May 3, 1906] 



NA TURE 



Eg-yptians conceived as inhabiting- the tomb-world ; 

 and occasionaliv sarcophagi were ornamented in tlie 

 same manner. Some of the best illustrations in Dr. 

 Budge's book are taken from the sculptures of Seti's 

 sarcophagus. 



The conceptions of the rewards and punishments 



C-r^fCliUil 



of the dead in the next world as given in these two 

 books are also well worth the attention of the 

 anthropologist. 



ANCIENT ECLIPSES. 

 H"" HE results of recent discussion of ancient eclipses 

 "^ may for convenience be divided into three sec- 

 tions. The conclusion of each section depends upon 

 the truth of the conclusions of the preceding sections, 

 but not vice versa, that is to say, the results of the 

 last section may be rejected without in the least im- 

 pairing the validity of the earlier conclusions. The 

 results are as follows : — 



(i) If an astronomer had been asked a year ago by 



NO. T905, VOL. 74] 



a historian or a chronologist whether the tables of 

 the sun and moon accurately accounted for the re- 

 corded phenomena of ancient eclipses, he could only 

 have replied that the tables failed altogether lo account 

 for the solar eclipses ; that they had been empirically 

 altered so as to account for the observed times of 

 certain lunar eclipses; and that the question whether 

 the tables so altered accounted for the magnitudes oi 

 the same lunar eclipses had not even been examined. 

 There seemed to be no possible modification of the 

 tables that would bring them into harmony with the 

 recorded solar eclipses, and it was therefore the 

 received opinion that the historical accounts of these 

 were untrustworthy. The first result is that two slight 

 modifications of the existing tables will cause them to 

 satisfy the records. 



The modifications in question may be stated as 

 follows : — Define the nodical month as the mean 

 period between one passage of the moon from south 

 to north of the ecliptic and the next passage, and 

 define the nodical year as the mean period between 

 one passage of the sun from south to north of the 

 plane of the moon's orbit and the next passage, purely 

 periodic variations being left out of account. Then 

 the eclipses show that the rale of change of length 

 of both the nodical month and nodical year as given 

 in the tables must be altered. 



(2) The second section of the results is concerned 

 with the question, " In order to alter the rate of vari- 

 ation of the nodical year, are we to alter the acceler- 

 ation of the node or of the sun? " Now the motion 

 of the node depends upon theory, and the same theory 

 which accounts for its motion at the present time 

 will suffice to calculate its motion at any time during 

 the last few centuries. The motion of the sun, how- 

 ever, is purely a question of observation. Unknown 

 causes mav easily be conceived as altering its motion. 

 The second result is therefore to ascribe an acceler- 

 ation to the sun's motion to account for the variation 

 in the nodical year inferred from ancient eclipses, or 

 in other words, we may leave out the word " nodical " 

 in our statement and say, " The ancient eclipses 

 indicate certain definite rates of change in the lengths 

 of the month and year." 



(3) We lastlv require some physical explanation of 

 the sun's acceleration. Here there are many possi- 

 bilities. The aether mav offer a sensible resistance to 

 the passage of the earth; or an electro-magnetic 

 theory of gravitation may compel us to take account 

 of the small, but not infinitesimal, ratio between the 

 velocity of a planet in its orbit and the velocity of 

 light; or again, electrical theories of matter somewhat 

 modify the old conception of mass, and with it the 

 fundamental equations of motion on which planetary 

 theory rests. But the explanation tentatively put 

 forward at the April meeting of the Royal Astro- 

 nomical Societv is as follows :— Let us suppose the 

 acceleration of 'the sun to be due to a change in the 

 length of the dav caused by tidal friction. Tlie tides 

 check the rotation of the earth, lengthen the day, and 

 therefore apparentlv increase all diurnal movements 

 by the same fraction of their whole amounts. Intro- 

 ducing numbers for greater definiteness, let us suppose 

 that in a centurv the day increases in length by a 

 two-hundredth part of a second of time. Then in a 

 century the sun's apparent rate of motion will increase 

 by one part in seventeen million, which is exactly the 

 change indicated bv the eclipses. If, however, the 

 moon's apparent rate of motion also increased by one 

 part in seventeen million the acceleration would be ten 

 times larger than that indicated by the eclipses. 



But if "the tides are checking the diurnal rotation 

 of the earth, it follows from the principle of conserva- 



