E. Loomis— Observations of the U. S. Signal Service. 3 
The general results - this eee correspond very closely with 
those deduced from the table in my 5th paper. The lowest 
latitude of any storm pee sais in this table is 10°°6 
The lowest latitude shown in my 5th paper was 10°3 N. The 
average velocity of these storms while moving westerly was 
i: glish statute miles per hour; the average velocity of 
the odin mentioned in my 5th paper while moving westerly 
was 17-4 miles per hour. In nine of these cases the course of 
the storm became due north before reaching the parallel of 30°. 
Storm No. 18 apparently moved directly west; and storms 
Nos. 9 and 16 apparently moved for a day or two a little south 
of west. The table in my 5th paper shows 31 cases in which 
.the course of storms was towards the north of west, and only 
- two cases in which the course was south of west, viz: one case 
in which the course was two degrees south of west, and the other 
eleven degrees south of west. From the two tables we perceive 
that the cases in which tropical storms move in a direction north 
of west are fifteen times as frequent as the cases in which they 
move in a direction south of west, and in none of the cases here 
reported was the southerly motion very decided. Since in the 
middle latitudes the average progress of storm centers corre- 
sponds pretty nearly with the average direction of the wind, it 
might have been inferred that within the region of the northeast 
trade winds the average progress of storms should be towards 
the southwest. 
n order to determine whether, during the period here con- 
sidered, there may not possibly have been other storms which 
moved in a direction south of west, I have made a careful com- 
parison of the International Observations. Five-sixths of all 
the storms enumerated in the table on page 2, occurred in the 
months of August, September and Octo ber. I therefore 
selected these three months for special comparison. For the 
years 1876, 7, 8 and 9 the barometric curves were drawn for 
these months for all the stations reported in the International 
Bulletin between the equator an 26° N. 
An examination of these curves shows that at all of these sta- 
tions the fluctuations of the barometer are very small, : PeHeUt 
larly for the stations nearest to the equator. At Paramaribo, 
lat. 5° 45’ N. the entire range of the barometer for ‘hese twelve 
months was only 0°20 inch, and there was no oscillation which 
can be identified with an oscillation at either of the other sta- 
tions. At Bridgetown, lat. 13° 4’ N., the entire range of the ba- 
rometer for these twelve months was 0-23 inch, Two or three of 
the barometric oscillations at this station can probably be iden- 
tified with oscillations at some of the other stations. The track 
of storm No. 9 can apparently be traced back to Bridgetown 
on the 10th of Aug., 1878. At Fort de France, lat. 14° 40’ N., 
