INVESTIGATIONS WERE CONCERNED WITH THE DESIGN OF A RADIO TELESCOPE WITH A ONE-MINUTE-OF-ARC BEAM. The feasibility of a one-minute-of-arc in one dimension was established. However, only a relatively small collecting area was used. To increase the area and to form a pencil beam, while keeping the structure close to the ground, one is led to consider echelon configurations. A multifocus arrangement of tiltable cylinders, stacked like the slats of a venetian blind was studied. The suppression of grating responses can be accomplished by a new tilt-independent technique in which a staggered pair of elements is switched against the remaining elements. The collecting area of a system in which some of the elements are switched, proves to be reasonable. Problems of interconnecting the tiltable elements so as to counteract the jump in path from one element to the next were partly met in achieving a one-minute fan beam. Multibeaming, a technique of halving observing time by doubling the investment in electronic equipment, is justified when the antenna cost is high and the beamwidth extremely narrow. (Author).