178 



SCIENCE. 



[N.S. Vol. XVII. No. 422. 



the spectrum upon dilution, unless some 

 change in the solution occurs. The strongly 

 absorbing region of these solutions shows 

 five bands. Ostwald shows that twelve 

 permanganates in dilute solution show 

 identical positions for four of these bands, 

 which suggests at once identical color for 

 common ions. Indeed, Ostwald gives a 

 large series of solutions of different com- 

 mon ions to support this conclusion. Still, 

 it must be readily recognized that the color 

 ■of a solution is determined by the magni- 

 tude of absorption, both inside and outside 

 the absorption band, as well as by the posi- 

 tion of the bands. This determination re- 

 •quires a spectrophotometric study, although 

 it is a tediously slow process in comparison 

 to the other method. Spectrophotometrie- 

 ally studied those two permanganates show 

 that the bands are identical for both sub- 

 stances in all concentrations. For the 

 potassium permanganate the relative trans- 

 parency in the band region increases 

 slightly upon dilution. The zinc perman- 

 ganate remains constant for all concentra- 

 tions in this region. Outside the charac- 

 teristic absorption bands, in both blue and 

 red, both solutions show marked increase 

 in relative absorption upon dilution. That 

 is, increased ionization has caused a change 

 outside the bands, not in the band region 

 itself. 



Note : Even in concentrated solutions, 

 permanganates would have a large disso- 

 ciation coefficient, hence a small difference 

 in ionization could only be realized upon 

 great dilution. Owing to the slight solu- 

 bility of several permanganates, one is still 

 farther restricted in the choice of sub- 

 stances. Hence so far I have only been 

 able to examine the two substances. 



The Magnetic Botary Dispersion of Solu- 

 tions of Anomalous Dispersive Sub- 

 stances: F. J. Bates, University of 

 Nebraska. Presented by D. B. Beace. 



The rotation of the plane of polarization 

 of a ray of light, when passed through a 

 substance in a direction parallel to the lines 

 of force, has been found on theoretical 

 grounds to be proportional to du/d^-, where 

 u is the index of refraction of the sub- 

 stance for the wave-length A. Conse- 

 quently in solutions showing anomalous 

 dispersion there should be an anomaly in 

 this rotation wherever there is an anomaly 

 in the refractive index. The author has 

 studied very dilute solutions of fuchsin, 

 cyanin, analine (blue) and litmus with an 

 improved form of polariscope. The mean 

 error of a setting for any wave-length 

 was less than .01° ; while the best results 

 claimed by previous investigators, who ob- 

 tained anomalous effects, is a probable 

 error of .03°. 



The first observations indicated that the 

 apparent anomalies were present in these 

 solutions ; but further investigation proved 

 them to be spurious. After eliminating 

 these effects no anomalies were obtained. 

 Hence, although anomalous dispersive sub- 

 stances may possess an anomalous Faraday 

 effect, its magnitude is much less than it 

 has heretofore been considered. 



The Investigation of the Atmospheric Cir- 

 culation in the Tropics: A. La^veence 

 RoTCH, Blue Hill Meteorological Ob- 

 servatory. 



It is generally believed that the currents 

 which ascend from the thermal equator 

 proceed immediately as southwest and 

 northwest anti-trades over the northeast 

 and southeast trades-winds, and that the 

 greater part of the anti-trade descends to 

 the surface of the ocean north and south 

 of the trades and continues to the poles 

 as the prevailing southwest or northwest 

 winds of the north or south temperate 

 zones. This hypothesis is not sustained by 

 the observations of the movements of vol- 

 canic dust and of upper clouds, which indi- 



