December 21, 1906.] 



SCIENCE. 



819 



f-cteristics of the larval fish. These were the 

 first observations made on the breeding Umbra 

 and should be repeated before they are fully 

 accepted. Although many years ago I kept 

 several specimens in a small aquarium, no 

 attempt to breed was noticed, and none has 

 been observed in an aquarium of the U. S. 

 Fish Commission containing a number of 

 them. I therefore call attention to the in- 

 teresting article by Carbonnier. 



Theo. Gill. 



N0TE8 ON PHYSICS. 



THE TUNGSTEN LAMP. 



Many readers of Science may be interested 

 to know that 'the electric lighting industry 

 is face to face with a change of almost revolu- 

 tionary character,' to quote from the con- 

 cluding paragraph of a paper read before the 

 American Institute of Electrical Engineers by 

 Dr. C. H. Sharp, of the Electrical Testing 

 Laboratories of New York City, on Friday 

 evening, November 23. 



The two papers of the evening were, a paper 

 by Dr. C. P. Steinmetz on the general aspects 

 of the problem of the transformation of elec- 

 tric power into light, and one by Dr. Sharp on 

 some tests of new types of incandescent lamps ; 

 and the subject was discussed by several in- 

 vestigators who are working upon the prob- 

 lem of the tungsten lamp in this country. 



It is generally conceded that within a year 

 an electric glow lamp, the tungsten lamp, will 

 be on the market and that the output of light 

 per unit of power consumed will be increased 

 at least threefold above that which is now 

 obtained by the carbon filament glow lamp; 

 which means that the light-producing capacity 

 of every electric lighting station in the world 

 will be at once multiplied by three, and that 

 there will be at once the possibility of greatly 

 reduced prices per unit of light and greatly 

 increased profits to the electric lighting com- 

 panies. 



Those who are interested in the scientific or 

 technical aspects of the problem of electric 

 lighting will find it worth their while to read 

 the papers of Dr. Steinmetz and Dr. Sharp in 

 the forthcoming monthly issue of the Pro- 



ceedings of the American Institute of Elec- 

 trical Engineers. 



NORMAL VERSUS SELECTIVE RADIATION. 

 SELECTIVE EXCITATION. 



To obtain a highly efficient lamp is either 

 to discover a substance which will stand an 

 excessively high temperature under which 

 conditions a very large percentage of the 

 radiant energy is light, or to discover a sub- 

 stance which at a moderately high tempera- 

 ture radiates selectively and gives off a large 

 percentage of luminous radiation. Thus the 

 Welsbach gas light owes its high efficiency 

 very largely to the selective radiation of tho- 

 rium and cerium oxides. 



The idea of selective radiation is, however, 

 profoundly modified in most illuminants and 

 made to depart widely from that form of the 

 idea which is based upon thermodynamics, 

 where the idea grows out of the necessarily 

 complementary character (in a substance 

 nearly in thermal equilibrium) of emission, 

 transparency and reflection.^ 



This modification of the idea of selective 

 radiation is so important in the problem of 

 light production that it should be more gen- 

 erally recognized, and its very intimate con- 

 nection with that principle in the kinetic 

 theory of gases which is known as the prin- 

 ciple of the equi-partition of energy should 

 be pointed out. Indeed, this modification of 

 the idea of selective radiation is intimately 

 connected with the apparent inapplicability 

 of the principle of the equi-partition of 

 energy. 



Jeans has shown that the apparent failure 

 of the principle of the equi-partition of en- 

 ergy in a gas may be explained by the hy- 

 pothesis that when energy is given to a gas 

 in a particular form, say as energy of trans- 

 lational molecular motion, it takes a very long 

 time for this energy to become properly par- 

 titioned among all the possible modes of mo- 

 lecular motion. 



The application of Jeans's idea to the ques- 

 tion of selective radiation is that when energy 



^See Nichols & Franklin's 'Elements of Phy- 

 sics,' Vol. III., chapter on Radiation for an out- 

 line of the argument. 



