362 PROCEEDINGS OP THE AMERICAN ACADEMY 



Mean cloudiness, 1.7. 

 Number of rainy days, one in five years. 

 Number of totally cloudy days, one in five years. 

 Number of wholly clear days, 21 each May. 



Up the Nile, as far as El-Akhmym, the conditions would be a little 

 less favorable. With Bagdad and Teheran in the month of May, the 

 case stands very similar. 



There are no readily available records with regard to the cloudiness 

 of Shanghai. 



(4) The duration of totality of the eclipse of 1882 is not very 

 favorable. At El-Akhmym totality lasts one minute and twelve 

 seconds. At Teheran the eclipse is total something like one minute 

 and two tenths. And at a point a few miles to the north of Shanghai, 

 the duration of totality is nearly half a minute ; but this short interval 

 would suffice for the purpose of verification simply. 



We may examine the conditions of applicability of the telegraphic 

 method to total eclipses subsequent to that of 1882 : — 



a. — The total eclipse of 1883, May 6. The path of totality lies 

 almost entirely on water, beginning to the east of Australia, and 

 going between New Caledonia and New Zealand, and leaving the 

 earth to the west of South America. In the longitude of the Mar- 

 quesas Islands, the duration of totality is about five minutes and fifty 

 seconds, but there is no other station where the eclipse will be observ- 

 able on land. 



b. — The total eclipse of 1885, September 8. In the longitude of 

 Wellington, New Zealand (the south point of the north island), the 

 duration of totality is one minute and fifty-four seconds. It does not 

 appear, however, that telegraphic communication with any other sta- 

 tion will be practicable. 



c. — The total eclipse of 1886, August 29. Here again we have an 

 ocean-track except for a few miles at the beginning, and a small por- 

 tion at the end. In Grenada the duration of totality is three minutes 

 and fifteen seconds ; on the Atlantic Ocean, off the coast of Liberia, 

 six minutes and thirty seconds ; and on the western coast of Africa 

 (latitude 12° south), about four minutes and forty-five seconds. 

 The time of the eclipse, however, is the storm-season in the AVest 

 Indies ; so that we are not warranted in expecting very much of this 

 eclipse, notwithstanding the extraordinary duration of totality. 



d. — The total eclipse of 1887, August 18. The totality-path of this 

 eclipse begins in central Germany, and passes to the north of Moscow, 



