On March 25th, at 8.15 P.M., very few clouds being visible, there appeared a luminous arch in a vertical plane, passing true E. and W. nearly through the zenith. This arch, to my surprise, gave everywhere a continuous spectrum. I had, in company with Professor Dewar, observed a phenomenon similar in all respects (continuous spectrum included) some seven years before, and was then inclined to fancy the arch a mere low-lying cloud, and its light due to the town lamps. But on the present occasion, after about forty minutes, when the arch had disappeared, an aurora, showing unmistakably in the spectroscope the characteristic green band, appeared near the magnetic north. And a few days afterwards there appeared in the Edinburgh newspapers several descriptions of a similar arch seen at almost exactly the same time at Colinton, &c., and even so far off as Dunfermline. Most of the observers remarked the entire absence of coruscations as a proof that the phenomenon could not be auroral. The facts stated seem, however, to leave no doubt that it was connected with the aurora and the peculiarity of its spectrum remains to be explained.