Dec. 26, 1878] 



NATURE 



i^y 



the motion of our satellite could be followed by their 

 means with the same accuracy as that of the other 

 heavenly bodies, after having been made the subject of 

 astronomical and mathematical research for two thousand 

 years. This expectation was soon proved to be far from 

 borne out. Prof. Xewcomb showed in 1870 that the 

 accuracy of the Tables since 1750 had been secured only 

 by sacrificing the agreement with observations previous 

 to that epoch, and that about 1700 the Tables deviated 

 more widely from the observations than the previous ones; 

 and it may be added that those who had been engaged in 

 examining the old eclipses were aware that Hansen's 

 Tables did not represent the phenomena only as far 

 back as the commencement of the eighteenth centun-, so 

 well as Burckhardt's or Damoiseau's, which had been 

 used for our ephemerides up to the date of their publi- 

 cation. 



Anew investigation of the subject for the purpose, if 

 possible, of ascertaining the cause of these unexpected 

 deviations was entered upon at the Naval Observator}-, 

 Washington, and was made a part of our author's official 

 duties in that establishment. In the present volume we have 

 the results of researches on the discordances in question, 

 based upon obser\ations before the year 1 750. This portion 

 of the work was originally intended to follow the study of the 

 mathematical theory- of the inequalities of long period in 

 the moon's mean motion, but for reasons explained in the 

 Introduction, that part of the inquiry is still incomplete 

 The year 1750 was fixed upon as the terminal point iii 

 the mvestigation in this first part, as it is the epoch when 

 exact meridian observations commenced, and that which 

 separates the period within which we are in possession of 

 obser\-ations reduced on modern data, from the period 

 durmg which neither really accessible obser^•ations nor 

 tables of reduction are available. 



Prof. Xewcomb supplies an historical introduction in 

 which the discovery of the secular acceleration, and the 

 exammations of ancient eclipses with the view to fix its 

 correct amount by obser%-ation are briefly noticed, as also 

 Ferrer s paper, published in 1S53, containing "the first 

 known attempt to calculate from theory the retardation 

 produced by the action of the moon upon the tidal wave " 

 and the researches of Adams and Delaunav. He pro- 

 ceeds to give a summary of the data now at our disposal 

 for determining the apparent secular acceleration of the 

 moon from observation alone. These include the state- 

 ments of ancient authors from which it has been inferred 

 that total solar eclipses have been wtnessed at certain 

 pomts of the earth's surface at dates approximately indi- 

 cated, and the author points out the uncertainty attending 

 our mterpretation of such records. He considers that th! 

 circumstance which we should regard as most unequivo- 

 cally markmg the totality of an eclipse is the 'Msibility of 

 stars, though he thinks that even this criterion is hardly 

 to be admitted as conclusive, because Venus may be seen 

 durmg a considerable partial or an annular eclipse, and 

 at certain times when there is no eclipse at all, there is 

 also another difficulty in some cases, in determining the 

 precise localities where the phenomena were obser^'ed. 

 we have also the series of lunar eclipses upon which 

 ^to.emy founded his theory and which are recorded bv 

 him m the Almagest. These are followed by the obser- 

 vations of the Arabian astronomers, chiefly contained in 



an Arabic manuscript belonging to the University of 

 Leyden, a translation of which was made by Caussin and 

 published by the French Government in 1804, under the 

 title " Le Livre de la Grande Table Hake'mite." Prof. 

 Xewcomb remarks that this work contains what are 

 entitled to be considered the earliest astronomical obser- 

 vations of eclipses which have reached us, for although 

 some of the data furnished by Ptolemy, Theon, Alba- 

 tegnuis, and others, may have been the results of astro- 

 nomical obser%'ations, in no case have the quantities 

 actually observed been handed down to us. The entire 

 number of eclipses in this collection is twenty-eight, and 

 the times of concluded beginning and ending were usually 

 determined by noting the ahitudes, which were recorded 

 sometimes in whole degrees only, at others " in coarse 

 fractions of a degree." There must remain a doubt how 

 nearly such times apply to those of actual contact, but 

 Prof. Xewcomb suggests in this part of his work that by 

 the mean of all the observed times the error in the moon's 

 mean longitude can be reduced to not more than a minute 

 of arc. The Arabian observations are followed by those 

 of European obser\'ers prior to the invention of the tele- 

 scope, including Regiomontanus, Bernard Walther, and 

 Tycho Brahe : of the latter, the author remarks, " It is 

 wonderful if so indefatigable an obser\-er never observed 

 an occultation of a star or planet by the moon, yet I hare 

 never succeeded in finding any such ; " he made a careful 

 examination of Tycho' s obser\'ations during periods in 

 which the bright star Aldebaran must have been occulted 

 to no purpose. The obser\-ation of eclipses and occulta- 

 tions with the aid of a telescope. Prof. Xewcomb remarks, 

 may be considered as commencing with Bullialdus and' 

 Gassendus, but they had no clock, and only fixed the 

 time by noting the altitude of the sun or a star. The 

 application of the clock commences with Hevelius, and in 

 the scarce volume of his " Machina Ccelestis" are found 

 a number of occultations thus obser\-ed. Then foUow the 

 obsenations of Flamsteed and the astronomers at the 

 obser^-ator)' of Paris, the Cassinis, La Hire, and Delisle, 

 the latter of whom also observed at St. Petersburg. 

 Prof. Xewcomb, during a visit to Paris, while Delaunav 

 was in charge of the Obser\-atory, was fortunate in having 

 all the archives of that estabhshment unreservedly placed 

 at his disposal. He found amongst them most of the 

 original note-books of the French observers since the 

 year 1675, in which were contained a great number of 

 occultations that had been quite forgotten, those which 

 had appeared in the Memoirs of the Academy, forming 

 but a smaU fraction of the whole. Agam, on visiting 

 Pulkowa, M. Struve gave him access to the records of 

 Delisle's observations, 1727 to 1747, forming a useful 

 supplement to those of Paris, which had diminished in 

 number after 1720. 



Prof. Xewcomb concludes this historical notice with 

 remarks on observations since the time of Bradley, and 

 granting certain fundamental premises, suggests that the 

 secular acceleration of the moon may admit of neariy 

 as accurate determination from the modem observations, 

 as from their combination wth the ancient ones. 



Having thus briefly recapitulated the data available for 

 investigation, the author adverts to ancient eclipses, pre- 

 sumed to have been total from the narrative of the 

 historians, at certain points of the earth's surface. The 



