January 26, 1893] 



NA TURE 



305 



able to take some photographs of the phenomena and to measure 

 the actinic intensity of the corona. Two years previously I 

 had been to Russia to observe the eclipse of August i8, 1887, 

 but the bad weather prevented any observations. If these ex- 

 peditions did not succeed as well as I had hoped, they were at 

 least useful in showing me all the difficulties to be met with in 

 such undertakings, and of convincing me that if one wishes to 

 thoroughly avail one's self of the precious moments during which 

 the eclipse lasts, it is necessary to gain a large experience of 

 these phenomena by omittingnoopportunity of observing them, 

 and by making a speciality of these expeditions. Therefore, 

 after the eclipse of 1889 I determined to go to observe the 

 following eclip-ie, that of the i6th of next April. 



This time the phenomenon is visible under particularly 

 favourable conditions. On the African coast to the south of 

 Dakar, where the expedition sent by the Bureau des Longitudes 

 will observe, and where I also propose to instal myself, the 

 duration of totality is four minutes thirteen seconds. Moreover, 

 a very important consideration is that we are certain of fine 

 weather. At a time when expeditions are being organised in 

 every country in view of this astronomical event, I think it will 

 be useful to draw your attention to the chief questions which 

 should be the object of astronomical study during the next 

 eclipse. 



You are aware that the passing of the moon before the sun has 

 the inestimable advant^ige of allowing us to see the circumsolar 

 regions which, under ordinary circumstances, because of their 

 faint light are lost in the general illumination of our atmosphere. 

 The regions thus revealed consist of a layer in immediate con- 

 tact with the sun, the chromosphere, m which are the rosy 

 flames which form the protuberances ; and a more or less exten- 

 sive luminous aureole, the corona. But since the celebrated 

 eclipse of 1868, which marks an epoch in the history of solar 

 physics, we are able, thanks to the great discovery of Messrs. 

 janssen and Lockyer, to study the protuberances at any time, 

 and consequently it is only to the corona that the attention of 

 astronomers turns during total eclipses. 



An invariable part of the day's programme is the study of the 

 structure of the corona, and the luminous intensity of its various 

 parts. We need to have recourse to photography for trustworthy 

 evidence as to this, for photography alone can give a faithful 

 representation of the phenomena ; even the best drawings always 

 leave much to be desired. Indeed, it is impossible in the space 

 of a few minutes to exactly represent a nebulous mass as com- 

 plicated as the corona, and without any definite outlines. We 

 can judge of the difficulty presented by the drawing of the 

 corona, by remembering that even the most skilful draughtsmen 

 have never yet been able to give us similar drawings of the great 

 Orion nebula, although this may be studied at leisure. The 

 brilliancy of the corona varies in intensity so much from one 

 eclipse to another, that it is difficult to determine beforehand 

 the length of exposure needed with given apparatus to obtain as 

 satisfactory a representation of the phenomenon as possible. 

 Moreover, the different parts of the corona differ in brilliancy, 

 so that when the photographic action is sufficient to give a good 

 image of the middle part, the lower portions which form the 

 interior corona are over-exposed, while the extreme limits of the 

 aureole are not reproduced. To satisfy all the conditions it is 

 necessary to take several photographs with different exposures. 



To advantageously discuss the results obtained it is very im- 

 portant that astronomers should place upon each plate a uniform 

 scale to measure the intensity of the photographic action upon 

 it. This intensity is equal to the product of three factors ; the 

 effectiveness of the object glass, the length of exposure, and the 

 sensitiveness of the plate. If we indicate the useful diameter 

 f the object glass by a and the focus by f, the effectiveness, 

 ilefined by the international congress of photography, is 



100 -. If we take plates of gelatinobromide of silver of 



normal sensitiveness as our unit, and let / be the length of 

 exposure in seconds, we have the following formula to express 



the photographic action :— 100 '^^if. In working with wet 



collodion plates it is necessary to multiply this expression by 

 ',7. and with plates of dry collodion it must be multiplied by 

 .Vix. The first photographs of the corona, taken with wet 

 collodion, from 1868 to 1878, were obtained with a photographic 

 action not greater than 2. Later, thanks to rapid plates, much 

 greater action could be obtained. Thus in 1883 a photograph 



NO. 



213. VOL. 47I 



obtained by M. Janssen had received a photographic action 

 equal to 918. On the negative thus obtained, the corona 

 extended to between 30' and 40' from the limb of the moon, 

 but on the other hand, details of the parts near the sun were 

 completely wanting. 



We might ask whether by still further increasing the photo- 

 graphic action we should also extend the limits of the corona ? 

 Certainly not ! for if the photographic action is too intense, the 

 faint contrast between the extreme parts of the corona and the 

 more or less illuminated sky is no longer appreciable on the 

 negative. We know, indeed, that if we wish to produce the 

 maximum contrast between two half tones we must only use 

 just enough light for the faintest of the half-tones to give a 

 perceptible image. In America, Mr. Burnham has been en- 

 gaged in determining the maximum length of exposure 

 necessary to obtain the best representation of the corona, and he 

 has made experiments on this subject by photographing the 

 moon and white clouds on a faintly lighted sky. 



In 1889, at the Salut Isles, I used five instruments, giving 

 photographic actions varying from 185 to 13, but, doubtless on 

 account of the peculiarly intense illumination of the atmosphere 

 due to the short duration of totality, and the great abundance 

 of water vapour in the atm.osphere, the most satisfactory 

 negative was that corresponding to a photographic action of 30. 

 It is very probable that an equally good result might have been 

 obtained with much less photographic action. Mr. Barnard, to 

 whom we owe the best photographs of the eclipse of January 

 I, 1889, worked with a photographic action equal to 058. 

 This result proves that with the sensitive plates now in use it 

 should be possible to obtain good images of the corona on a 

 large scale by using secondary magnifiers to increase the size of 

 the image given by the object glass. In any case we can em- 

 ploy object glasses of two or three metres focal length, virhich 

 would give images sufficiently large to show all the details of the 

 corona without having resort to enlargement of the plates. 

 Nevertheless, in spite of the use of a long focus, the instrument 

 must remain luminous enough for the time of exposure to be 

 short. The displacements of the image on the plate, caused by 

 the imperfect adjustment of the equatorial mounting, or by an 

 irregularity in the clockwork movement, or by the movement of 

 the sun in declination are thus rendered inappreciable. 



The photographs, when obtained, should be examined from the 

 following different points of view. First of all we wish to find 

 if the corona, which will be observed in the month of April, 

 1893, at a period of great solar activity, and at an epoch when 

 the south pole of the sun is projected on the visible part of its 

 disc, resembles, as we have every reason to think it does, the 

 corona of 1883, which was studied under the same conditions. 

 A great resemblance between the forms of the corona in 1889 and 

 1878, at the periods of minimum sun-spots, has already been 

 noted, and if it can be established that the corona, seen under 

 similar conditions, presents the same appearance, it will be 

 proved that the diversity of appearance hitherto noticed depends 

 solely upon the state of agitation of the solar surface, and the 

 position of the observer with respect to the solar equator. 



If the corona should present an axis of symmetry it must be 

 ascertained whether the poles of this axis coincide with the poles 

 of the axis of rotation of the sun ; or if, as is more usually the 

 case, the poles of the corona are inclined at some degrees from 

 the poles of the sun, thus resembling the position of the magnetic 

 poles of the earth with regard to its geographical poles. If the 

 corona shares in the movement of rotation of the sun, it must be 

 the same with its axis of symmetry, and therefore if we once 

 observe the northern pole of the corona to the east of the northern 

 pole of the sun, we ought to find it after an uneven number of 

 half- revolutions, of the sun to the west of the north pole of the 

 sun. To ascertain if this is so or not, it is of the greatest im- 

 portance to know exactly the orientation of the images upon the 

 photographic plates. The most exact and simple method to 

 orient the images is to place the photographic apparatus in the 

 position which it occupied at the moment of the phenomena, 

 and, in the night, to photograph the trails which the stars leave 

 in passing across the field of the lens. 



If the photofjraphs should show the structure of the corona 

 clearly, we shall be able to study the form of those luminous 

 rays which we notice at the poles of the sun, and of those cur- 

 vilinear structures which seem to extend from the middle lati- 

 tudes of the sun. The study of the curvature of these structures 

 will be very useful in verifying the exactness of one of the most 

 favoured theories of the corona, which explains the phenomena 



