322 NATIOXAL GEOGRAPHIC SOCIETY'S ECLIPSE EXPEDITION 



As the second contact or beginning of totality aj)i)roachecl, the 

 observer was struck b}' the clearness with which a long arc of the 

 chroinos-phere came out on each side of the diminishing arc of the 

 sun's limb some seconds before totalit}'. 'I'lie observation of the be- 

 ginning of totalit}' was very satisfactory, the distinct arc of the sun's 

 liinb contracting slowly at first, then more rapidly, until it at length 

 vanished at the following moment: 



Uis. Mill. Sec. 



Contact 2. Watch time 8 .')2 54..") 



Corr. of watch .. — 2S.:i 



Standard time 8 52 2(1.2 



Given wich time 1 52 20.2 



Hrs^. .'Mill. Sec. 



Contacts. Watch time 8 54 22 rb 1 



Corr. of watch. . . — 28.;; 



Standai'd time 8 5."5 5."!. 7 



Duration of totality, 1 min. 27.5 sees. 



The observation of contact 3 was less certain than that of con- 

 tact 2, tiie doul)t Ijeing ±: 1 second. 



The tabuhir times of contact for Norfolk as given in the American 

 Eit/iemeris for 1900, with the corrections thus derived from o'oserva- 

 tion, are : 



Tahiilar time. Corr. 



Contact 2 8 In-s. 52 min. 32.6 sec. — (5.4 sec. 



Contact 3 53 59.1 — 5.4 



Duration 1 26.5 +1.0 



PHOTOGKAPHS 



The })hotograj)hs were taken by Messrs \V J McGee and \V. H. 

 Holmes with a large camera of 6 centimeters aperture and about 50 

 centimeters focus. It was fixed upon an inclined table without clock 

 or otiier motion to follow the sun. Three plates were successfull}'' 

 exposed. A fourth was s|)oil('il by the end of totality coming on 

 before it was cut ott'. 



These })lates were specially ))repared for the use of the expedition 

 by Professors E. C. and \V. H. Pickering, of the Harvard Observatory, 

 and were sent to that observatory for development, examination, and 

 report. The re()ort is not expected in time I'ov this nuinljer. 



SHADOW BANDS 



Many unexplained optical phenomena have been described in con- 

 nection with total eclipses. The irregular, sporadic, and inconsistent 



