As an amateur astronomer, John Michel was also an active telescope builder. His main instrument was a self-made 10-foot [focal length] reflector of 30-inch aperture. This instrument was puchased by William Herschel after Michell's death, and served as a model for a similar instrument. At that time it was no more usable, as the prime mirror was damaged.
The work of Michell had some impact on William Herschel, and in particular, stimulated his work on double stars.