The main problem in building large mirrors is that the mirror’s thickness should be proportional to the diameter. In fact, if the disk is too thin, the glass bends and loses its shape. This was an insuperable obstacle at the time the multi-mirror was being developed, whereas present-day technology allows mirrors over 4 metres in diameter, monolithic, to be produced.
Hence the idea of Horn-d’Arturo - director of the University Astronomical Observatory in Bologna from 1920 to 1954, with brief interruptions - to build a mirror made up of small hexagonal mirrors with converging foci in the same focal plane so as to obtain an integral stellar image by bringing together the data from each individual mirror. The resolving power of a small hexagonal mirror is less than that of a single mirror with diameter equal to the total size of the composite mirror; it is however also true that what determines the quality of the observation in astronomical photography, as indeed in photoelectric photometry and spectroscopy, is not the diameter of the mirror but the image of "seeing", i.e. the dimensions of a stellar image altered by atmospheric turbulence.

The mirror consists of 61 small hexagonal mirrors, total diameter 180 cm, and forms the object lens of the zenithal telescope located on the first floor of the Specola tower; it lay horizontally 50 cm from the floor and received light from a circular hole made in the floor of the terrace on the penultimate storey, under the dome which held the meridian circle of Ertel & Sohn [file 22]. To get round the immobility of the mirror, a movable plate-holder was used that could follow the path of a star during zenithal transit. When seeing was good, it was possible, with exposition times just under seven minutes, to achieve the eighteenth magnitude using the photographic material of the time, usually ultrasensitive Cappelli plates produced by the Ferrania company, size 9 x 12 and 9 x 24 cm.
Construction of the instrument was slow and in stages - from the 10 small trapezoidal mirrors in 1935, supplied by Filotecnica Salmoiraghi with a reflecting surface of just over 10 square decimetres, to the 19 hexagonal mirrors produced by the Specola’s technician Aldo Galazzi, with a reflecting surface of about 4 square decimetres. This prototype, installed in 1950 in the middle of today’s "Turret Room", using for observation purposes the zenithal hole already present in the vault, was approximately one metre in diameter.
The instrument was finally completed in 1952 with 61 mirrors.
The mirrors have a spherically curved polished surface, with a radius - the same for all - equal to 20.82 m and a focal distance of 10.41 m. They are placed on a marble plate which acts indirectly as support and which has holes for the three screw pegs that hold each mirror. The screw plate, fitted with slides for sideways movement, is attached to the marble and, by operating from the small room beneath the mirror, the pegs can be moved as desired, and with them the mirrors, both laterally and vertically.
The hole was closed in the 1980s; in the same period the marble plate supporting the prototype of the instrument was destroyed.
At the end of the 80s all the mirrors were recovered, realuminized, mounted in their original positions and covered with a crystal plate. The more than ten thousand plates obtained with this telescope were packed in boxes and put in the small room, under the mirror, which was used to make adjustments to the mirrors.

R. Berry (1988), pp. 42-47.
G. Horn-d’Arturo (1932-1955)
L. Jacchia (1978), pp. 100-102.