8 K LootHlx—CohtrHniHons t<> Mdeorology. 



In No. 14 it seems probable that the rain-fall reported at 

 Keokuk did not all fall during- the preceding eight hours, but 

 represents the entire rain-fall for the preceding twenty-four 

 hours. This is indicated by a comparison of the record at 

 Keokuk with that at the neighboring stations St. Louis and 

 Davenport. The following is a copy of the record. 





January 2.2 Kaln . 



Weather. 1 Bfe. Weather. 



Rain. 



Weather. 1 Bain. 1 ** L ' 



Keokuk Light rain ! .... I| Lkht rain 

 Heavy i a 0*30 Foggy 



Dav.-npr.rt Cloudy .... Li-ht rain 



1-01 



Liui.r -now IM V0:! 

 Cloudy 0-:t2 1 -J.J 

 I.iu-ht rain 0-10 i! I'll 



At St. Louis, up to Jan. 2.1, the rain in sixteen hours 

 amounted to 0-94 inch ; and at Davenport it amounted to 101 

 inch, while at Keokuk (an intermediate station) no rain is re- 

 corded in the rain column although rain was reported in the 

 weather column both at Jan. 1.3 and Jan. 2.1. Tt seems evi- 

 dent that the observer at Keokuk neglected to measure the 

 rain until Jan. 2.2, and that a portion of the amount recorded 

 for Jan. 2.2 must have fallen before Jan. 2.1. If this conclu- 

 sion is correct, then the observations indicate that the principal 

 rain-center for Jan. 2.2 was on the east side of the low center. 



In No. 26 the principal rain-center was 137 miles northwest 

 of the low center, and during the succeeding eight hours the' 

 low center advanced 156 miles towards the northwest. 



In Nos. 46 and 47 the principal rain-center was northwest of 

 the low center, and the low center advanced towards the north- 

 west, as was fully shown in my seventh paper. 



In No. 53 the principal rain-center was 290 miles southwest 

 of the low center. At Nov. 23.3 the lowest isobar formed a 

 very elongated oval whose principal axis extended from Vicks- 

 burg to Lake Erie, and the heaviest rain occurred near Vicks- 

 burg, around which place prevailed a very decided cyclonic 

 movement of the winds. There was at the same time in the 

 neighborhood of Lake Erie a fall of rain and snow of less 

 amount, but extending over a greater area. 



From this comparison we find that during the entire period 

 of the published tri-daik observations (thirty-seven months), if 

 we exclude the cases in which the rain-center was more than 

 500 miles soatb of the low center, there were but six cases of 

 gr.-a' rain storms in which the principal rain-center was south- 

 west of the low center, and in each of these cases the lowest 

 isobar formed an oval whose longest diameter was from three 

 to five times its shortest diameter, and the longest diameter 

 was dire-., i towards the northeast. The principal rain-center 

 appeared to be in the southwest portion of this oval ; but at 



