ON RADIOTELEGRAPHIC INVESTIGATIONS. 41 



the eclipse finished at sunset near Madagascar. The effects of the 

 moving shadow were investigated under three heads: — 



(1) Strays. 



(2) Signals not crossing the denser parts of the shadow. 



(3) Signals crossing through or near the umbra. 



Strays. 



These were bad on the day of the eclipse and on the preceding day 

 in Europe, North America, and in temperate latitudes on the Atlantic 

 Ocean. They were very few in Central and South America and in the 

 central equatorial Atlantic. In Central America the conditions were 

 exceptional meteorologically, the day having less rain than nearly every 

 day of the preceding three weeks. The preliminary survey ol the 

 results recorded throughout the part of the globe reaching from Con- 

 stantinople to Eio Janeiro suggests that there was no outstanding 

 occurrence in regard to frequency or intensity of strays that could 

 be directly ascribed to the passage of the shadow. 



Signals not traversing the cleanse shad'OW. 



Many observations were made in Northern Europe and America 

 on the signals from the Azores, which were arc signals of 4,700 metres 

 wave-length. The observing points extended from Berhn through 

 Holland, France, Italy, Spain, and Great Britain to stations near the 

 Atlantic coast of the United States. There were no unusual variations 

 in the strength of the signals from the Azores. 



Anothei; class of experiment comes under this heading. It was 

 suggested by the effect sometimes observed at sunset or sunrise, in 

 which the twilight band when on one side of a transmitting station 

 appears to strengthen as if by reflection the waves received at a station 

 on the other side of the transmitting station. In order to test whether 

 such reflections oocm-red during an eclipse certain stations on the 

 south of the central line of the eclipse were asked to listen to Ascension, 

 which was also south of the central line. The stations at Durban and 

 Port NoUoth (South- West Africa.) found no trace of the effect, and in 

 fact the former concluded that the signals from Ascension were rather 

 worse after the eclipse began. An analogous experiment on the 

 northern side was carried out by one o^f the Malta stations and also at 

 Rosyth, listening to Cairo, with similar conclusions. 



Effect on Signals passing across the Central Line. 



Arrangements were made for the transmission of signals from the 

 Darien station of the Panama Canal zone, and several stations in South 

 America attempted to receive the signals. The report from the Falkland 

 Islands has not yet come to hand, and the other stations in South 

 America did not succeed in picking up the signals. The only observa- 

 tion made on the earlier stag:es of the eclipse are those of Demerara 

 listening to Ascension. Fluctuations in signal strength are repoi'ted, 

 but no steady increase or decrease in strength. Ships at sea within 

 the penumbra report a strengthening of all signals during the eclipse. 



