Polarization of Sky Light. 93 



October 22. 



G. r. s. 



3.36 P.M. 60i° -168 -38 Near the hill to the left. 

 3.40 59| -165 -40 „ „ „ 



At this time the polarization at the highest point was too 

 strong to be measured. 



After this I seldom took readings except near the middle 

 of the day. 



2. The effect of the ground being covered with snow is to 

 materially diminish the polarization. Thus add to the readings 

 just quoted the following : — 



October 22. 9.0 a.m. 52^° -215 '55 

 9.15 51i -221 -57 



and then compare them with the readings a few days later 

 after five inches of snow had fallen : — 



October 26. 10.15 a.m. 41° '318 -93 



On this day there were a few small clouds near the horizon, 

 not enough to produce any appreciable effect. 



The hot sun swept away the fresh snow very quickly, and 

 by the next morning the greater part of it had gone. 



October 27. 11.40 a.m. 47f° -251 -67 



Again, after some bad weather, during which three feet of 

 snow fell, came a fairly good day. 



November 6. 10.0 a.m. 37° -366 l-i6 

 10.6 36i -377 1-21 



The point of observation was perfectly clear, though there 

 were a few scattered clouds about. An hour later the sky 

 was nearly covered with clouds, so the weak polarization on 

 this day may be partly attributed to that circumstance 

 (see 4). I will therefore cite also : — 



Ct. r. s. 



November 13. 12.50 A.M. 41° -318 -93 

 12.57 42^ -301 -86 



The sky was cloudless at the time of observation and nearly 

 so all day. The white covering on the ground was more 

 broken than on the 6 th, and the snow had gone from the 

 trees. 



After this date the snow covering was practically perfect 

 except for the dark trees. The previously dark lake was 



