Janu.\ey 1, 1894.] 



KNOWLEDGE 



21 



questions, important in connection with future investiga- 

 tions of the sun, let us sum up some of the conchisions to 

 which we have been led. 



The facul* are the elevated regions of the photosphere. 

 They form an irregular reticulation over the entire surface 

 of the sun," and in the sunspot zones appear as irregular 

 bright regions of varying extent. The spectrum of the 

 faculiP is similar to the general spectrum of the sun, but is 

 somewhat brighter, and contains the doubly-reversed 

 calcium lines H and K. I The hot calcium vapour from 

 which these lines emanate may be diffused throughout the 

 mass of a facula, or confined to its upper surface. In the 

 latter case, the base of the chromosphere and the upper 

 surface of the facula would practically coincide. The 

 white-hot particles giving a continuous spectrum would 

 ordinarily be found only in the facula proper, and (more 

 sparsely scattered) in the lower region of the chromosphere. 

 Eruftive prominences are closely related to facuhe, and 

 probably rise from them. It thus occasionally happens that 

 a violent eruption carries some of the white-hot particles 

 to a considerable distance above the photosphere. In such 

 a case the prominence gives a continuous spectrum in 

 addition to its bright lines. While some faculiB are 

 covered by prominences, others do not appear to be so 

 covered. Certain exceptionally bright eruptive prominences 

 have been photographed in projection on the solar disc. 

 Ordinary prominences in the region of the sun's poles are 

 not shown in spectro-heliograms. Sunspots, even in their 

 central parts, seem to be covered by the chromosphere (or 

 by overhanging prominences). The chromosphere (or 

 prominences) is frequently so bright as to completely 

 hide small spots in spectro-heliograms. 



As M. Deslaudres has discussed the instruments and 

 methods employed in my photographic investigations of 

 the sun, I may perhaps be allowed to express an opinion 

 as to the most advantageous manner of continuing these 

 researches. 



The spectro-heliograph used in the greater part of my 

 photographic work has a pair of slits arranged to move in 

 the focal planes of the collimator and observing telescope 

 of a large diffraction spectroscope, attached to a 12-inch 

 equatorial refractor.,': It has proved itself a thoroughly 

 practical instrument, and from two to twenty or more 

 photographs of the forms of prominences and calcium 

 reversals have been made with it on every clear day (with 

 few exceptions) since -January, 1892. The photogravure 

 reproductions which accompany this article were made 

 from negatives secured by the aid of this apparatus. 

 Although they fail to bring out the more delicate details 

 shown on the original i}lates, the illustrations surely 

 demonstrate that the combination of a fixed spectroscope 

 with two moving slits is not without certain merits. 

 Neither my assistant, Mr. EUerman, nor myself have 

 experienced any diflQculty in using a second slit 0-005 inch 

 wide, and this width could be decreased were it considered 

 desirable. By increasing the width of the second slit, and 

 making a series of photographs of the K line at various 

 points on the solar disc with the slit stationary, the 

 character of the double reversals can be studied. This 

 method of successive sections, which I first employed in 

 1891, seems to me quite as convenient as that used by 

 M. Deslaudres. As the slits always move together, there 

 are no troublesome adjustments to be made by hand. 



* At least during the maxiiiuiin period of simspots. 



t The less refrangible hydi'ogen lines may also be preseiil, but 

 under ordinary conditions they are too faint to be recognized. 



X This telescope will be removed to the Yerkes Observatory, where 

 the spectro-heliograph will be employed, as at present, in securing a 

 daily record of solar phenomena. 



But after studying and e.^perimenting with a great 

 variety of instruments, I came to the conclusion early in 

 the year 1893 that a spectroscope with collimator and 

 observing telescope parallel (or nearly so) to each other, 

 and slits fixed in the axis of each, the whole instrument 

 being arranged to move on wheels at right angles to the 

 axis of the large telescope, would possess important 

 advantages over all other forms of spectro-heliograph. 

 An instrument of this type, suitable for use with a heliostat 

 for automatically photographing the sun, was described in 

 my paper entitled " The Spectro-heliograph " [Astnmomij 

 and Astro-Physk's, March, 1893, ]). 256). I have since 

 designed a spectro-heliograph on this principle for the 

 40-inch Yerkes telescope. A short description of this 

 instrument will be found in the January (1^94) number of 

 Antriinoini/ ami Astro- I'liyt^ics. The proposed " rotating 

 spectroscope," for which M. Deslaudres' various papers 

 have claimed many advantages has, I believe, been 

 recently abandoned for a spectro-heliograph of this class. 

 A similar instrument is to be used at the Maharajah 

 Takhtasingji Observatory at Poona, India. 



The question raised by M. Deslaudres, in regard to 

 the best dispersion to employ, is a most interesting and 

 important one. On the one hand, a feeble dispersion would 

 seem to offer important advantages on account of the 

 narrowness of the K line and the greater brightness of 

 the image. On the other hand, it must not be forgotten 

 that the fainter details of the reversals may be lost if the 

 dispersion is insufficient. M. Deslaudres recognizes this 

 fact when he states that the exceedingly faint H and 

 K reversals in the general spectrum of the sun are best 

 obtained with a "spectroscope puissant" (Knowledge, 

 December, 1893, p. 231) ; and again, when he remarks that 

 for the study of the details of the reversals on the solar 

 disc " une grande dispersion est necessaire " (Uiiil, p. 232). 

 This advantage of high dispersion is due to the fact that 

 the width of the bright lines is not proportional to the 

 dispersion, at least up to a certain limit. If by doubling 

 the dispersion the lines were doubled in width, there would 

 evidently be no change in their brightness as compared 

 with the solar spectrum in which they lie. But, within 

 certain limits, the brightness of the lines relatively to the 

 solar spectrum increases with the dispersion. It is e\ident, 

 therefore, that we must not employ too feeble a dispersion. 

 I have found the fourth order spectrum of a Kowland 

 grating (14,438 lines to the inch) very suitable, though I 

 should have preferred prisms had circumstances permitted 

 their use. The separation of the H and K lines at the 

 focus of the observing telescope, where the photographic 

 plate is placed, is nine millimetres. M. Deslandres 

 uses a single prism, giving a separation of H and K 

 of two millimetres, and magnifies the image of the 

 second slit — and, consequently, the width of the K line — 

 three diameters. The resulting separation of H and K 

 on his photographic plate would thus be six millimetres, 

 and the width of the K line two-thirds the width of the 

 K line in my instrument, if the line were supposed to have 

 a width proportional to the dispersion. As has already 

 been pointed out, the width of a line is not proportional 

 to the dispersion, and it is probable that there is no 

 great difference in the efi'ective width of the line in the 

 two instruments. Thus, any question in regard to dis- 

 placement due to motion in the line of sight, width of 

 the second slit, &c., would apply almost equally in both 

 cases ; but the greater dispersion of the grating would 

 increase the brightness of the K line as compared with 

 the solar spectrum, and fainter and more delicate calcium 

 reversals should be obtained by its means. 



It would probably be advisable to have a set of three 



