756 



The analysis of the Chinese wax itself corresponds with the for- 

 mula CiosHios O4, which admits of a simple explanation of the nature 

 of its decompositions ; its decomposition by potash being explained 

 by the equation 



C108 H,o8 O, + KO HO = H,3 O3 + KO) + C54 H56 O^, 



and its decomposition by heat by the equation 



C108 H108 O4 = C54 H54 O4 + C54 H54, 



the substance itself belonging to the class of compound ethers. 



The author announces his intention of following up this paper by 

 a third on the constitution of myricine. 



May 18, 1848. 



The MARQUIS OF NORTHAMPTON, President, in the Chair. 



" On a new case of the Interference of Light." By the Rev. 

 Baden Powell, M.A., F.R.S. &c., Savilian Professor of Geometry in 

 the University of Oxford. 



The principal experiment exhibiting the newly- observed case of 

 interference described by the author, is the following : into a hollow 

 prism or trough containing oil of sassafras, anise, cassia, &c., a plate 

 of glass is dipped so as to intercept the light passing through the 

 upper or thicker half of the prism, and leave the lower half clear : 

 the spectrum of a line of light or narrow slit, is then immediately 

 seen to be covered with dark bands parallel to the slit or edge of the 

 prism. 



Some substances, as oU of turpentine, water, &c,, give no bands 

 with this arrangement, but will give them, if a complementary 

 arrangement be adopted, viz. by placing a narrow slip of glass at 

 the bottom of the prism, or so as to intercept the light passing 

 through the thinner half. 



The production of the bands in general, and the reason of the dif- 

 ference existing in the cases referred to, are fully accounted for on the 

 theory of waves and interference : whence the following formula is 

 deduced, giving the number of bands (n), when /x^ and yu^, the in- 

 dices for any ray for the plate and for the medium, are known, of 

 which X is the wave length ; and taking these quantities for any two 

 rays (1) and (2) of the spectrum, r being the thickness of the plate, 

 we have the number of bands between them expressed by 



The numbers thus calculated agree with observation for many 

 media. The distinction of the two cases alluded to depends on 

 whether n result positive or negative : that is, on the relative values 

 of the indices ; also agreeing with observation. 



