16 



KNOWLEDGE. 



January, 1915. 



by Sir J. Herschel on the appearance of Hansen's tables 

 some sixty years ago. In giving the Gold Medal of the 

 Royal Astronomical Society to Professor Hansen, he said 

 his results " might be considered as a practical completion 

 of the lunar theorj', at least for the present astronomical 

 age, and as establishing the entire dominion of the New- 

 tonian theory over that refractor^' satellite." 



Not many years had passed before the tables began to 

 show errors that developed at an alarmingly rapid rate. 

 Newcomb made an e.xhaustive search for the cause, which 

 involved a ransacldng of old astronomical libraries to find 

 observations of occultations and ecMpses. He was able 

 in this way to trace back the place of the ^Nloon a century 

 further than Hansen had done ; he found that a Venus 

 term, inserted by Hansen, ought not to be there at all, 

 being really' quite insensible. He removed this, and aiso 

 altered the rate of the Moon's motion, as well as reducing 

 the acceleration in a century from 12" to 8". He found 

 that there was still a large wave in the motion, with a 

 period of about t\vo hundred and se\'enty years, and am- 

 plitude about 13". Vain attempts were made to explain 

 this term gra\-itationally ; it is now conclusi\-ely established 

 that it cannot arise from the action of any known matter 

 in the solar system, so that it is probably non-gravitational. 

 A similar wave has recently been foimd with a period of 

 about sixtv'-five years. 



It need hardly be said that the value of Professor Brown's 

 researches and new tables is in no way diminished by the 

 fact that we cannot expect the ]Moon to move in strict 

 conformity with them. It is a great point to know that the 

 tables correctly represent the action of all kno%\Ti forces, 

 and that any departure from them indicates either some 

 unknown body, or some force other than gra\-itation. 

 \\liile Professor Brown feels some natural disappointment 

 that his twent>' years' devotion to this great problem 

 has not completely unravelled all the mysteries of the Moon's 

 movements, he has taken a long step forward, and has 

 separated what is understood from what is mysterious. 

 He has the consolation that problems that await a final 

 solution are far more interesting than those completely 

 solved- Professor Brown gives a diagram showing that 

 there are unexplained fluctuations in the motion of the Sun 

 and Mercury in the same direction, though smaller in 

 amount than those in the Moon. He says : " We must 

 look for some kind of a surge, spreading through the solar 

 system, and affecting planets and satelhtes the same way, 

 but to different degrees." 



As an illustration of the extreme difficulty of predicting 

 the place of the Moon with accuracy, even for a few years 

 in advance, a comparison has been made of the new French 

 tables (based on Delaunay) with the obser\-ations made at 

 Green\\-ich this year. The observed R.A. is greater than 

 the predicted by 0*-20, that is, 3" in arc. This is only a 

 quarter of the error of The Nautical Almanac, but still 

 it is unexpectedly large, considering that the new tables 

 made use of recent obser\'ations, and introduced two em- 

 pirical terms in the endeavour to represent the observations. 



Professor Brown notes that the lunar methods give 

 1 /294 for the Earth's compression, while pendulum methods 

 give 1 /297. He suggests photographic observations of the 

 Moon at new stations to settle the question. Such observ- 

 ations, made at several points on the Equator, should also 

 decide whether it [the Equator] is a circle or an ellipse 

 with its axes differing by a mile or t^vo. 



Professor Brown adds some interesting information on 

 the progress of the construction and printing of his tables. 

 There are o\'er one thousand terms in his expressions for 

 the longitude, latitude, and parallax, but, o%ving to the care 

 taken in combining as many terms as possible into a single 

 table, the actual computation of a position will take very 

 little longer than with Hansen's tables, w-hich included only 

 three hundred terms. The new tables will be used for the 

 first time in The Nautical Almanac for 1919, as the 

 places there have to be computed several years in advance, 

 and 1918 is already computed, 



FUTURE TOTAL ECLIPSES.— O. M. Mitchell's " Orbs 

 of Heaven " has the following fine passage on the w-onderful 

 power which the knowiedge of gravitation gives to the mind 

 of the astronomer : " With resistless energy it rolls back 

 the tide of time, and lives in the configuration of rolhng 

 worlds a thousand years ago ; or, more wonderful, it sw-eeps 

 aw-ay the dark curtain from the future, and beholds those 

 celestial scenes which shall greet the vision of generations 

 when a thousand years shall have rolled away, breaking 

 their noiseless waves on the dim shores of eternitj'." 



Oppolzer's " Canon of Eclipses " is a good instance of 

 such calculations. It contains the details of all eclipses 

 from the j'ear —1207 to +2161, and has proved of vast 

 service to students of chronology. The research was pushed 

 further into the future by Rev. S. J. Johnson in " Eclipses 

 Past and Future." He carried the examination of eclipses 

 visible in Great Britain to the year 2491. A still greater 

 extension has been pubUshed in the last few months by 

 O. Schrader, who gives full particulars of all eclipses that 

 are large in Central Europe from 2166 to 3045. It is now 

 possible to form a trustworthy list of the totalities in the 

 British Isles for the next eleven hundred and fifty years. 

 I think it may be guaranteed that no eclipses are omitted 

 from Table 5 that could possibly be total in our 

 islands — -at least as far as the year 2290. Beyond that 

 point it may not be quite exhaustive, but all the favour- 

 able eclipses are included. I give enough data to draw 

 the tracks on a map. The altitude of the Sun can be 

 estimated from the time of day. The duration of totality 

 is given, but where this is given as an exact number of 

 minutes it must be taken as approximate only. 



Tills makes twenly-six totalities in one thousand one 

 hundred and fifty years, but two or three of them may not 

 quite reach our shores. Mr. Maguire [Monthly Notices of 

 R.A.S., Volume XLV, page 400) gives a similar list of 

 totalities for the past thousand years. He makes the 

 number thirteen. Probably one should be added, namely, 

 968, December 22nd (Johnson). The barren interval of 

 two hundred and three years from 1724 to 1927 is much 

 the longest in the w^hole two thousand two hundred years ; 

 the next longest is one hundred and forty-five years (1185 to 

 1330). The average interval between British totaUties is 

 about fifty- five years. In the tr^venty-second century 

 there are at least six, possibly seven, British totalities in 

 a period of sixty-seven years. Edinburgh is highly favoured, 

 having had five totaUties in the last one thousand years, 

 and with at least four in the coming period. London had 

 none between 878 and 1715. It will possibly have totality 

 in 2151, 2600, 2864 ; but in each case it will be a good w^ay 

 from the central line, according to the tables used. 



I made a rough calculation some years ago that any given 

 spot has totalitv about three times in one thousand years, 

 or seven times in tw-o thousand two hundred years. This 

 is fairly borne out by these figures, Edinburgh having 

 distinctly more than its share, London distinctly less. 



It is interesting to note that in 3045 periheUon, which 

 now happens about January 2nd, will have changed to 

 January 21st. The obliquity of the ecliptic, now 23° 27', 

 will have diminished to 23° 18'. 



The author calls attention to the curious partiality for 

 Wedne^iday in the mid-European totaUties of the next 

 twelve hundred years. The fact that the 300-year cycle 

 and the 521 -year cycle are both exact numbers of weeks 

 partly accounts for particular weekdays being favoured. 

 The table affords many instances of these cvcles : thus 

 there are European totalities in 1842, 2142, 2442 ; in 1927, 

 2227 ; in 1945, 2245, 2545 ; in 2081, 2381. 2681 (aU total 

 in our islands) ; in 2126, 2426. 2726, 3026 ; and so on. 

 The 521 -year cycle gives European totalities in 1860, 2381, 

 2902, and so on. The triple Saros is well illustrated by 

 2081, 2135, 2189 (aU total in our islands). The 521-year 

 cycle brings back the latitude of eclipse tracks with great 

 accuracy ; but there is often a considerable shift in longitude, 

 also in the diameter of the Moon, so that a total eclipse 

 may be succeeded by an annular. It will probably surprise 



