CONTRIBUTIONS TO METEOROLOGY. l9 



West India Islands, and lor which definite )>atlis liave been determined. (Jolumn 1 shows the 

 number of reference ; column 2 gives the date of commencement of the storm so far as ascertained ; 

 column 3 shows the latitude of the storm's center, when it first became violent; column 4 shows 

 the average course of the storm while moving west \\ard; column .") shows the hourly velocity of 

 progress in the preceding i)art of its course; column (i shows the latiiude at which the storm was 

 movinjr due north ; column 7 shows the average conrsi^ of the storm after turning eastward, until 

 it reached the ])aralle] of 40°; column S shows the hourly velocity of ])rogress during the pre- 

 ceding i)eriod; column t) shows whether lain was mentioned as accomi)auying the storm; column 

 10 gives the name of the person by whom the phenomena of the storm were investigated, and 

 column 11 shows where the record of the inv(>stigation maybe found. 



29. It will be noticed that the least latitude of any storm path here recorded is 10°; that is, 

 over the Atlantic Ocean no storm path has been traced within 10° of the equator. 



The courses of the storms nientioned in this table (while moving westward) range from 11^° 

 south of west to 62° north of west. In two cases the course was a little south of west ; in a third 

 case the course was only one degree north of west, and in a fourth case the course was only five 

 degrees north of west. Tropical stoi ms do therefore sometimes travel towards the equator, and it 

 maybe suspected that this directioii occurs more frequently than the table would indicate, since 

 many of the storms here recorded would never have been selected for investigation if they had not 

 advanced into the middle latitudes. The average course of the storms here enumerated, while they 

 were moving westward, was west 26° north ; and the average hourly velocity in this part of their 

 course was 17.4 miles. 



The average latitude of the storm's center when moving due north was 29^°, and the latitudes 

 range from 23^° to 34°. During the three summer months the average latitude is 30°. 6; in 

 Septeuiber it is 29°. 7, and during the other months of the year 2()°.7, indicating that the i)oint 

 where the course changes from west to east is somewhat more northerly in summer than in winter. 

 The average course of these storms while traveling easti^ard to the parallel of 40° was E. 38J° N., 

 ranging from 17° to 60°. The average hourly velocity in this part of their course is 20.5 miles, 

 which is a little less than the average velocity of storms in the United States for the months of 

 August and September, according to the Signal Service observations. It will be seen from column 

 9, that rain generally accompanies cyilones. In three of the cases I have found in the published 

 reports no mention of rain, but it is presumed that this is simply an oversight, since in most of the 

 other cases rain is only incidentally mentioned. In all the iuvestigatious of Redfield and Reid the 

 circumstances upon which they insist as specially important are the direction and force of the wind, 

 and it is only by consulting the extracts from .he log books which they have furnished us that I 

 have discovered anj mention of accompanyiug rain. It is believed that tropical cyclones never 

 occur without rain, and generally the rain is described as descending in torrents. The letter V, in 

 column 9, signifies very. 



30. In order to obtain more complete information respecting the tracks of tropical cyclones in 

 the neighborhood of the West India Islands, I have compared all the storm tracks delineated on 

 the maps of the Monthly Weather Review, and also these delineated on the international charts. 



