1889.] on Beacon Lights and Fog Signals. 439 



there was still the objectionable crater at the points, whether direct 

 or alternating currents were employed, involving flickering from the 

 incessant shifting of position at the points. A considerable loss of 

 radiant light was also involved, particularly when condensing it 

 optically. The flickering was somewhat reduced by an improvement 

 of Messrs. Siemens, in providing the carbons with a graphite core, 

 but with the increasing powers of currents and in the necessary 

 dimensions of carbons the results were far from satisfactory. With 

 the fluted form of carbon shown on the diagram the formation of the 

 crater is prevented, and the arc is held centrally at the points of the 

 carbons; there is thus, in addition to comparatively steady light, nearly 

 uniform radiation in azimuth, and over a greater vertical angle for 

 optical condensiition. It now appears to me, after some practical 

 experience with this form of carbon, that it is impossible to determine 

 a practical limit to the dimensions of carbons that may be efficiently 

 emj)loyed. With carbons of the actual size shown on the diagram an 

 intensity of about a million candle units should be produced in the 

 arc, and about 150 millions of candle units in the condensed flashes 

 from the optical apparatus of the dimensions now employed for oil 

 and gas flames in lighthouses. Such an intensity is about 400 times 

 that possible at the focus of such aj^paratus with a flame luminary. 

 Such results as these were probably in the mind of Faraday when he 

 reported that " in principle this light may be accumulated to any 

 degree." Flashes of the great intensity here referred to could only 

 be employed in atmosphere impaired for the transmission of light. 

 In clear weather they would be found to be far too dazzling to the 

 eyes of the mariner, while an intensity of about 50,000 candle units 

 is found to be sufficient for his guidance, and in thick fog no possible 

 intensity can be of practical value for navigation. There are, how- 

 ever, various gradations of impaired atmosphere, between clear weather 

 and thick fog, in which the highest available intensity is doubtless 

 desirable at many important landfall stations for obtaining the greatest 

 possible range of visibility. On the other hand, at the majority of 

 stations in narrow waters, the maximum intensity now obtained with 

 flame light is found to be more generally efficient for navigation than 

 higher intensities. 



In 1881 the question of the relative merits of the three light- 

 house illuminants — electricity, gas, and mineral oil — was receiving the 

 attention of the lighthouse authorities of this country, whicli resulted 

 in the Trinity House accepting the responsibility of carrying out an 

 investigation at the South Foreland, of universal importance to the 

 mariner. In the photometrical and electrical portions of this work the 

 Trinity House were aided by the labours of Professor Harold Dixon, 

 F.R.S. and Professor W. Grylls Adams, F.E.S. which contributed 

 very largely to the success of the investigation. The experiments 

 were carried on during a period of over twelve months, and a vast 

 amount of very valuable evidence was collected from numerous 



