94 Mr. J. C. McConnel on the 



hidden on November 27 by a wbite mist, and in January by 

 both ice and snow. I quote the following readings : — 



G. 



27. 



11.6 A.M. 



381° 



•345 



1-06 





11.30 



40 



•326 



•97 



4. 



12.37 



421 



•301 



•86 



17. 



1.0 P.M. 



41 



•318 



•93 



Nov, 

 Jan. 



3. The polarization at different points, distant 90° from the 

 sun, is affected by the brightness of the ground below. The 

 following successive observations were taken on a cloudless 

 day : — 



October 21. 

 G. r. s. 

 9.0 A.M. 53° -209 ^53 At the highest point. 

 551 -191 -47 Near hills to the right. 

 511 -221 -57 At the highest point. 

 9.30 531 ^205 -51 Near hills to the right. 



The altitude of the points observed in the second and fourth 

 of these readings may have been about 5°. In this instance 

 the polarization was materially stronger near the horizon. 

 This was due to the instrument being pointed down the 

 long straight valley, which was nearly clear of snow and 

 was bordered by dark Avoods. Later on in the same day this 

 general principle led to an opposite result. 





October 21. 



G. 



r. s. 



501° 



•228 -59 



471 



•257 ^69 



48 



•249 ^67 



12.55 P.M. 501° -228 ^59 At the highest point. 



Near hills to the left. 

 1.10 48 ^249 ^67 Near hills to the right. 



The weaker polarization in the two last observations was due 

 to the line of sight passing over neighbouring and far- 

 reaching tracts of snow. On the following day similar 

 results were obtained. 



October 22. 

 G. r. s. 



8.55 A.M. 541° ^199 ^50 Near hills to the right. 

 9.0 521 .215 -55 At the highest point. 



9.7 541 ^199 -50 Near hills to the right. 



9.15 511 -221 -57 At the highest point. 



4. The formation of clouds seemed to be heralded by a 



