February 20, 1885. 



SCIENCE. 



151 





ful color, the light growing redder the farther 

 it traversed the atmosphere. 



In a recent article x Professor Langley states 

 his belief that much of this diffusion of the 

 blue rays, as also the general absorption of 

 the whole spectrum, is due to fine dust-parti- 

 cles in the atmosphere. The very strong ab- 

 sorption of the blue rays of the arc-light by 

 fog would seem to suggest the inquiry whether 

 the average size of the minute water-drops 

 forming this fog has any thing to do with the 

 remarkably selective effect upon the blue wave- 

 lengths, or whether this is simply the absorp- 

 tion effect of water en masse. 



With the failure of the arc-light to penetrate 

 fog comes the natural inquiry, whether the 

 incandescent lamp will be any better for light- 

 house and search-light purposes. Now, the 

 part of the solar spectrum most free from at- 

 mospheric absorption-lines is in the orange, 

 with part of the neighboring yellow and red ; 

 and some experiments have shown that this re- 

 gion — or the 3"ellow part of it, at any rate — is 

 that in which the incandescent carbon filament 

 is especially rich, relatively more so than the 

 solar spectrum, and it is the brightest part of 

 that. So that there would seem to be every 

 probability that the incandescent lamp would 

 prove very effective in fog penetration, perhaps 

 most efficiently so at a slightly lower tempera- 

 ture and brilliancy than the present average. 

 The difficulty for lighthouse and search-light 

 purposes would be in concentrating a sufficient 

 amount of luminous radiating filaments in a 

 very small space near the focus of a lens or 

 mirror, which is a strong point in the effective 

 use of the arc-light. With single-filament 

 lamps this would be impossible ; but the writer 

 can see no insuperable difficulty in arranging 

 a whole bunch or cluster of interlacing loops, 

 joined in multiple arc within the same ex- 

 hausted globe, so as to present almost a com- 

 plete network of filaments over a vertical 

 projection of an inch or two square, and yet 

 not have them touch each other ; unless, in- 

 deed, the great heat might soften the globe 

 enough to let it collapse ; and this could prob- 

 abh- only be determined by experiment. The 

 suggestion that a slightly lower temperature 

 might be about as effective in fog penetration 

 would help a little, but not very much, on ac- 

 count of the rapid decrease of luminosity, with 

 slight fall in temperature. Special care would 

 need to be taken to make each of the filaments 

 of the cluster of equal resistance with the 

 others ; but no more so than in any set of 

 lamps on the same circuit, and no doubt all 



1 Philosophical magazine, October, 1884. 



the difficulties could be speedily surmounted. 

 Some experiments upon the fog-penetrating 

 power of the incandescent lamp would cer- 

 tainly seem to be worthy the attention of those 

 engaged in these matters ; for there can be no 

 question about the far greater convenience, 

 cleanliness, safety, and reliability, of the in- 

 candescent lamp over all others, even if it is 

 not so economical. But in government light- 

 houses and war-ships the economy is not so im- 

 portant, reliability and fog-penetrating power 

 being the prime requisites. H. M. Paul. 



RECENT DETERMINATIONS OF LON- 

 GITUDE ON THE WEST COAST OF 

 SOUTH AMERICA. 



The recent completion of the longitude 

 measurements on the western coast of South 

 America by the U. S. naval officers, under the 

 command of Lieut. -Commander Charles H. 

 Davis, U.S.N. , affords a remarkable proof of 

 the accuracy of the methods and instruments 

 now in use for such operations. Lieut. -Com- 

 mander Davis commenced his measurement 

 in November, 1883, at Valparaiso, and ter- 

 minated it in March, 1884, at Panama ; con- 

 necting there with the chain of measurements 

 made in 1875 by Lieut. -Commander F. M. 

 Green, U.S.N. , and measuring from Valparaiso 

 to Arica, Arica to Payta, Payta to Panama, 1 

 and in December, 1883, with the aid of Dr. 

 B. A. Gould, director of the Cordoba obser- 

 vatory, from Valparaiso to Cordoba. This 

 work completes the telegraphic measurement 

 of the polygon Washington-Key West, Key 

 West-Havana, Havana-Santiago de Cuba, 

 Santiago-Kingston, Kingston-Aspinwall, As- 

 pin wall-Panama, 2 Panama-Pay ta, Payta- 

 Arica, Arica-Valparaiso, Valparaiso-Cordoba, 

 Cordoba-Buenos Aires, Buenos Aires-Monte- 

 video, Montevideo-Rio de Janeiro, Rio de 

 Janeiro-Bahia, Bahia-Pernambuco, Pernam- 

 buco-St. Vincent, St. Vincent-Madeira, Ma- 

 deira-Lisbon, Lisbon-Greenwich, 3 Greenwich- 

 Washington. 4 



This great chain of longitude measure- 

 ments, consisting of twenty links, closes with 

 but an insignificant discrepancy ; the longi- 

 tude of the Cordoba observatory by way of 

 Lisbon, Rio de Janeiro, and Buenos Aires, 

 being 4 h. 16 m. 48.06 s., and by way of Wash- 



1 Report of the U. S. coast-survey for 1875, appendix No. 11. 



2 Telegraphic longitudes in the West Indies and Central 

 America, Washington, 1877. 



3 Telegraphic longitudes on the east coast of South America, 

 Washington, 1880. 



4 U. S. coast-survey report for 1870. 



