300 PROCEEDINGS OP THE AMERICAN ACADEMY 



surfaces. The photographs, however, can be classified according to 

 intensity in order of numbers as follows, — number 1 indicating the 

 greatest intensity: 1, steel; 2, gold; 3, platinum; 4, palladium; 

 5, silver ; 6, tellurium ; 7, copper. 



It was evident from these experiments that the selective absorption 

 of metals is far less than the absorption exercised by the earth's 

 atmosphere. We therefore resolved to employ the light of the electric 

 spark between metallic terminals, in order to ascertain whether any 

 limit of absorption could be reached. For this purpose, the light of 

 the spark between copper terminals was reflected, by means of a mirror 

 of the metal whose selective absorption we wished to examine, upon 

 the slit of the spectroscope. To protect the surface of the mirror from 

 the effects of the spark, a thin plate of quartz was placed in front of 

 it. It was found that the copper mirror showed no limit of selective 

 absorption by reflection for wave-lengths of light produced by burning 

 copper at the limits of the copper spectrum, that is, at wave-length 

 21UU. The photographic plate taken by this method showed all the 

 lines that the plates showed which were taken by the direct light of the 

 spark unreflected and unabsorbed by any medium. The palladium 

 mirror was substituted for the copper mirror, and also showed no 

 limit of selective absorption above wave-length 2100. We are led to 

 conclude, therefore, that the metallic surface of the speculum metal 

 upon which the lines are ruled which form the diffraction grating 

 does not fix by selective absorption the limit of metallic spectra at 

 1800 to 2100. This limit more likely resides in the materials form- 

 ing the sensitive emulsion with which the sensitive plates are coated. 

 We have found that a marked difference exists in different emulsions 

 in regard to sensitiveness to ultra violet light. The various staining 

 processes, which enhance to such a marked degree the sensitiveness 

 of photographic plates to wave-lengths of greater length, do not seem 

 to affect the limit of metallic spectra in the ultra violet. Thus, plates 

 stained with erythrosine, which are extremely sensitive to yellow 

 and green light, continue to give the same limit in the ultra violet 

 after staining as they did before they were submitted to the staining 

 process. 



