On the other hand, the spots afford a ready means by which the observer may for himself determine approximately the rotation period of the sun. A spot will generally appear to travel across the solar disc in about 13 days 14-1/2 hours, and to reappear at the eastern limb after a similar lapse of time, thus making the apparent rotation-period 27 days 5 hours. This has to be corrected, as the earth's motion round the sun causes an apparent slackening in the rate of the spots, and a deduction of about 2 days has to be made for this reason, the resulting period being about 25 days 7 hours. It will quickly be found that no single spot can be relied upon to give anything like a precise determination, as many have motions of their own independent of that due to the sun's rotation; and, in addition, there has been shown to be a gradual lengthening of the period in high latitudes. Thus, spots near the equator yield a period of 25·09 days, those in latitude 15° N. or S. one of 25·44, and those in latitude 30° one of 26·53. This law of increase, first established by Carrington, has been confirmed by the spectroscopic measures of Dunér at Upsala. His periods, while uniformly in excess of those derived from ordinary observations, show the same progression. For 0° his period is 25·46 days, for 15° 26·35, and for 30° 27·57. Continuing his researches up to 15° from the solar pole, Dunér has found that at that point the period of rotation is protracted to 38.5 days. Reference has already been made to the bright and fantastically branched features which diversify the solar surface, generally appearing in connection with the spots, and best seen near the limb, though existing over the whole disc. These 'faculæ,' as they are called, will be readily seen with a small instrument--I have seen them easily with a 2-inch finder and a power of 30. They suggest at once to the eye the idea that they are elevations above the general surface, and look almost like waves thrown up by the convulsions which produce the spots. The rotation-period given by them has also been ascertained, and the result is shorter than that given by the spots. In latitude 0° it is 24·66 days, at 15° it is 25·26, at 30° 25·48. These varieties of rotation show irresistibly that the sun cannot in any sense of the term be called a rigid body. As Professor Holden remarks: 'It is more like a vast whirlpool, where the velocities of rotation depend on the situation of the rotating masses, not only as to latitude, but also as to depth beneath the rotating surface.' Plate VII., from a photograph of the sun taken by Mr. Hale, in which the surface is portrayed by the light of one single calcium ray of the solar spectrum, presents a view of the mottled appearance of the disc, together with several bright forms which the author of the photograph considers to be faculæ. M. Deslandres, of the Meudon Observatory, who has also been very successful in this new branch of solar photography,