284 Scientific Intelligence. 



International Union for Solar Research, to take as secondary stand- 

 ards of wave-length the means of the data obtained by interfer- 

 ometer methods at three independent laboratories. The radiations 

 were to be obtained from the iron arc under specified condi- 

 tions, and the primary standards were to be the three cadmium 

 lines which bad been measured with an extremely high degree of 

 accuracy, first by Michelson and later by Benoit, Fabry, and 

 Perot. Fabry and Bnisson, at Meudon, determined the wave- 

 lengths of 115 lines (chiefly due to iron), from A 2373*737 to 

 A 6494*994. Eversheim at Bonn and Pfund at Johns Hopkins 

 took up independently the same problem. The final means of the 

 three sets of determinations were tabulated in the Astrophysical 

 Journal for October, 1910. These means comprised 49 lines from 

 A 4282-408 to A 6494*993 inclusive. At this time the wave-lengths 

 of the ultra-violet lines as determined by Fabry and Buisson had 

 not been verified by Eversheim and Pfund. 



Very recently Eveesheim has published the wave-lengths of 

 26 iron lines from A 3370-787 to A 4282-408. In order to work in 

 the ultra-violet the glass portions of the apparatus were replaced 

 by quartz. The interferometer plates were obtained by subject- 

 ing the quartz to a special process of fusion and pressing. A 

 noteworthy improvement in the interferometer proper consisted 

 in making the emergence " plate" in the form of a plano-convex 

 lens, which was half-silvered on the plane surface, of course. 

 This device did away with one lens and minimized several pos- 

 sible sources of error. Another improvement was made in the 

 silvering process. Eversheim says that the receipt was sent to 

 him in a private communication from Pfund. The details are 

 given in the paper under consideration. The cadmium and iron 

 radiations were photographed simultaneously. Finally, the cor- 

 rections for the variations with wave-length of the change of 

 phase on reflection were very carefully determined by studying 

 the interference curves of equal thickness, so-called. 



The extremes of the probable errors of the newly measured 

 lines are given as 0-0002A and O-OOIOA. The greatest arithmet- 

 ical difference between the wave-lengths of the same lines as 

 determined by Fabry and Buisson and by Eversheim amounts to 

 0*003 A, while the simple average of the numerical differences only 

 equals 0-0014A. Five wave-lengths have identically the same 

 values. The interferometer method seems, therefore, to be about 

 as reliable in the ultra-violet as in the visible spectrum. Evers- 

 heim expresses the hope of extending his table up to A 2300 as 

 soon as possible. — Ann. d. jPhys., No. 15, 1911. h. s. u. 



11. The Sun's Energy -Spectrum and Temperature. — The distri- 

 bution of energy in the solar spectrum outside of the earth's 

 atmosphere has been deduced by C. G. Abbot from the spectro- 

 bolometric observations made at Washington, Mount Wilson, and 

 Mount Whitney during the interval 1903 to 1910. The corre- 

 sponding means obtained at each of these three stations agree 

 with one another as well as do the constituent data used to calcu- 



