The reference print on which the 2017 restoration is based belongs to the Cineteca Italiana di Milano; it was taken from the original negative dated 1933, and given by Henri Langlois to Luigi Comencini in 1947. It is longer than the camera negative, which was reedited and corresponds to the 1945 release of the film and other prints known so far. As it turns out, the two edits were done in the same year, but ‘Milan’s’ is the earlier one.
Two different versions of a film existing at the same time was not exceptional. Why is it so, in the present case? One reason would be that the longer one exceeded the standard length of a supporting short feature (1200 m), as set by the contract between producer Nounez and GFFA. Or perhaps, after the initial disappointing audience reactions, Vigo, in an attempt to balance action and style, “opted for the second solution, reserving the right to add, if necessary, some explanatory intertitles” (Paulo Emilio Sales Gomes). Also, in the print matching the negative, some ‘censorable’ scenes were either shortened or cut: a silhouette of Surveillant Pète-Sec, injecting himself before going to sleep (a shot cut in the known version), the chemistry professor spitting into a handkerchief (in the known version only the sound is heard), his hand caressing Tabard’s (the shot was shortened), etc.
Finally one episode, that of the sleepwalker, was displaced and is now re-integrated into the night-time context of the dormitory and linked to other short scenes: preparations for the revolt; punishing an ‘informant’; the sleepwalker; Tabard and Bruel tenderly saying good night to each other, spied upon by Surveillant-Général Bec-de-Gaz.

Bernard Eisenschitz