ON RADIOTELEGRAPHIC INVESTIGATIONS. 



43 



station of the Eastern Telegraph Company, and the Sierra Leone 

 station of the African Direct Telegraph Company. There is evidence that 

 on certain occasions the same cause is affecting both stations though 

 they are separated by 2,500 miles, mostly across mountain and desert. 



As regards disturbances observed at night-time, these are also fre- 

 quently very local and due to convective weather, but there is probably 

 a greater proportion of non-local storms than appears in the day records. 

 In this connection may be noted a report from Australian stations that 

 the worst and most continuous type of disturbance (apart from local 

 thunder storms) occurs on calm nights when the sky is blue and starry. 



As a contrast to the prevalence of strays during convective weather 

 may be instanced the fact, reported by Lieut. E. R. Macpherson from 

 Sierra Leone, that a very dry wind which blows periodically for several 

 days on the West Coast of Africa causes an almost complete cessation of 

 strays immediately it starts and allows of their resumption immediately 

 it stops. On the other hand, the monsoon period on the same coast is 

 one marked by great X storms. 



Correlation of Records at Distant Stations. 



The daily records of strays received at the above-mentioned stations 

 in Malta and Sierra Leone have been examined carefully for the period 

 August 1914 to May 1915 inclusive. This period has been treated in 

 four sections of two and a half months each. It will be sufficient to 



give the following figures, which 

 2 A.M., Greenwich mean time: — 



refer to the night hours 10 p.m. to 



M. indicates Malta, S. L. indicates Siena Leone. 



Mean M. . . . 



Probable error 

 Mean S. L. 



Probable error 

 Standard Deviation, M 



„ ,, S. L. 



Correlation Coefficient 



Probable error 



Graphic Records. 



Many observers have made for the Committee precise observations 

 of individual strays by making, on lines graduated to represent time, 

 marks corresponding to each stray as heard in the telephones, the zero 

 of time being fixed by aid of radiotelegraphic time signals within range of 

 the obsei-ver. Comparison of the records made in the British Isles has 

 shown that on an average night many of tlie strong strays are heard by 

 all the observers, and on days free fi'om X storms the same remark 

 applies. Coincidences have also been noticed between pairs of American 

 stations not too widely separated. The analysis for very distant 

 stations has not yet been carried out except for a veiy few in Europe. 

 For example, in the month of June 1914, coincidences have been traced 

 in the strays heard at Southampton and Dresden, Gibraltar and Dresden, 



