10 



MEMOIRS OF THE NATIONAL ACADE.MY OF SCIENCES. 



from 45 to 50 miles per lioiir. We see tliat (with a few exceptions, whicb may generally be ascribed 

 to local causes) the winds all have a tendency inward toward the low center, and at the same time 

 tiiey circulate around this center in a direction contrary to the motion of the hands of a watch. 

 We also see that the average velocity of the wind is greatest on the northwest side, where the 

 gradient is steepest, and that the winds are generally feeblest where the gradients are least. 

 This remark must not be construed as applicable rigorously to the observation.s at each locality, 

 but rather to the average velocity of the wind over a considerable district. 



4. This storm was attended by a great fall of rain and snow, the precipitation at Louisville 

 Ky., having been 1.56 inches during the nine hours preceding 4'' 35™ p. m., January 17, and 2.13 

 inches during the seventeen hours jireceding 4'' 35'" p. M. The greatest rain fall was pretty near 

 the center of low i)ressure, and was situated upon the southeast side of it. The area over which 

 there was a fall of at least a quarter of an inch of water (in the form of raiii or snow) was G50 

 miles in diameter, measured in a direction from NW. to SE.; and 1,2.50 miles in diameter, measured 

 in a direction from SW. to NE. 



5. The contrasts between the temperatures prevailing on tlie op])osite sides of the low center 

 were uncommonly great. On the northwest side the thermometer was extremely low, viz, at 

 Pembina, — 22°; at Breckenridge,— 21°; at Fort Garry, — 18°, and at Yankton, — 13°. On the 

 southeast side of the low center the temperature was unusually high for midwinter, viz, at Key 

 West, 85°; at Pnnta Rassa,740; at .Tacksonville and New Orleans, 73°; at Savannah and Augusta, 

 70<5. Thus at the same instant of time, from Key West to Breckenridge (distant 1,700 miles), the 

 dift'erence of temperature was 106^, showing an average diflerence of one degree for each 16 

 miles of distance. In the neighborhood of the low center, the contrasts of temperature were still 

 more remarkable. At Memphis, the thermometer stood at 01°, showing a ditterence of temi)erature 

 from Yankton to Memphis (distance 690 miles) amounting to 74°, being an average difference 

 of one degree for each 9 English miles of distance. It wilt be seen hereafter that the phenomena 

 here noticed are in their main features characteristic of the violent storms of the United States, 

 particularly during the colder, months of the year. 



6. The isobars represented on Plate I are not circles. Occasionally we find examples iu which 

 the isobars about a low center approach more nearly to a circular form, but the Signal Service 

 mai)s do not, on an average, show more than one case in a year in which the isobars do not differ 

 sensibly from circles. 



From an actual measurement of the greatest and least diameters of the isol)ars represented 

 on the Signal Service maps for 7'" 35™ A. M. during a period of three years, the following average 

 results have been obtained: 



The average ratio of the longest diameter of the isobars to the shortest was 1.94. 



In 59"! per cent, of the whole imniber of cases, ( 1.5 



33 I the ratio of the longest J 2 • 



11 ( diameter to the shortest | 3 



nJ 



was greater than 



7. The longest diameter of the isobars may be turned in any azimuth, but it is most frequently 

 directed towards a point somewhat east of north. The following table shows the number of cases 

 in a hundred in which the longest diameter of the isobars was directed towards each of the ten- 

 degree intervals of azimuth, counting from the north point around by east towar<ls the south: 



The point towards which the hmgest diameter is most frequently directed is N. 36° E. If we 

 make a separate comparison of the cases occurring in the Mississippi Valley and those near the 



