342 ' J. E. Willet on Indian Summer. 
south” (toward the Antarctic). If the burnings of the» 
were the cause of the smoky atmosphere, then the smokiness 
should have been more apparent toward the equator ; whereas 
the reverse was so strikingly true, that he abandons his previous 
notions of a connection, between these burnings and the smo- 
kiness, of the nature of cause and effect. 
3. The season when it appears and the period of continuance 
of Indian Summer seem, from the above, to be slightly different 
in different countries. 
. In North America, it occurs in September, October, or Novem- 
ber in different years. It is sometimes confined to one well-de- 
4. Anot 
ual days of smoky weather, distributed through the whole eee 
and undistinguishable from Indi i artic 
every month, excepting about one month in wit 
ter and one in summer. me Ae Lieut. A. W. Whipple 
A., of Psychrometric and Climato ogical observations, 
Ocean, 2 ap H 
