No^-EMBER 2, 1896.] 



KNOWLEDGE. 



249 



to build large stora,f,'e reservoirs at the level of the top of 

 the tide. They would fill at the Hood, and gradually run 

 down to reservoirs on low ground, which would be emptied 

 at the ebb. By means of turbines and dynamos this 

 power could be transformed into the energy of the electric 

 current, which would turn all the machinery in the city — 

 if not in the whole county — besides lighting and heating 

 the private houses and public buildings. Another im- 

 portant point is that our stores of coal are by no means 

 inexhaustible. In fact, they are dimiaishing rapidly, and 

 the most liberal estimate does net give our coal supply 

 more than one hundred and fifty ytars' duration at the 

 present rate of consumption. Long before that time the 

 depths of the existing shafts will have to be increased so 

 much that the cost of getting the coal will be very much 

 augmented, and we shall be placed at a great disadvantage 

 as compared with countries that have not drawn largely 

 on their hidden store. If we are able to utilize the 

 natural powers of the tides and rivers we shall be able to 

 eke out our supplies of fuel to an almost indefinite extent, 

 and the countries like our own with a large coast-line, 

 round which run strong tides, will still lead the van in 

 industrial progress. 



M 



OBSERVATIONS DE L'tCLIPSE TOTALE DU 

 SOLEIL DU 16 AVRIL, 1893.* 



DESL ANDRES' report comes at a most appro- 

 priate time, and it will be the more welcomed 

 since M. Deslandres has taken so prominent a 

 ^ part in that rapid advance in solar physics 

 which has been so marked during the last 

 few years, and since in the 1893 eclipse he attacked an 

 entirely new department of the solar problem. 



The report opens with remarks on the importance of 

 eclipses, on the special conditions of that of 1893, on the 

 necessity for using the photographic method in the study 

 of coronal forms, and on the threefold character of the 

 coronal spectrum. This is : first, continuous, from incan- 

 descent solid or liquid particles ; second, bright line from the 

 coronal gases ; third, a feeble solar spectrum from reflected 

 sunlight. M. Deslandres then speaks of his design to push 

 his researches into the ultra-violet region of the coronal 

 spectrum, and to determine the speed of rotation of the 

 corona by the relative displacement of its lines east and 

 west of the sun. This latter was not only a new investi- 

 gation, but one of the highest importance ; for if the 

 corona be composed of a vast network of meteoric streams, 

 its rate of motion would be much more rapid than if it 

 turned with the sun. 



The description of the instruments employed, and of the 

 manner in which they were arranged, wliich follows mxt, 

 is of great interest, but need not detain us now. The mc st 

 striking detail was the use of a double polar heliostat in 

 connection with three of the principal spectroscopes. The 

 day of the eclipse was not particularly fine, the sky being 

 by no means so clear as on the preceding days. However, 

 twenty-seven photographs of the corona and of its spectrum 

 were secured. 



The photographs of the corona itself were obtained with 

 three objectives of different ratios of aperture to focus, all 

 mounted on the same equatorial. It is, M. I)eslandres 

 points out, practically impossible to fix what will be the 

 best exposure beforehand ; this combination of three lenses 

 of very different relations gave him, however, a very wide 

 range, and the greatest extension seems to have been 



• Par M. H, Bcsliuidics. (Paris: Onulliii rVillars ct Fits.) 



secured with a telescope of fa 12^^, an exposure of forty 

 seconds, and a plate of medium sensitiveness. In all, 

 seven plates were obtained with each of the three tele- 

 scopes ; but the last plate of each set was spoiled by the 

 return of sunlight. The evaluation of the depth of 

 deposit caused by the coronal light shows that for the 

 region between 3' and 9i' from the limb, the light equalled 

 18 of a " decimal bougie" placed at one metre distance. 



Comparison of the West African photographs with those 

 taken in Brazil and Chili show no great change in the 

 corona in the interval of time ; but the comparison with 

 those of earlier eclipses enforces the conclusion long since 

 arrived at that the corona changes its form with the sun- 

 spot cycle. The threefold character of the coronal light, 

 alluded to above, was again manifested : the bright line 

 spectrum being strongest in the lower regions, the central 

 region showing rather a continuous spectrum, and a feeble 

 spectrum of reflected sunlight being given from remoter 

 districts. 



In spectroscopic work M. Deslandres was anxious to 

 push his researches far into the ultra-violet, thus breaking 

 up new ground. Five slit spectroscopes in all were 

 employed, but two yielded no results. The other three 

 gave good spectra, yielding the means for determining the 

 positions of a large number of bright coronal lines, and 

 the spectrum of the greatest dispersion enabled M. 

 Deslandres to obtain the first determination of the speed 

 of rotation of the corona. This, by far the most striking 

 result of his work, and worthy of the strongest emphasis, 

 gave a relative motion of the corona, east and west of the 

 sun, of 68 km. ; approaching, that is to say, the relative 

 motion of particles rotating with the sun. This result 

 would alone have been worth all the heavy labour 

 M. Deslandres devoted to the eclipse ; and though the more 

 powerful spectroscope he designed for this research gave 

 no result, and though we must hesitate to base theories on 

 a single set of observations, as yet unconfirmed and not 

 incapable of misconstruction, he is abundantly to be con- 

 gratulated on the success obtained with his second weapon. 



Other spectroscopic results were : the demonstration that 

 the continuous spectrum, which forms the major part of 

 the coronal light, is intensest towards the red, relatively to 

 the light of the disc, and this is more marked the higher 

 from the limb the point examined ; that the dark line 

 spectrum due to reflected sunlight cannot be detected at 

 5' from the limb ; and that the bright lines vary in 

 difl'erent parts of the corona and at different heights, and 

 do not usually correspond to known terrestrial elements. 

 The concluding chapter of the report is one of especial 

 interest. It is a review of the progress of our knowledge 

 of the solar atmosphere up to the present day, with 

 suggestions to which the long and careful consideration of 

 solar problems has given rise in M. Deslandres' own 

 mind. 



The first point touched upon is the relationship of the 

 periodic changes in the sun and its atmosphere, as, for 

 example, the association of a particular type of coronal 

 form with a solar spot minimum. The cause of this 

 periodicity is still unknown, but M. Deslandres favours the 

 sunspot theory of M. Faye, which regards sunspots as 

 having analogies to terrestrial cyclones, and explains the 

 differences in the surface speeds of the photosphere by 

 great movements of gas in the vertical direction. For the 

 corona an electric origin is favoured : M. Schaeberle'a 

 " mechanical theory,'' which is discussed somewhat at 

 length, being irreconcilable with the results mentioned 

 above as to the coronal rotation. In propounding an 

 electric theory, M. Deslandres, as ho points out, follows in 

 distinguished footsteps, Tacchini, Fi/eau, Secch\ lluggins, 



