Sept. 26, 1889] 



NATURE 



5'9 



I should say, perhaps, that the birds maintain the kite-like in- 



■cline, all round the wimhvard portion of the spiral, thus A < ' #1 



and appear to gain elevation in sweeping round while thus at an 

 angle ; at first facing the breeze, and as they sweep round 

 being side on, as per diagram, the centrifugal element coming 

 into play. If very low down, indeed, the bird can be seen to 

 rise distinctly as it wheels round the windward edge «f the 

 curve, and reaches its greatest elevation just as it turns tail to 

 wind. Then it seems to shoot along down a slight incline with 

 wings horizontal, and in rounding the leeward part of the 



spiral, the outer wing is again spiral, thus 



The 



speed of the bird so greatly exceeding that of the wind, 

 renders this necessary to enable it to turn. 



Viewed from below, the tracks look circular, and often vary 

 in size a good deal. 



In all cases, so far as I have seen, these birds when rising, as 

 they sail round and round, make leeway, say a mile for every 1000 

 feet rise, or less. 



They ne%'er rise on a straight course, though they often descend 

 in a straight line, and, after a time, flocks that drift over the 

 hills recover their position over the plains by descending to 

 windward. 



But in all these cases of land-birds rising steadily as they sail 

 round and round, there is no possibility of lifting by waves, as 

 at sea, nor yet differences in the speed of air strata every 20 feet 

 or so, as they sail at all heights, from 500 feet up to 10,000 or 

 more. 



The problem is much more clearly seen, and more wonderful, 

 than in the case of the albatross, and, as far as I can see, the 

 explanation which I gave years ago is the on I" one feasible. 

 The momentum gained in descending a given distance with the 

 -wind is expended in lifting the bird kite-like as it turns z.nA. faces 

 the wind, on the shorter windward course, where also a certain 

 amount of centrifugal reaction comes into play and does some of 

 the work. S. E. Peal. 



Sibsagar, Assam, August 8. 



Bishop's Ring and Allied Phenomena. 



Is there any connection between the sunset glows seen by Mr. 

 S. E. Bishop in the Sandwich Islands (see p. 415) and the 

 phenomena which have occurred lately in Western Europe? 

 There has been for two months a feeble reappearance of a great 

 corona round the sun. I do not know whether I can call it 

 Bishop's ring, as it has generally been larger than that caused 

 by the Krakatab dust, and also more dirty-looking in colour, and 

 doubtless at a lower level. Bishop's ring itself had never 

 altogether ceased t > be visible about sunset and sunrise, but I 

 had not certainly seen it at any time when the sun has had a 

 considerable altitude for nearly two years, and then only in the 

 clear Alpine air. On May 14 and August 26, 1888, there was 

 indeed a great corona, but on the latter occasion the air was not 

 very clear, so I concluded it probably had some other cause than 

 volcanic dust ; and there have been a few other occasions before 

 that when I have had the same impression. On May 14, 1888, 

 it is also questionable whether the circle was caused by volcanic 

 dust. 



The present corona appeared all at once on June 30 last, while I 

 was at Macugnaga, at the foot of Monte Rosa ; and since that 

 date, both travelling in the Alps and here (since I came home on 

 August 8) it has continued more or less visible at all times of the 

 day when not ol)scured by mist or smoke, except from the after- 

 noon of July I until the afternoon of the 2nd. It is plainest 

 about sunset. It first appeared on cloudy streaks like very thin 

 indefinite cirrus, but these disappeared during July i. Since it 

 appeared the sky has never been such a pure blue as it was 

 before ; this applies to all altitudes from sea-level to 10,000 feet. 



I have not seen any remarkable sunset phenomena, so that it 

 seems difficult to connect this circle with the phenomena noticed 

 by Mr. Bishop a fortnight later than this commenced ; still the 

 connection is not impossible, as he mentions seeing the central 

 (and by far the brightest) part of the corona. 



Although it is improbable that the dust of Krakatab can have 

 entirely settled to the ground, the above description will show 

 that the present phenomenon can hardly be caused by it ; but 



there must be a very extensive diffusion of some kind of dust — it 

 may be one of the kinds alluded to by Prof. Tyndall and others 

 in the discussion which occurred in your pages last year. 

 Sunderland, September 21. T. W. Backhouse. 



OBSERVATIONS OF TWILIGHT AND ZODT- 

 ACAL LIGHT DURING THE TOTAL ECLIPSE 

 OF THE SUN, DECEMBER 21, 1889. 



'T* H E attention of all friends of astronomy and meteoro- 

 -^ logy, and especially of navigators on the ocean, and 

 of meteorological observers in India and Mexico, is re- 

 spectfully called to the fact that they may make valuable 

 observations during the coming total eclipse of the sun. 

 The observations referred to are of a most elementary 

 character, and pertain to the phenomena of twilight and 

 of the zodiacal light. The importance of the observa- 

 tions to be described was first appreciated by German 

 meteorologists, and the following paragraphs are based 

 upon suggestions made by Prof. Bezold and Dr. Zenker, of 

 Berlin. 



On December 22, 1889, at 47 minutes before noon, 

 Greenwich time (which corresponds with 6 hours and 13 

 minutes a.m. of standard Eastern time) our globe becomes 

 tangent to the long cone that constitutes the shadow of 

 total darkness behind the moon, and we say that 

 " totality has begun on the earth " ; this occurs at a point 

 in the Caribbean Sea, north of Venezuela, at about 15° 

 N. lat. and 721° W. long. In about three hours and a 

 quarter this zone of total eclipse has passed the earth ; its 

 last point of contact, ortangency, occurs at a point on the 

 eastern coast of Africa at about 5^° N. lat. and 49° E long. 

 The path of totality on the earth's surface between these two 

 limits is a narrow belt, about 100 miles broad, and astro- 

 nomers will station theoiselves at favourable localities in 

 order to observe the phenomena visible immediately 

 around the sun, but meteorologists and local amateurs 

 residing entirely outside the path of totality can also do 

 good work. 



The accompanying diagrams, I., II., III., IV., will 

 illustrate the conditions of the eclipse and the nature of 

 the meteorological observations desired to be made. 

 Thus in diagram I. we see the point R, or the locality 

 where the totality begins to an observer on the earth ; 

 at that point the sun rises while it is totally eclipsed, and 

 the band from R to P shows the path of the shadow as it 

 flies eastward. Diagram II. takes up the path at its 

 other end, and shows it still moving eastward from P 

 towards s, where it finally leaves the earth, so that an 

 observer at S sees the sun in his western horizon, but 

 totally eclipsed as it is setting out of sight. 



If we now imagine a vertical plane passing through the 

 observer and the eclipsed sun, it will give us a section of 

 the earth, its atmosphere, and the cone of darkness very 

 much as shown in diagrams II. and IV. In diagram II. 

 R represents, as before, the locality where the sun rises 

 eclipsed, the moon is to the east of the observer, and the 

 cone of darkness passes over him towards its apex, 

 which is far off towards the west. Those who are 

 located to the west of R, as at D, looking eastward, can- 

 not see the sun because it is below their horizon ; they 

 can see only the light of the morning dawn where the sun is 

 about to rise, but these observers, if they look directly 

 toward the brightest part of the dawn, immediately over 

 the sun, will, on this morning of December 22, find that 

 they are looking right into the shadow of the cone, and 

 will therefore observe that the light of the dawn is for a 

 few minutes much feebler than usual. 



Similarly diagram IV. shows the condition of things at 

 the point s, where the observer at sunset has the cone of 

 darkness imediately above him ; observers to the east of 

 his location, as at x, will be enjoying twilight, and as they 

 look toward the west, immediately above the sun, they 



