414 Prof. Rudberg o« the Variations xvJiich Temperature 



my observations jirove. But, on the other hand, it apjiears 

 singular, that while the dilatation of bodies commonly dimi- 

 nishes tiieir refractive power, the extraordinary raj', though 

 the crystal is dilated in the direction of its axis, becomes never- 

 theless less refracted, and that the refraction of the ordinary 

 r.ay, notwithstanding the contraction of the crystal in a di- 

 rection perpendicular to its axis, does not change, or dimi- 

 nishes if there is any variation. An analogous phoenonienon, 

 however, has already been observed by M. Arago in water *, in 

 which refraction always goes on increasing from the tempera- 

 ture of the maximum density to the point of congelation. 



2. Itock Crystal. — Refracting angle of the prism =45° 20' 5". 

 In both spectra a decrease of the deviation was observed, 

 which was also sensibly the same; viz. 



= 48"-0 



u r 1 • 0-000051 ^ ^^_ ,^, 



whence, on account of the correction = ,7,,^^ -rT tang 28° 15' 



= 6''"0, it follows that the total diminution of the deviations 

 in both spectra is 



= 42 "-0. 



The indices calculated for the ray F become in the extra- 

 ordinary spectrum =1*55868, or 0-00028 less than at the or- 

 dinary temperature; and in the ordinary spectrum, =r549i4, 

 or 0-00026 less. 



3. Arragonite. — The experiments were made with the 

 prisms A, No. 1 ; A, No. 2; B, No. 2; and C, No.2.f 



In all of them I found for the spectrum polarized perpendicular 

 to the axis of crystallization a diminidion, of the deviation pro- 

 duced by an increase of temperature. I also observed a change 

 in the refracting angle of the prisms, with the exception of 

 prism A, No. 1 ; with which, in this respect, on account of the 

 magnitude of the refracting angle, no experiment with the 

 heating apparatus could be made. The following were the 

 observed results : 



A, No. 1. A, No. 2. B, No. 2. C, No. Z. 



Variation of the \ _^i ^<t —V 53" -4' 3" -2' 58" 



deviation ... J 

 Variation of re- f not mea- \ jg,,.^ _ j, ^3,, _^^„,^ 



iractuig angle J surecl. j 



If we correct these values for the deviation of the plate of 

 mica, we obtain 



* See Ediiil). Eiicydopicdia, Art. Expansion, vol. ix. [). 257; aiiu also 

 the observatioiib in the next page. — Edit. 

 ■)■ bee present vohuni, p. 131). — Edit. 



