16, p. 231) and failed to distinguish in the dispersed image the light contrast between crescent and cloud. Five minutes later I exposed a plate in the 4-prism spectrograph, using the light reflected from the clouds near the sun ; the resulting negative has unexpectedly great density. Four minutes before totality I was able, through a gap in the cloud, to adjust the visual objective grating (see 17, p. 233), and 2 minutes later to perform a similar adjustment for the photographic objective grating, an operation which 8 minutes earlier I had given up as hopeless. The cloud was a sort of mackerel cirro-cumulus, giving momentary clear views of the crescent, which was more generally partially obscured by the flocculent masses. As the sunlight faded with the waning of the crescent, the visibility of this veil of cloud decreased, but its presence was clearly indicated by the coloured rings. Cloud continued in greater or less degree throughout the whole of totality, as will be gathered from the record, given in the following paragraphs, of the observations made with the various instruments.
8. Huts far the Instrument and Dark Room.
Two huts wooden framework and Willesden canvas were taken out to Sumatra.
One was 15 feet 2 inches long, 10 feet 10 inches wide (the length of the sides being multiples of 2 feet 2 inches, the width of the canvas), 6 feet high at the eaves, 8 feet 6 inches high at the ridge. The ridge- pole was not central, but 4 feet 2 inches from one side and 6 feet 6 inches from the other ; two canvas-covered doors were provided, one in each long side, at the ends of a diagonal, and the framework of the hut was covered over with Willesden canvas, except over about two- thirds of one side of the roof. To cover this part six thin sheets of the lightest galvanised iron (6 feet by 2 feet, 10 Ibs. to a sheet) were used as a roof which could be easily taken off when it was desired to make observations or to let in light and air into the hut, and which, when replaced, formed a water-tight roof that resisted even tropical rain. Over the hut, and quite separate from it, a light bamboo frame- work was erected by native workmen, and a thatching of rough grass (" alang-alang ") was fitted over it, leaving only a space of a foot and a-half between it and the canvas roofing ; the thatching was, of course, omitted over the galvanised-iron sheets. This hut was used to cover all the instrument, and was found extremely convenient.
The other hut was 10 feet 10 inches long, 8 feet 8 inches wide, 6 feet high at the eaves, 8 feet 6 inches high at the ridge. It had been used in Algiers for the instrument hut ; on the present occasion it was converted into a photographic dark room. For this purpose a roll of Willesden waterproof paper, 5 feet wide, ordinarily used for under- slating, had been taken out to Sumatra, and was used to make the
