A solar eclipse (SE) pertains to rare high-energy natural phenomena. For instance, a change in the internal (thermal) energy of the air in a layer of only a 100-m height attains 10¹⁸ J while the power of the process is of order of TW. The energetics of the processes produced by the SE in the upper atmosphere and geospace is significant. For instance, the thermal energy of the ionospheric plasma in a volume of ~10¹⁹ m³ decreases by 10¹¹ J, the magnetic field in a volume of ~10²¹ m³ decreases by 50 nT, and its energy by 10¹⁵ J. SEs are accompanied by disturbances in all subsystems of the Earth — atmosphere — ionosphere — magnetosphere system. Disturbances in the upper atmospheric and ionospheric parameters act to inevitably produce geomagnetic field variations. At present, geophysicists have no consensus on how SE manifests itself in the geomagnetic field. The available data are inconsistent. Most of the researchers believe that the geomagnetic effect of SE exists. In some cases, the temporal variations in the geomagnetic field, as a whole, repeat the changes in the illumination of the Earth’s surface, in other cases they occur ~1 hour in advance or with time delays of ~1 hour relative to the changes in illumination. Most often, the geomagnetic effect is studied in the region of the total SE where it should be the most pronounced. The further the observatory is located from the umbra, the more difficult the magnetic variations are to be relate to the SE. Finding the response of the geomagnetic field to the SE is a formidable task. A possible response is masked by variations of another nature. Moreover, the magnitude and sign of the geomagnetic field disturbance significantly depend on the state of space weather, season, local time, location of the magnetic observatory, and, of course, on the magnitude of the eclipse. Therefore, the study of the effect of SEs on the geomagnetic field remains an urgent task. The purpose of this work is to present the results of analysis of temporal variations in the geomagnetic field observed by the International Real-Time Magnetic Observatory Network (INTERMAGNET) during the SE of 10 July 2021. The main feature of this eclipse was that the SE was annular (maximum magnitude M_max ~ 0.943). The annular SE occurred on June 10, 2021 with a commencement time 08:12:20 UT over Canada. The Moon’s shadow moved across the Atlantic Ocean, the Greenland, the Arctic Ocean, the North Pole, the Northern Parts of Europe and Asia. The partial SE occurred in Mongolia and the People’s Republic of China, and it ceased at 11:33:43 UT. The annularity was observed to occur from 10:33:16 to 10:36:56 UT over the Greenland. The analysis of the geomagnetic effect was based on the INTERMAGNET database. Data were processed with 1-min temporal resolution, and with 0.1-nT level resolution, and temporal variations in the X-, Y-, and Z-components recorded at 15 magnetic observatories have been studied. The SE was determined to be associated with an aperiodic decrease in the X component by 31…36 to 2…3 nT. A decrease in the effect attained a maximum value during the maximum magnitude of the annular eclipse, and the magnitude of the effect rapidly decreased with distance southward. For the most southern observatories, we have not managed to determine a decrease in the mean value. Other geomagnetic field components virtually did not change in the course of the SE. The SE was also accompanied by quasi-periodic variations in the level of the X component. The amplitude of these variations decreased from 4.8 to 0.4 nT with distance away from the maximum magnitude area of the annular eclipse. The period of the quasi-periodic disturbances was observed to be 40 ± 2 min. These disturbances are suggested to be produced by acoustic and atmospheric gravity waves under the action of the solar eclipse. Changes in atmospheric wave pressure were estimated to be ~1-2%. The estimates of both aperiodic and quasi-periodic effects are substantially in agreement with the observations, which confirms the mechanism for their generation.