258 NINTH ANNUAL REPORT OP 



tember, then August, then April, and June stands the sixth in order, 

 and only two degrees above November. In neither of the three years 

 preceding was July the warmest month. In 1851 the warmest months 

 occurred in the following order : August, October, September, June, 

 July, April. In 1852: September, July, August, June, October, No- 

 vember. In 1853: October, September, June, May, July, August. To 

 the daily visitations of the cold ocean wind in the summer is owing this 

 great variation from the order of the months as to comparative temper- 

 ature in other climates. 



January was the coldest month, then February, then December, and 

 next March. In other years, December sometimes takes the precedence 

 of January. February, which in the Atlantic States is often the cold- 

 est month in the year, is not so here. 



Rain fell on 54 days in the year, 22.12 inches in depth. This is our 

 average supply, though only half the quantity that falls in the Atlantic 

 States. In 1853, the quantity was 19.03 inches; in 1852,25.60 inches; 

 and in 1851, only 15.12 inches. The old inhabitants tell of occasional 

 seasons when scarcely any rain has fallen, and when the cattle have 

 perished from want. Such very dry seasons are said to recur at in- 

 tervals of eight or ten years. 



The greatest amount of rain was in February, next comes January, 

 next April, then March. This differs from the ordinary arrangement. 

 Taking the last four years iuto viev»^, December gives the most rain, 

 and March comes next, while the intervening months are comparatively 

 dry. In fact, we have the early rains, beginning in November and con- 

 tinuing through December into the early partof Januajy ; and tlie later 

 rains, beginning in March, and continuing at times through April. 



Lightning is seen at San Francisco on an average three or l()ur time? 

 a year, and thunder is heard less frequently. On the 15lh January. 

 Hashes of lightnino- were ol)served in the evenins:, during a cold rain 

 Storm from the south; and on the 22d February, under similar circum- 

 stances, lightning was again noticed. But no thunder was audible it: 

 either case ; nor was there any further exhibition of atmospheric elec- 

 tricity during the year. There was thunder three times in 1851, five 

 times in 1852, and twice in 1853. 



No exhibition of auroral light was observed in the year. Since mj^ 

 residence here, from August, 3850, 1 have seen the aurora borealis only 

 on two occasions, once in January and once in February, in the 

 year 1852. 



There was no unusual display of shooting stars during the year. In 

 September, 1851, in August, 1852, and in August, 1853, the}' were 

 numerous for several nights in succession. 



Earthquake shocks were distinctly felt on the mornings of the 9th' 

 of January and 21st of October. 



