640 



SCIENCE. 



[N. S. Vol. III. No. 69. 



no longer problems, though the world waits for 

 a successfal machine to clean the fiber for 

 market. 



There are many hundreds of fiber plants in 

 the world, and the fiber expert is constantly 

 asked to give information concerning the more 

 promising species, not always with a view to 

 cultivation, but often that useless expense in 

 experimentation may be avoided through proper 

 knowledge of their value. The question to be 

 asked in considering a new form of fiber is not 

 ' 'Can we grow the species'?' ' but ' 'What commer- 

 cial fiber will it compete with, or become a sub- 

 stitute for?" With a definite knowledge of the 

 subject, as it relates to the fibers of the world, 

 the expert need never be in doubt regarding the 

 economic value of any species that may be sub- 

 mitted to him for an opinion. 



The commercial fibers represent, in a sense, 

 the survival of the fittest, and until these are 

 crowded ovit by new conditions there is little 

 chance for the other fibers, unless a particular 

 species is found adapted to some new and 

 special use for which the standard forms are 

 not available. 



The second paper was on Geographic Names 

 by Henry Gannett. Bernard R. Green, 



Secretary. 



APRIL 11, 1896. 



Mr. S. p. Langley read a paper on ' More 

 recent observations in the infra-red spectrum.' 



He referred to a communication to the So- 

 ciety more than two years ago, in which the 

 expectation was held out of an early publica- 

 tion of a map of the infra-red spectrum made 

 by the bolometer, and he desired to explain 

 some of the difficulties which had caused its 

 delay. 



It was the misfortune of the astro-physical 

 observatory here that appropriations for its 

 maintenance were made in such a form that a 

 proper building could not be erected in some 

 site free from tremor, and under circumstances 

 providing against local disturbance. As had 

 already been stated in official reports, such local 

 causes had introduced numerous errors in the 

 record, in the form of tremors and oscillations 

 in the photographic trace of the movements of 

 the needle controlled bj' the bolometer, which 



it was almost impossible to exclude in the pre- 

 sent installation. The linear spectra which had 

 been shown here and before the British Asso- 

 ciation were all produced by a nearly automatic 

 process, the minutest line in the spectrum im- 

 plying a corresponding minuteness in the orig- 

 inal curve ; and in this connection he desired 

 to call attention to the statement in a previous 

 report, to the effect that all the minuter details, 

 such as had been shown here and at Oxford, 

 had not been verified ; and to the fact that 

 illustrations of the minuter detail in linear form 

 were given at that time, with the caution that 

 they were presented ' only in illustration,' and 

 were ' not to be treated as a criterion of the 

 final results.' 



The amount of local error is roughly propor- 

 tional to the minuteness of the detail sought. 

 Thus, in the spectrum shown here, and later at 

 Oxford, giving the leading lines discovered by 

 the new method, nearly everything has stood 

 the test of subsequent investigation ; while of 

 the minuter detail in the curves of which a 

 linear translation was then given, in illustration 

 of the process, a large proportion had been sub- 

 sequently found to lie under suspicion. 



The extent to which the character of the 

 work had been influenced by these local condi- 

 tions having been more and more recognized, 

 the labor of the past two years had consisted 

 largely in weeding out errors arising from them, 

 and the process had involved the slow recon- 

 struction or modification of nearly every por- 

 tion of the apparatus, with special reference to 

 the difficulties imposed by the site and the in- 

 sufficient installation. 



Details of the new apparatus were then given 

 with lantern illustrations, particular attention 

 being directed to the introduction of the system 

 of suspending the galvanometer so that ground 

 tremors were not conveyed to it, or were con- 

 veyed in diminished intensity, a change which 

 was stated had been a most essential improve- 

 ment, and which had done away, not entirely, 

 but more than might have been thought pos- 

 sible, with the inconveniences of a site sur- 

 rounded by city traffic. 



Many holographs had been taken during the 

 past year, but onlj- within the past months had 

 the apparatus been brought to such a condition 



