OUR CHIEF TIMEPIECE LOSING TIME. 53 



have swept on in their courses for ages with none 

 to note them. Gradually, observant men began to 

 notice and record the more remarkable phenomena. 

 But such records, made with very insufficient instru- 

 mental means, have in general but little actual value. 

 It has been found easy to confirm them without any 

 special regard to accuracy of calculation. 



Buc there is one class of phenomena which no 

 inaccuracy of observation can very greatly affect. A 

 total eclipse of the sun is an occurrence so remark- 

 able, that (1) it can hardly take place without being 

 recorded, and (2) a very rough record will suffice to 

 determine the particular eclipse referred to. Long 

 intervals elapse between successive total eclipses 

 visible at the same place on the earth's surface ; and 

 even partial eclipses of noteworthy extent occur but 

 seldom at any assigned place. Yery early, therefore, 

 in the history of modern astronomy, the suggestion 

 was made, that eclipses recorded by ancient historians 

 should be calculated retrospectively. An unexpected 

 result rewarded the undertaking: it was found that 

 ancient eclipses could not be fairly accounted for 

 without assigning a slower motion to the moon in 

 long-past ages than she has at present ! 



Here was a difficulty which long puzzled mathe- 

 maticians. One after another was foiled by it. Halley, 

 an English mathematician, had detected the difficulty, 

 but no English mathematician was able to grapple 



