[10] This observation is not without a certain amount of confirmation by more recent ones, in which certain lunar objects and regions have been suspected of mist or vapour. Mr. Birt (‘English Mechanic,’ vol. xxviii. no. 725) mentions two—the cloud-like appearance of the white patch west of Picard, and the interior of Tycho, which at one time always misty and ill-defined, is now become perfectly distinct and sharply defined. December 4, 1878, 4h 45m. I observed Klein’s crater as a dull dark spot, larger than the true object; and while definition was good and other objects were well defined, “the floor of Klein’s object, the oval spot near, and also Agrippa (especially), all had _an odd misty look as if vapour were in or about them_” (‘English Mechanic,’ vol. xxviii. no. 727). The mystery of different observers seeing and not seeing Klein’s object on the same night is hardly to be accounted for by the angle of illumination. [11] Some doubt has been cast on this observation, on the ground that nothing unusual was seen, and that the appearances were only those ordinarily presented by the moon at its then phase. I simply give the account as it appears in the scientific journal in which it was published. [12] The question of a connexion between the waxing and waning of the solar corona and the prevalence of sun-spots is now being mooted, and may have an important bearing on the subject of the constitution of the corona. It would seem that when the corona has been examined about the time of minimum of sun-spots, it has proved fainter though more extended, while the bright lines of the spectrum have been absent, indicating a change or variance in the gaseous part of it at those periods. [13] There seems to be some confusion as to the W.L. here given; 5567 is usually accepted as Ångström’s line, while Prof. Smyth refers to it as 5579. The position, too, when examined with a spectroscope of greater dispersion, is not exactly over the citron-line of acetylene, both the above referred to lines lying somewhat more towards the violet end of the spectrum (see Plate V. fig. 7). [14] Ångström’s drawing, in giving this character to the two Aurora-bands which are said to correspond with violet-pole bands about 47 and 43, is incorrect, and calculated to mislead by giving the Aurora-bands a feature corresponding to the violet-pole bands which they do not possess. I am not aware of any Aurora-line or band which is described as distinguished by degrading towards the violet. [15] The tubes generally seem marked Si Fl instead of the ordinary notation Si F. Si Fl₆ is probably, in fact, Si F₄. [16] Dr. Schuster has found that while the line-spectrum of lightning is attributable to N, it has also a band-spectrum, which he considers due to O and a slight trace of CO₂ (Phil. Mag. 5th ser. vol. vii. p. 321). [17] _In these observations some suggestions made by Mr. Capron have been incorporated._