William Gilbert formulated over 400 years ago a postulate that can be considered as the main principle of modern natural sciences : All theoretical constructs that claim to be scientific must be verified and confirmed experimentally.
Despite of past centuries, this principle has not lost its relevance today. In the modern physics there are some conventional theories, which do not satisfy to Gilbert’s postulate . In physics of microcosm there are models which cannot be compared with the measurement data as they do not allow to calculate the basic characteristic parameters (such as masses or magnetic moments) of elementary particles. In this article an alternative approach to these problems is considered. It is shown that an attraction in the proton-neutron pair can occur due to the exchange of relativistic electron. The estimation of this exchange energy is in agreement with the experimental values of the binding energy of some light nuclei. At that neutron is regarded as a composite corpuscule consisting of proton and relativistic electron that allows predicting the neutron magnetic moment, its mass and energy of its decay. It is shown that the standard Maxwell’s theory of electromagnetic field describes a possibility to initiate in free space (in empty ether) a magnetic ϒ
-quantum (a splash of magnetic field), devoid of the electric component and having spin
. Since magnetic monopoles do not exist, a characteristic feature of the magnetic ϒ
-quantum is the weakness of its interaction with matter, which is many orders of magnitude smaller than that of the electromagnetic wave. These properties suggest that the magnetic ϒ
-quantum can be identified with neutrinos. On this basis, we get a fresh look on the nature of π
-me-sons and μ
-mesons and calculate their masses.