are directed as in the present diagram. This would also happen in the same way if the division of the stone were through the plane of a tropick, and the mutual separation of the divided parts and the interval between them were the same as before, when the loadstone was divided through the plane of the æquator, and the parts separated. For the cusps are repelled by C, are attracted by D; and the versoria are parallel, the poles or the verticity in both ends mutually requiring it.
Half a terrella by itself and its directions, unlike the directions * of the two parts close to one another as shown in the figure above. 
All the cusps have been touched by A; all the crosses below except the middle one tend toward the loadstone, not straight, but obliquely; because the pole is in the middle of the plane which before was the plane of the æquator. All cusps touched by places distant from the pole move toward the pole (exactly the same as if they had been rubbed upon the pole itself), not toward the place where they were rubbed, wherever that may have been in the undivided stone in some latitude between the pole and the æquator. And for this reason there are only two distinctions of regions, northern and southern, in the terrella, justas
