2 REPORT~1839. 



compound bands are seen. In all eases the plate must advance /ro»z 

 the blue end of the spectrum ; the other way no bands are formed. 

 Hence, Sir D. Brewster considers the eiFect due to a new and peculiar 

 sort of polarity in the light of the spectrum. 



It is allowed that the bands simply may be accounted for by interfe- 

 rence, but not the polarity. 



The author has repeated and varied these experiments ; and con- 

 ceives that all the facts are easily explicable by the principle of inter- 

 ference, combined with the simplest considerations relative to the un- 

 dulatory course of the rays. 



On fixing an opaque border, about the breadth of the pupil, on one 

 half of the edge of the plate, while the other is left bare, bands are 

 formed by the open part, but not by the opaque. 



The inquiry was extended by using substances of different refractive 

 and dispersive powers, both as plates and prisms, as well as plates of 

 different thicknesses. According to these differences in the retardation, 

 the bands were closer or wider. 



The smallest breadth of any plate (if of the requisite thickness) will 

 act. Hence the effect of the oblique edge is explained by a succession 

 of edges, and was imitated artificially, by combining several plates with 

 their edges slightly overlapping. Each plate gives its own set of bands, 

 and thus compound systems of superposed bands are produced. 



Some plates are found with natural differences of thickness at some 

 points, on simply looking through which bands are seen. 



These phaenomena, as well as those ascribed to polarity, appear per- 

 fectly explicable on the same principle as that applied in a previous 

 communication to the Section, viz. that the two pencils which interfere 

 are the two halves of the pencil of each ray which converges in the eye, 

 and whose breadth is equal to the aperture of the pupil ; the inter- 

 cepted half being that which has passed through the thinner part of 

 the prism, and this part is the least retarded. 



The intercepted part has its retardation nearly equalized with that 

 of the other half of the pencil by the plate ; while the differences in 

 retardation for different rays of the spectrum are successively odd and 

 even multiples of the half wave-length. 



On certain points in the Wave-theory as connected with Elliptic Polar- 

 ization, &c. By Prof. Powell, F.P.S. 



The object of this communication is to draw attention to a discrepancy 

 between certain investigations relative to the theory of undulations, 

 when applied to the case of elliptically polarized light, as deduced 

 from the general equations of motion. 



All the investigations of MM. Cauchy, Kelland, and others, set out 

 from certain general equations of motion, which are then reduced into 

 other forms, and being simplified by the express introduction of the 

 condition that certain terms vanish, are shown to be directly integrable 

 in forms which give the expression for a wave involving the relation 

 which explains the dispersion. 



