A novel type of a continuously zoomable telescope is based on two pairs of adjacent toroidal lenses ("saddle lenses") in combination with standard optical components. Its variable magnification is adjusted by a mere rotation of the four saddle lenses around the optical axis. This avoids the necessity of classical zoom systems to shift multiple lenses along the longitudinal axis of the setup. A rotationally tunable pair of saddle lenses consists of two individual saddle lenses (also known as quadrupole lenses, or biconic lenses), which are arranged directly behind each other, acting as a "combi-saddle lens." The transmission function of such a combi-saddle lens corresponds to that of a single saddle lens, but with an adjustable optical power which depends on the mutual rotation angle between its two components. The optical system contains two of these combi-saddle lenses, and acts as a cylindrical Kepler telescope in one plane, and as a cylindrical Galilei telescope in the orthogonal plane. The two orthogonal Kepler/Galilei telescopes stay aligned and change their magnification factors in the same way when the telescope is zoomed by adjusting the optical powers of the two combi-saddle lenses. Altogether this produces a sharp image, which is mirrored with respect to the axis of the Kepler telescope. Thus, in addition to the zooming capabilities of the telescope, it is also possible to rotate the resulting image by a rotation of the whole telescope, or of all included saddle lenses. The operation principle of the telescope will be explained in both a ray-optical and a wave-optical description.