ON March 16 I observed a peculiar halo here; its form is best shown by a rough sketch. The angles were taken with a pocket slide-rule held at arm's length, and are, therefore, only approximations, but the relative values are probably fairly correct. The halo was brightest at the point above the sun, and faded off somewhat on each side; it ended rather abruptly at the points shown in the sketch. The colours, with red nearest the sun, were not very pure, but they were purer in the arms than in the centre. The phenomenon was visible from 14.45 to 15.46, with intervals of disappearance when a sheet of alto-stratus became so thick that nothing could be seen through it but the glare of the sun. It was not possible to see any higher layer of cloud, but the halo probably had its origin in a layer of cirro-stratus. The measurements were taken at 15.40; a few minutes later the halo disappeared for the last time.