350 



SCIENCE 



[N.S. Vol. XXV. No. 635 



the nicols, since the intermediate colors are 

 elliptically polarized. Colors are white for 

 thick plates and one may note in addition that 

 (i-e — f'o and d are reciprocally related. 



Interference. Diffraction. — For the case of 

 two slits, real or virtual, at a distance c apart 

 and r from the screen, if d is the distance of a 

 lateral dark band from the central fringe for 

 light of wave length A, we may write 



n= (2c/r) • d/\, 



where odd numbers for n determine the posi- 

 tion of the successive minima. Hence if the 

 phase difference in which the rays meet 

 (f, = Trn be introduced, 



0= {Zttc/v) • d/\ 



and the chart is applicable at once, with the 

 underetanding, however, that the dark hori- 

 zontal bars now denote maxima, the dotted 

 bars minima. This, however, is not necessary, 

 for the bars may be moved up by inserting in 

 one of the rays a thickness of lamina t of the 

 refracted index /* (compensator) by the 

 amount 27r(/i — V)t/\, until a minimum re- 

 places the central maximum. The compen- 

 sator thus has the same effect as the rotation 

 of the nicols in the first paragraph. The 

 chart shows the lateral extent of spectra of 

 different orders very clearly, the first having 

 a breadth oi d = the other breadths being 

 proportional to the intercepts of the succes- 

 sive bars between the oblique lines blue and 

 red. The overlapping of these spectra is also 

 well shown. The spectra must be less spread 

 out as the slopes ^irc/rX are steeper. If the 

 slit of the spectroscope be placed at a given 

 distance, as at d ^ a, for instance, the vertical 

 section indicates the channeled spectra which 

 will be observed and the dark lines may be 

 sharp enough to suffice the standardizing the 

 spectrum. 



With certain well-known changes the same 

 remarks apply for most cases of the diffrac- 

 tion of white light. 



Colors of Thin, Films. Ordinary Light. — ^If 

 d denote the thickness of the film or a given 

 section of the wedge of refractive index fx., X 



the wave length of light in air, r the angle of 

 refraction corresponding to the angle of inci- 

 dence i we may write 



=^ Ztt/j. cos r • d/\ 



where minima correspond to even numbers 

 of TT. Hence the positive field of the chart 

 applies at once. The equation as written is 

 primarily useful for the wedge. What the 

 diagram points out very well is the resolution 

 of doublets in the spectra of very high order. 

 If the incidence is normal and the wedge be 

 of air as in the interferometer <f, = 2ird/X. If 

 the thickness is constant and the angle of 

 incidence varies the oblique lines are still ap- 

 plicable if instead of d the quantity cos r be 

 laid off as the abscissa. They are real be- 

 tween r = and the angle of total reflection. 



In practice the oblique lines may be drawn 

 to the scale on a sheet of paper, the black bars 

 on a sheet of celluloid capable of sliding up 

 and down over the former and the vertical 

 lines may be represented by threads movable 

 to right and left over the celluloid. The whole 

 is to be serviceably framed on a sheet of tin 

 plate. 



C. Babus 



Brown University, 

 Providence, R. I. 



THE CAUSES OF THE GLACIAL EPOCH* 



Op the various hypotheses advanced to ac- 

 count for the occurrence of a protracted gla- 

 eiation of a large portion of the earth's sur- 

 face in post-tertiary times, none has thus far 

 met with universal acceptance; partly because 

 of the insufficiency of data regarding critical 

 geographical areas, but mainly because of the 

 feeling that although the causes discussed by 

 Croll, Chamberlin and others are or may be 

 true ones, they are inadequate to account, 

 quantitatively, for all the facts observed. The 

 paleontological evidence of the prevalence of 

 temperate and even semi-tropic floras and 

 faunas in the late Tertiaries within what are 

 now arctic regions, gives such forcible evi- 

 dence of the comparative uniformity of tem- 



' Read at tenth session of the International Geo- 

 logical Congress, Mexico, September 6-14, 1906. 



