********************************************************************************

10675 07/03/1891 M= 6  1 SNBR= 282 NOT NAMED   XING=1 SSS=9
10675 07/03/1891 M= 6  1 SNBR= 293 NOT NAMED   XING=1 SSS=1
                               ***                        *

10680 07/03*  0   0   0    0*217 930  35    0*220 932  45    0*224 935  55    0
10685 07/04*229 939  65    0*234 942  75    0*240 945  80    0*247 948  85    0
10685 07/04*229 939  65    0*234 942  75    0*240 945  80    0*247 948  80    0
                                                                        **

10690 07/05*254 951  85    0*262 954  85    0*271 956  85    0*281 956  85    0
10690 07/05*254 951  80    0*262 954  80    0*271 956  80    0*281 956  80    0
                     **               **               **               **

10695 07/06*292 954  80    0*303 951  70    0*312 947  60    0*319 943  50    0
10695 07/06*292 954  70    0*303 951  60    0*312 947  55    0*319 943  50    0
                     **               **               **

10700 07/07*325 938  45    0*331 931  40    0*337 923  40    0*342 911  35    0
10700 07/07*325 938  45    0*331 931  40    0*337 923  35    0*342 911  30    0
                                                       **               **

10705 07/08*350 881  35    0*  0   0   0    0*  0   0   0    0*  0   0   0    0
10705 07/08*346 897  25    0*350 881  25    0*  0   0   0    0*  0   0   0    0
            *** ***  **      *** ***  **

10710 HR
10710 HRBTX1CTX1
        ********

Partagas and Diaz (1996b) did not introduce any changes for this storm from
that shown in Neumann et al. (1999).  A peripheral pressure of 990 mb (at
0230Z on the 6th) suggests winds of at least 62 kt from the Gulf of Mexico
wind-pressure relationship - 80 kt chosen for best track, which reduces
the peak intensity originally in HURDAT slightly.  Decay to tropical 
depression stage over land included before dissipation.  Additional
six-hourly position added at end of track to allow for reasonable
translational speed of system.

********************************************************************************

10715 08/17/1891 M=13  2 SNBR= 283 NOT NAMED   XING=0 SSS=0
10715 08/17/1891 M=13  2 SNBR= 294 NOT NAMED   XING=0 SSS=0
                               ***

10720 08/17*  0   0   0    0*133 244  35    0*136 255  35    0*138 266  35    0
10725 08/18*140 277  35    0*142 288  40    0*144 299  50    0*146 310  60    0
10730 08/19*149 320  70    0*152 330  75    0*154 340  80    0*156 349  85    0
10730 08/19*149 320  65    0*152 330  65    0*154 340  65    0*156 349  65    0
                     **               **               **               **

10735 08/20*158 358  85    0*161 367  85    0*163 375  85    0*165 383  85    0
10735 08/20*158 358  65    0*161 367  65    0*163 375  65    0*165 383  65    0
                     **               **               **               **

10740 08/21*167 390  85    0*170 398  85    0*173 406  85    0*177 415  85    0
10740 08/21*167 390  65    0*170 398  65    0*173 406  65    0*177 415  65    0
                     **               **               **               **
 
10745 08/22*182 426  85    0*187 436  85    0*192 446  85    0*196 455  85    0
10745 08/22*182 426  65    0*187 436  65    0*192 446  65    0*196 455  65    0
                     **               **               **               **

10750 08/23*200 463  85    0*205 471  85    0*209 480  85    0*214 489  85    0
10750 08/23*200 463  65    0*205 471  65    0*209 480  65    0*214 489  65    0
                     **               **               **               **

10755 08/24*218 498  85    0*224 508  85    0*230 518  85    0*237 529  85    0
10755 08/24*218 498  65    0*224 508  65    0*230 518  65    0*237 529  65    0
                     **               **               **               **

10760 08/25*245 540  85    0*253 552  85    0*260 563  85    0*267 573  85    0
10760 08/25*245 540  65    0*253 552  65    0*260 563  65    0*267 573  65    0
                     **               **               **               **

10765 08/26*274 583  85    0*281 592  85    0*288 600  85    0*295 608  85    0
10765 08/26*274 583  65    0*281 592  65    0*288 600  65    0*295 608  65    0
                     **               **               **               **

10770 08/27*302 615  85    0*309 621  85    0*316 627  85    0*324 633  85    0
10770 08/27*302 615  65    0*309 621  65    0*316 627  65    0*324 633  65    0
                     **               **               **               **

10775 08/28*335 637  85    0*347 641  85    0*360 644  80    0*373 646  70    0
10775 08/28*335 637  65    0*347 641  65    0*360 644  65    0*373 646  65    0
                     **               **               **               **

10780 08/29*388 645  65    0*403 644  55    0*419 641  35    0*433 640  25    0
10780 08/29*388 645  65    0E403 644  55    0E419 641  35    0E433 640  25    0
                            *                *                *         

10785 HR    

No major changes from Partagas and Diaz (1996a), who made reasonable small
alterations to the track shown in Neumann et al. (1999).  Gale force and 
greater observations available for this system were the following:  997 mb
sea level pressure (suggestive of at least 53 kt from the subtropical 
pressure-wind relationship) at Bermuda on the 27th, a ship (the steamer
"Dunsmurry") capsized in the "hurricane" on the 29th (but no specific 
observations were provided), and 50 kt S wind on the 30th and 31st from the
steamer "La Touraine".  Thus available observational evidence suggests 
that the system may have achieved minimal hurricane intensity, but not
reaching Category 2 status as shown originally.  Winds reduced for much of 
the system's lifecycle.

********************************************************************************

10790 08/18/1891 M= 8  3 SNBR= 284 NOT NAMED   XING=1 SSS=9
10790 08/18/1891 M= 8  3 SNBR= 295 NOT NAMED   XING=1 SSS=1
                               ***                        *

10795 08/18*  0   0   0    0*  0   0   0    0*139 578  35    0*147 597  40    0
10795 08/18*  0   0   0    0*  0   0   0    0*133 580  90    0*139 594 100    0
                                              *** ***  **      *** *** ***

10800 08/19*155 614  65    0*162 629  75    0*168 640  80    0*174 649  85    0
10800 08/19*147 611 110  961*153 625 110    0*160 640 105    0*165 650 100    0
            *** *** ***  *** *** *** ***      ***     ***      *** *** ***

10805 08/20*179 657  85    0*184 665  80    0*190 672  80    0*196 679  80    0
10805 08/20*170 661  95    0*175 671  90    0*180 680  85    0*187 684  85    0
            *** *** ***      *** ***  **      *** ***  **      *** ***  **

10810 08/21*202 686  85    0*207 693  85    0*213 700  85    0*218 709  85    0
10810 08/21*196 686  85    0*203 689  85    0*210 695  85    0*215 702  85    0
            ***              *** ***          *** ***          *** ***  

10815 08/22*222 718  85    0*226 727  85    0*230 736  85    0*234 744  85    0
10815 08/22*218 710  85    0*221 717  85    0*225 726  85    0*229 735  85    0
            *** ***          *** ***          *** ***          *** ***  

10820 08/23*237 752  85    0*240 759  85    0*243 767  85    0*246 775  85    0
10820 08/23*233 745  85    0*238 755  85    0*243 767  85    0*246 775  85    0
            *** ***          *** ***          

10825 08/24*248 782  85    0*251 789  80    0*253 797  75    0*255 806  65    0
10825 08/24*248 782  80    0*251 789  75    0*253 797  70    0*255 806  55    0
                     **               **               **               **

10830 08/25*258 815  60    0*260 826  50    0*262 837  45    0*262 848  35    0
10830 08/25*258 815  50    0*260 826  45    0*262 837  40    0*262 848  35    0
                     **               **               **

10835 HR
10835 HRCFL1
        ****

No major changes from Partagas and Diaz (1996a), who made large alterations
to the track shown in Neumann et al. (1999).  These track changes are
found to be reasonable.  A central pressure reading of 961 mb (01Z on the 
19th) suggests winds of 99 kt from the southern wind-pressure relationship 
- 110 kt used in best track because of indications of a small radius of 
maximum wind (from Father Benito Vines' analysis quoted in the Partagas
and Diaz report) as well as extensive destruction in Martinique.  Hurricane 
is considered Category 1 (70 kt) at landfall in South Florida, but such 
designation is quite uncertain given the lack of observations near the 
landfall location.  Complete lifecycle of this hurricane is not available 
as neither the genesis nor the decay of the system was not documented.  The 
hurricane is also known as "San Magin", due to the rainfall-induced 
flooding that occurred in Puerto Rico.

********************************************************************************

10840 09/02/1891 M= 9  4 SNBR= 285 NOT NAMED   XING=0 SSS=0
10840 09/02/1891 M= 9  4 SNBR= 296 NOT NAMED   XING=0 SSS=0
                               ***

10845 09/02*  0   0   0    0*193 582  35    0*197 592  40    0*199 600  40    0
10850 09/03*201 607  45    0*204 615  50    0*208 623  55    0*213 631  60    0
10855 09/04*218 639  70    0*223 647  75    0*228 655  80    0*234 662  85    0
10860 09/05*239 669  85    0*246 676  85    0*252 683  85    0*260 690  85    0
10865 09/06*271 697  85    0*283 703  85    0*296 710  85    0*310 714  85    0
10870 09/07*327 715  85    0*347 713  85    0*368 703  85    0*395 680  85    0
10875 09/08*426 646  80    0*458 609  75    0*486 579  70    0*509 555  65    0
10880 09/09*529 533  60    0*547 511  55    0*562 492  50    0*575 475  45    0
10880 09/09E529 533  60    0E547 511  55    0E562 492  50    0E575 475  45    0
           *                *                *                *

10885 09/10*585 459  40    0*592 445  35    0*597 433  30    0*  0   0   0    0
10885 09/10E585 459  40    0E592 445  35    0E597 433  30    0*  0   0   0    0
           *                *                *

10890 HR           

Partagas and Diaz (1996b) did not introduce any changes for this storm from
that shown in Neumann et al. (1999).  Only intensity alteration is to
indicate extratropical stage for the hurricane north of 52N.
 
********************************************************************************

10895 09/16/1891 M=11  5 SNBR= 286 NOT NAMED   XING=0 SSS=0
10895 09/16/1891 M=11  5 SNBR= 297 NOT NAMED   XING=0 SSS=0
                               ***

10900 09/16*193 465  35    0*198 471  35    0*202 478  35    0*207 488  35    0
10905 09/17*213 498  35    0*218 507  35    0*223 516  40    0*228 524  40    0
10910 09/18*232 532  45    0*237 539  50    0*242 547  55    0*248 556  60    0
10915 09/19*254 566  65    0*260 575  70    0*266 583  70    0*272 590  75    0
10920 09/20*277 596  80    0*282 601  80    0*288 607  85    0*294 613  85    0
10920 09/20*281 600  80    0*288 607  80    0*295 615  85    0*300 621  85    0
            *** ***          *** ***          *** ***          *** ***

10925 09/21*300 618  85    0*306 623  85    0*312 628  85    0*318 631  85    0
10925 09/21*305 627  85    0*310 633  85    0*315 637  85    0*320 640  85    0
            *** ***          *** ***          *** ***          *** ***

10930 09/22*324 632  85    0*330 631  85    0*336 630  85    0*342 628  85    0
10930 09/22*325 641  85    0*329 641  85    0*333 640  85    0*340 635  85    0
            *** ***          *** ***          *** ***          *** ***

10935 09/23*349 624  85    0*356 617  85    0*363 607  85    0*370 591  85    0
10935 09/23*348 627  85    0*356 618  85    0*363 607  85    0*370 591  85    0
            *** ***              ***          

10940 09/24*375 573  85    0*379 555  85    0*382 538  85    0*382 523  85    0
10945 09/25*382 508  80    0*382 493  80    0*382 478  75    0*384 463  65    0
10950 09/26*388 448  55    0*394 433  40    0*402 418  35    0*413 397  30    0
10950 09/26*388 448  55    0*394 433  40    0E402 418  40    0E413 397  40    0
                                             *         **     *         **

10955 HR   

No major changes from Partagas and Diaz (1996a), who made reasonable small
alterations to the track shown in Neumann et al. (1999).  A peripheral 
pressure reading of 980 mb (05Z on the 22nd) suggests winds of at least 
75 kt from the subtropical wind-pressure relationship - 85 kt retained in 
the best track.

********************************************************************************

10960 09/29/1891 M=10  6 SNBR= 287 NOT NAMED   XING=0 SSS=0
10960 09/29/1891 M=10  6 SNBR= 298 NOT NAMED   XING=0 SSS=0
                               ***

10965 09/29*  0   0   0    0*207 546  35    0*212 547  35    0*217 550  40    0
10970 09/30*222 554  45    0*228 558  45    0*233 562  50    0*238 566  50    0
10975 10/01*244 571  55    0*249 576  60    0*255 582  60    0*261 590  65    0
10980 10/02*267 600  70    0*274 611  75    0*280 620  80    0*283 626  80    0
10980 10/02*266 597  70    0*271 604  75    0*277 613  80    0*283 620  80    0
            *** ***          *** ***          *** ***              ***

10985 10/03*287 633  85    0*291 639  85    0*295 645  85    0*301 653  85    0
10985 10/03*289 627  85    0*294 634  85    0*300 640  85    0*308 646  85    0
            *** ***          *** ***          *** ***          *** ***

10990 10/04*310 662  85    0*318 671  85    0*328 680  85    0*338 684  85    0
10990 10/04*318 654  85    0*327 660  85    0*335 667  85    0*346 675  85    0
            *** ***          *** ***          *** ***          *** ***

10995 10/05*350 686  85    0*364 688  85    0*380 688  85    0*399 678  85    0
10995 10/05*353 681  85    0*365 687  85    0*380 688  75    0*399 678  65    0
            *** ***          *** ***                   **               **

11000 10/06*425 650  85    0*453 616  85    0*472 580  80    0*484 549  75    0
11000 10/06E425 650  55    0E453 616  50    0E472 580  50    0E484 549  50    0
           *         **     *         **     *         **     *         **

11005 10/07*496 517  70    0*506 484  65    0*516 450  65    0*525 414  60    0
11005 10/07E496 517  50    0E506 484  50    0E516 450  50    0E525 414  50    0
           *         **     *         **     *         **     *         **

11010 10/08*534 373  60    0*542 331  55    0*550 295  50    0*568 255  45    0
11010 10/08E534 373  50    0E542 331  50    0E552 295  50    0E568 255  45    0
           *         **     *         **     ****             *         

11015 HR     

No major changes from Partagas and Diaz (1996a), who made reasonable small
alterations to the track shown in Neumann et al. (1999).  A peripheral 
pressure reading of 981 mb (01Z on the 4th) suggests winds of at least 
74 kt from the subtropical wind-pressure relationship - 85 kt wind retained 
in the best track.  Winds reduced from the 5th to the 8th due to 
observations supporting tropical storm intensity south of and over Canada.  
Position altered slightly on last day of system to allow a more realistic 
translational velocity.

********************************************************************************


1891/07 - 2003 REVISION:

11020 10/01/1891 M=10  7 SNBR= 288 NOT NAMED   XING=1 SSS=0
11020 10/01/1891 M=10  7 SNBR= 299 NOT NAMED   XING=1 SSS=0
                               ***

11025 10/01*  0   0   0    0*  0   0   0    0*175 598  35    0*176 608  40    0
11030 10/02*177 619  40    0*177 629  45    0*178 639  45    0*178 649  45    0
11035 10/03*179 659  45    0*179 670  45    0*179 680  45    0*180 690  45    0
11040 10/04*180 701  45    0*182 712  40    0*183 723  40    0*185 734  40    0
11045 10/05*189 745  40    0*193 756  45    0*198 767  45    0*204 778  45    0
11050 10/06*210 788  45    0*218 798  40    0*228 807  40    0*239 812  45    0
11055 10/07*250 812  45    0*262 810  40    0*273 804  40    0*284 797  40    0
11060 10/08*295 789  40    0*306 781  40    0*317 772  45    0*326 763  45    0
11065 10/09*334 753  45    0*342 743  45    0*350 732  45    0*360 718  45    0
11070 10/10*371 702  40    0*384 682  40    0*398 661  35    0*  0   0   0    0
11075 TS                    

No changes from Partagas and Diaz (1996a), who made no alterations to the 
track and intensity shown in Neumann et al. (1999).  (This storm along
with storms 8 and 9 is being further investigated by the re-analysis team.
Alterations - if any - will await the collection of all possible ship and
land based observations.)


1891/07 - 2004 REVISION:

11555 10/01/1891 M=10  7 SNBR= 299 NOT NAMED   XING=1 SSS=0
11020 10/04/1891 M= 7  7 SNBR= 300 NOT NAMED   XING=1 SSS=0
         **        **          ***  

(The 1st through the 3rd are removed from HURDAT.)
11560 10/01*  0   0   0    0*  0   0   0    0*175 598  35    0*176 608  40    0
11565 10/02*177 619  40    0*177 629  45    0*178 639  45    0*178 649  45    0
11570 10/03*179 659  45    0*179 670  45    0*179 680  45    0*180 690  45    0

11575 10/04*180 701  45    0*182 712  40    0*183 723  40    0*185 734  40    0
11025 10/04*  0   0   0    0*  0   0   0    0*150 790  35    0*160 795  40    0
            *** ***  **      *** ***  **      *** ***  **      *** ***  

11580 10/05*189 745  40    0*193 756  45    0*198 767  45    0*204 778  45    0
11030 10/05*170 800  45    0*175 805  45    0*180 810  45    0*187 815  45    0
            *** ***  **      *** ***          *** ***          *** ***  

11585 10/06*210 788  45    0*218 798  40    0*228 807  40    0*239 812  45    0
11035 10/06*195 820  45    0*205 823  45    0*215 825  45    0*226 823  45    0
            *** ***          *** ***          *** ***          *** ***  

11590 10/07*250 812  45    0*262 810  40    0*273 804  40    0*284 797  40    0
11040 10/07*237 820  40    0*248 815  45    0*260 810  40    0*271 803  40 1004
            *** ***  **      *** ***  **      *** ***          *** ***     ****
 
11595 10/08*295 789  40    0*306 781  40    0*317 772  45    0*326 763  45    0
11045 10/08*282 793  40    0*293 782  40    0*305 770  40    0*318 755  40    0
            *** ***          *** ***          *** ***  **      *** ***  **

11600 10/09*334 753  45    0*342 743  45    0*350 732  45    0*360 718  45    0
11050 10/09E332 740  40    0E346 725  40    0E360 710  45    0E370 695  40    0
           **** ***  **     **** ***  **     **** ***         **** ***  **

11605 10/10*371 702  40    0*384 682  40    0*398 661  35    0*  0   0   0    0
11055 10/10E375 675  35    0E378 650  35    0*  0   0   0    0*  0   0   0    0
           **** ***  **     **** ***  **      *** ***  **

11610 TS                    

U.S. Tropical Storm Landfall Data
---------------------------------
#/Date         Time   Lat    Lon    Max    States
                                   Winds  Affected
7-10/7/1891$   0000Z 25.0N  81.2W   45kt     FL
7-10/7/1891$   0800Z 25.2N  81.3W   45kt     FL
               ****  ****   ****

Three tropical systems in early October were first suggested by Mitchell 
(1924), which was utilized in the HURDAT database as well as Neumann et al. 
(1999) (storms 7, 8 and 9).  In contrast, the Monthly Weather Review summary 
of the era suggested one primary low forming in the Caribbean on the 6th,
moving across Cuba and Florida, impacting the U.S. mid-Atlantic states
and dissipating near Nova Scotia on the 14th.  MWR also had a secondary
low pressure forming near the Florida Keys on the 9th and merging with
the main low on the 11th.  Partagas and Diaz (1996a) believed that, "most 
likely, only one storm ... was what happened in reality".  However, they 
"did not find enough evidence to entirely disprove the existence of the 
three storms and, consequently, [they] decided to keep unchanged the tracks 
for Storms 7, 8 and 9, 1891."  

Subsequent research by the re-analysis team has uncovered evidence to
support a different conclusion to all the above:  two storm systems existed - 
1) a moderate tropical storm forming in the Caribbean on the 4th, moving
across Cuba and Florida, being absorbed in a frontal boundary and decaying
on the 10th (storm 7); and 2) a weak tropical storm also forming in the 
Caribbean on the 7th and becomming extratropical storm system near near the 
Florida Keys on the 9th, crossing Florida, slowing and becoming a strong 
"Nor'easter" on the 11th to the 13th and decaying on the 15th and 16th over 
Nova Scotia and Newfoundland (was storms 8 and 9, now combined into 8).  
Storm 9 apparently never existed as a separate storm system, but was in fact 
part of the extratropical storm stage for storm 8.  Thus the two original 
tropical storms and one original hurricane in HURDAT have been replaced with 
two tropical storms. 

Evidence for this scenario comes from both the COADS ship database, U.S. 
station data obtained from NCDC, and Bermuda observations provided by
Mike Chenoweth.  These were then plotted and analyzed twice daily from 1 to 
15 October, 1891.  (Figures showing the 
station and ship observations and the team's analyses are provided.)

The remainder of this writeup focusses upon storm 7.  The discussion for
the storm 8 (originally storm 8 and storm 9) is contained in that storm's
metadata file. 

The early portion of original storm 7's track (1st through the 5th) has
been discounted partly by ship data (especially on the 4th and 5th) but
primarily by the climatological studies of Puerto Rico (Salivia 1972),  
Hispanola (Garcia-Bonnelly 1958), and Cuba (Sarasola 1928).  The first two
comprehensive tropical cyclone listings indicated that no tropical storm
or hurricane impacted those locations in October 1891.  Perez (2003 - 
personal communication) reconfirmed the earlier Cuban historical study
that the tropical system in October 1891 formed in the Caribbean and
made landfall in south central Cuba late on the 6th of October, not moving 
in along eastern Cuba as suggested in HURDAT and Neumann et al. 

Ship data first indicate a closed circulation late on the 4th in the
western Caribbean.  Peripheral pressures of 1004 and 1005 mb on the 5th and 
6th from ships and Havana suggest winds of at least 39 and 36 kt, 
respectively.  These along with ship observations suggest a maximum 1 min
wind of about 45 kt for this time period.  (Some small weakening over
Cuba is accounted for on the 7th with a return to 45 kt intensity for
landfall in south Florida.)  Station observations clearly locate the
center of the storm during its trek across Florida on the 7th.  A sea
level pressure of 1004 mb in Jupiter, Florida at 1940 UTC may have been
a central pressure.  This suggests winds of about 40 kt, which is 
utilized for the HURDAT revision.  Over water observations are somewhat
sparse on the 8th and 9th, but enhanced winds in North Carolina (peak of
39 kt at Kitty Hawk and 35 kt at Cape Hatteras) suggest a relatively
close pass east of the state early on the 9th.  This likely occurred
soon after the system's extratropical transformation.  The storm then likely 
dissipated north of Bermuda on the 10th.  It is noted that the track 
provided here is quite similar from that found in HURDAT and Neumann et al. 
(1999) for storm 7 from the 7th to the 10th.


********************************************************************************

1891/08 - 2003 REVISION:

11080 10/06/1891 M= 6  8 SNBR= 289 NOT NAMED   XING=1 SSS=0
11080 10/06/1891 M= 6  8 SNBR= 300 NOT NAMED   XING=1 SSS=0
                               ***

11085 10/06*  0   0   0    0*159 830  35    0*164 832  40    0*171 837  40    0
11090 10/07*178 841  45    0*186 844  45    0*193 847  45    0*200 849  45    0
11095 10/08*208 850  45    0*215 850  45    0*223 848  45    0*233 844  45    0
11100 10/09*246 837  45    0*259 831  45    0*270 825  45    0*278 819  40    0
11105 10/10*284 813  35    0*291 807  35    0*298 801  40    0*307 793  45    0
11110 10/11*316 784  45    0*326 774  40    0*337 762  35    0*  0   0   0    0
11115 TS                    

No changes from Partagas and Diaz (1996a), who made no alterations to the 
track and intensity shown in Neumann et al. (1999).  (This storm along
with storms 7 and 9 is being further investigated by the re-analysis team.
Alterations - if any - will await the collection of all possible ship and
land based observations.)



1891/08 - 2004 REVISION:

11615 10/06/1891 M= 6  8 SNBR= 300 NOT NAMED   XING=1 SSS=0
11615 10/07/1891 M=10  8 SNBR= 301 NOT NAMED   XING=0 SSS=0
         **        **          ***                  *

(The 6th removed from HURDAT.)
11620 10/06*  0   0   0    0*159 830  35    0*164 832  40    0*171 837  40    0

11625 10/07*178 841  45    0*186 844  45    0*193 847  45    0*200 849  45    0
11625 10/07*  0   0   0    0*  0   0   0    0*180 850  30    0*185 850  30    0
            *** ***  **      *** ***  **      *** ***  **      *** ***  **

11630 10/08*208 850  45    0*215 850  45    0*223 848  45    0*233 844  45    0
11630 10/08*190 850  35    0*195 850  35    0*200 848  35    0*207 844  35    0
            ***      **      ***      **      ***      **      ***      **

11635 10/09*246 837  45    0*259 831  45    0*270 825  45    0*278 819  40    0
11635 10/09*215 840  35    0*225 837  40    0*240 835  40    0E255 828  35    0
            *** ***  **      *** ***  **      *** ***  **     **** ***  **

11640 10/10*284 813  35    0*291 807  35    0*298 801  40    0*307 793  45    0
11640 10/10E270 815  30    0E284 807  30    0E298 800  30    0E310 788  30    0
           **** ***  **     ****      **     *    ***  **     **** ***  **

11645 10/11*316 784  45    0*326 774  40    0*337 762  35    0*  0   0   0    0
11645 10/11E320 770  30    0E326 758  35    0E332 750  40    0E338 745  45    0
           **** ***  **     *    ***  **     **** ***  **     **** ***  **

(The 8th to the 11th from storm 9 removed.  The track from storm 9 on the
12th to the 16th incorporated into storm 8's track.)
11655 10/08/1891 M= 9  9 SNBR= 301 NOT NAMED   XING=0 SSS=0
11660 10/08*  0   0   0    0*238 572  35    0*245 582  40    0*251 591  40    0
11665 10/09*256 601  45    0*262 612  50    0*268 625  55    0*274 639  60    0
11670 10/10*280 653  70    0*287 666  75    0*293 680  80    0*299 694  85    0
11675 10/11*305 708  85    0*312 721  85    0*321 735  85    0*330 740  85    0

11680 10/12*340 742  85    0*350 741  85    0*359 740  85    0*366 737  85    0
11680 10/12E344 740  50    0E348 737  55    0E350 735  55    0E354 733  55    0
           **** ***  **     **** ***  **     **** ***  **     **** ***  **

11685 10/13*373 733  85    0*380 728  85    0*388 721  85    0*395 715  85    0
11685 10/13E360 731  55    0E370 729  55    0E380 725  55    0E390 715  55    0
           **** ***  **     **** ***  **     **** ***  **     ****      **

11690 10/14*403 706  85    0*412 694  85    0*422 681  85    0*433 665  85    0
11690 10/14E400 705  55    0E410 695  55    0E420 685  50    0E433 665  45    0
           **** ***  **     **** ***  **     **** ***  **     *         **

11695 10/15*446 647  80    0*460 626  75    0*475 602  70    0*500 571  65    0
11695 10/15E446 640  40    0E460 620  40    0E475 602  35    0E500 571  30    0
           *    ***  **     *    ***  **     *         **     *         **

11700 10/16*530 522  65    0*  0   0   0    0*  0   0   0    0*  0   0   0    0
11700 10/16E530 522  30    0*  0   0   0    0*  0   0   0    0*  0   0   0    0
           *         **  

11650 TS                    


U.S. Tropical Storm Landfall Data
---------------------------------
#/Date         Time   Lat    Lon    Max    States
                                   Winds  Affected
8-10/9/1891$   1400Z 25.8N  81.7W   45kt     FL
(Removed from listing)

9-10/12/1891*  0600Z 35.0N  74.1W   60kt     NC
(Removed from listing)

Three tropical systems in early October were first suggested by Mitchell 
(1924), which was utilized in the HURDAT database as well as Neumann et al. 
(1999) (storms 7, 8 and 9).  In contrast, the Monthly Weather Review summary 
of the era suggested one primary low forming in the Caribbean on the 6th,
moving across Cuba and Florida, impacting the U.S. mid-Atlantic states
and dissipating near Nova Scotia on the 14th.  MWR also had a secondary
low pressure forming near the Florida Keys on the 9th and merging with
the main low on the 11th.  Partagas and Diaz (1996a) believed that, "most 
likely, only one storm ... was what happened in reality".  However, they 
"did not find enough evidence to entirely disprove the existence of the 
three storms and, consequently, [they] decided to keep unchanged the tracks 
for Storms 7, 8 and 9, 1891."  

Subsequent research by the re-analysis team has uncovered evidence to
support a different conclusion to all the above:  two storm systems existed - 
1) a moderate tropical storm forming in the Caribbean on the 4th, moving
across Cuba and Florida, being absorbed in a frontal boundary and decaying
on the 10th (storm 7); and 2) a weak tropical storm also forming in the 
Caribbean on the 7th and becomming extratropical storm system near near the 
Florida Keys on the 9th, crossing Florida, slowing and becoming a strong 
"Nor'easter" on the 11th to the 13th and decaying on the 15th and 16th over 
Nova Scotia and Newfoundland (was storms 8 and 9, now combined into 8).  
Storm 9 apparently never existed as a separate storm system, but was in fact 
part of the extratropical storm stage for storm 8.  Thus the two original 
tropical storms and one original hurricane in HURDAT have been replaced with 
two tropical storms. 

Evidence for this scenario comes from both the COADS ship database, U.S. 
station data obtained from NCDC, and Bermuda observations provided by
Mike Chenoweth.  These were then plotted and analyzed twice daily from 1 to 
15 October, 1891.  (Figures showing the 
station and ship observations and the team's analyses are provided.)

The remainder of this writeup focusses upon storm 8.  The discussion for
the storm 7 is contained in that storm's metadata file. 

Original storm 8 and 9 were each depicting a portion of the same storm 
system that occurred.  The genesis of the revised system is delayed a day
until the 7th in the northwestern Caribbean.  By the time it reached
the Florida Keys on the 9th, it had merged with a pre-existing baroclinic 
zone and became an extratropical storm.  During the two day period
when the system maintained tropical cyclone status, peak observed winds
were 35 kt N from a ship at 14 UTC on the 9th at 21.0N 86.0W and lowest 
observed pressures were from same ship:  1004 mb at 22 UTC on the 7th at
20.0N 84.0W and 1005 mb at 10 UTC on the 9th at 21.0N 86.0W (though a time
series of pressure from this ship suggests that the values may be 
consistant 2-4 mb too low.  1004 mb peripheral pressure suggests winds of 
at least 39 kt from the southern pressure-wind relationship.  Peak estimated 
winds as a tropical storm are 40 kt on the 9th.  However, by the
time the system reached the Florida Keys as an extratropical system, either 
it had weakened slightly or had not actually attained tropical storm 
intensity.  Peak conditions observed were only 21 kt and 1012 mb in Key West
as the system passed just to the west of the city.  The extratrpical storm 
then moved slowly northeastward across Florida into the Atlantic and then 
drifted to the north beginning on the 11th for about 36 hours southeast of 
Cape Hatteras.  During this time a high built in from the north and west and 
in conjunction with the extratropical storm caused strong northeasterly winds 
along the U.S. mid-Atlantic and New England coasts.  Peak (uncorrected) 5 min 
sustained winds reached 57 kt at Kitty Hawk, 63 kt at Cape Hatteras, 41 kt at 
Atlantic City, 63 kt at Block Island, and 50 kt at Nantucket.  The estimated
maximum 1 min winds for this system during its extratropical stage were
about 55 kt.  On the 13th to the 15th, the baroclinic low moved northeastward 
and weakened.  The baroclinic nature of this system is quite clear - it had 
at times a 25F east-west temperature gradient while along the Atlantic coast.
The early portion of the original storm 9 also appears to be incorrect 
based upon ship and Bermuda data on the 8th to the 11th.  There is no
indication that a low (tropical or baroclinic) came toward the U.S.
Atlantic seaboard from the southeast.  However, the portion of original
storm 9's track from the 12th to the 15th does closely match the analysis 
here of the extratropical storm stage for this revised storm 8.  However,
it is to be noted that the evidence for retaining this system in HURDAT
at all as a tropical storm is marginal given one gale force report and
a couple suspect low pressure readings.

********************************************************************************

1891/09 - 2004 REVISION:

Note:  Storm was originally 1891/10, but became 1891/09 after the removal
of the original 1891/09 - May 2004.


1891/09 - 2003 REVISION:

11175 10/12/1891 M= 9 10 SNBR= 291 NOT NAMED   XING=0 SSS=0
11175 10/12/1891 M= 9 10 SNBR= 302 NOT NAMED   XING=0 SSS=0
                               ***

11180 10/12*  0   0   0    0*  0   0   0    0*125 613  35    0*136 620  45    0
11180 10/12*  0   0   0    0*  0   0   0    0*125 613  35    0*136 620  35    0
                                                                        **

11185 10/13*147 627  60    0*156 632  70    0*162 636  80    0*167 639  80    0
11185 10/13*147 627  40    0*156 632  40    0*162 636  45    0*167 639  45    0
                     **               **               **               **

11190 10/14*172 641  85    0*177 644  85    0*182 646  85    0*187 648  85    0
11190 10/14*172 641  50    0*177 644  50    0*182 646  55    0*187 648  55    0
                     **               **               **               **

11195 10/15*192 650  85    0*197 652  85    0*202 654  85    0*213 656  85    0
11195 10/15*192 650  60    0*197 652  60    0*202 654  65    0*213 656  70    0
                     **               **               **               **

11200 10/16*224 658  85    0*234 660  85    0*245 662  85    0*256 663  85    0
11200 10/16*224 658  75    0*234 660  75    0*245 662  75    0*256 663  75    0
                     **               **               **               **

11205 10/17*267 662  85    0*277 661  85    0*288 660  85    0*297 659  85    0
11205 10/17*267 662  75    0*277 661  75    0*288 660  75    0*297 659  75    0
                     **               **               **               **

11210 10/18*303 659  85    0*310 658  85    0*320 657  85    0*334 655  85    0
11210 10/18*303 659  75    0*310 658  75    0*320 657  75    0*334 655  75    0
                     **               **               **               **

11215 10/19*353 653  85    0*372 649  85    0*390 639  85    0*409 623  85    0
11215 10/19*353 653  75    0*372 649  75    0*390 639  70    0*409 623  65    0
                     **               **               **               **

11220 10/20*430 602  80    0*448 582  65    0*465 570  60    0*485 575  35    0
11220 10/20*430 602  60    0*448 582  50    0*465 570  40    0*485 575  35    0
                     **               **               **               **

11225 HR

Partagas and Diaz (1996a) did not introduce any track changes from 
that shown in Neumann et al. (1999) for this hurricane.  Gale force and 
greater observations available for this system were the following:
"gales of hurricane force" on the 17th east-northeast of the Bahamas, 
60 kt SE-SW wind and 992 mb on the 18th at Bermuda (this peripheral
sea level pressure suggests winds of at least 61 kt from the subtropical
wind-pressure relationship - 75 kt utilized), 70 kt wind on the 19th 
("Ocean Prince") at 36 N, 62 W.  Available observational evidence 
suggests that the peak intensity for this hurricane was a minimal 
hurricane (Category 1), rather than the standard Category 2 (85 kt) 
originally in HURDAT.  Winds are reduced accordingly from the 13th to 
the 20th.  Hurricane intensity attained after passing through the Lesser 
Antilles.

********************************************************************************

1891/10 - 2004 REVISION:

Note:  Storm was originally 1891/11, but became 1891/10 after the removal
of the original 1891/10 - May 2004.


1891/10 - 2003 REVISION:

11230 11/03/1891 M= 4 11 SNBR= 292 NOT NAMED   XING=0 SSS=0
11230 11/03/1891 M= 4 11 SNBR= 303 NOT NAMED   XING=0 SSS=0
                               ***

11235 11/03*241 738  35    0*250 725  35    0*257 716  40    0*268 703  45    0
11240 11/04*279 687  45    0*291 668  50    0*302 647  50    0*313 621  50    0
11245 11/05*325 592  50    0*338 562  50    0*352 538  45    0*380 512  40    0
11250 11/06*416 490  35    0*  0   0   0    0*  0   0   0    0*  0   0   0    0
11255 TS

Partagas and Diaz (1996b) did not introduce any changes for this storm from
that shown in Neumann et al. (1999).  The only change is to renumber the 
storm number for the year.

********************************************************************************

1891 - Additional Notes - 2004 REVISION:


1) The NHC Best Track Change Committee requested further investigation into
the possible storm system first identified by Partagas and Diaz (1996b)
in mid-July 1891:

   "1891  additional system #1 (July)  MWR mentions 'gale' winds.  
    This system may warrant further research.  Is there any COADS?"

Upon investigation of this system in the Monthly Weather Review, from
the COADS ship database, and coastal station data, this system was of
tropical depression intensity at its peak.  The COADS data were sparse
in the vicinity of the system until the 12th, when it was east of the
U.S. mid-Atlantic states.  Peak ship observations were 25 kt and 1014 mb,
though a weak closed circulation was analyzed.  Station data were also
obtained for Jacksonville, Jupiter, Titusville, Savannah, Charleston,
Wilmington, Cape Hatteras, Kitty Hawk, Baltimore, Atlantic City, New 
York City, New London, New Haven, Block Island and Nantucket.  Peak 
observed winds were 36 mph at Kitty Hawk (10th and 11th) and at 
Cape Hatteras (11th).  These observations also support tropical depression
status for this system.  While "fresh to strong gales" were mentioned
in the Monthly Weather Review, no evidence for these were to be found
from any source.  Thus this system is not added into HURDAT.


2) September 11-12, 1891:  Partagas and Diaz (1996b) had investigated 
this system for possibility of inclusion into HURDAT.  The re-analysis team 
agreed with Partagas and Diaz to leave it out of HURDAT as there was not
enough evidence for tropical storm intensity.


3) September 14-15, 1891:  Partagas and Diaz (1996b) had investigated 
this system for possibility of inclusion into HURDAT.  The re-analysis team 
agreed with Partagas and Diaz to leave it out of HURDAT as there was not
enough evidence for tropical storm intensity.


4) The NHC Best Track Change Committee requested further investigation into
the possible storm system first identified by Partagas and Diaz (1996b)
in mid-September 1891:

   "1891  additional system #4 (Sept)  MWR gives TS force winds at 
    coastal stations.  This system is also given a high probability 
    by P+D, and bears further investigation."

Upon investigation of this system in the Monthly Weather Review, from
the COADS ship database, and coastal station data, a closed circulation for
this system could not be found.  The COADS ship data did not provide any
observations near the system, though the station data was quite thorough
with observations obtained for Key West, Jupiter, Titusville, Tampa, 
Jacksonville, Pensacola, Mobile, Port Eads, New Orleans and Galveston.
These data indicate that a disturbance did cross the Gulf of Mexico
being located roughly along longitude 78W on the 17th, 80W on
the 18th, 82W on the 19th, 85W on the 20th, 88W on the 21st, and 90W
on the 22nd.  Peak winds recorded in association with this system were
40 mph E at Titusville on the 19th, 48 mph NE at New Orleans on the 20th,
and 40 mph NE at Galveston on the 20th.  However, only easterly winds were 
reported from these locations and lowest pressure recorded was only 
1014 mb at Key West on the 19th.  It is possible that this was a
tropical storm, but confirming observations for having a closed 
circulation were not found.  (It is also possible that the system was
a vigorous easterly wave with no closed circulation and a NNE-SSW oriented
wave axis.)  Thus because of the uncertainty and lack of having an observed 
closed circulation, this system is not included into HURDAT.


5) Storm 9 in Partagas and Diaz (1996a) and Neumann et al. (1999) apparently 
did not exist as a separate tropical cyclone, but was in fact part of the 
extratropical storm stage for storm 8.  Thus this system is removed from
HURDAT.  Details on the observations for this removal can be found within
the discussion of storm 8.  (Figures showing 
the station and ship observations and the team's analyses are provided.).  
Below is the original HURDAT entry for this system:

11120 10/08/1891 M= 9  9 SNBR= 290 NOT NAMED   XING=0 SSS=0
11125 10/08*  0   0   0    0*238 572  35    0*245 582  40    0*251 591  40    0
11130 10/09*256 601  45    0*262 612  50    0*268 625  55    0*274 639  60    0
11135 10/10*280 653  70    0*287 666  75    0*293 680  80    0*299 694  85    0
11140 10/11*305 708  85    0*312 721  85    0*321 735  85    0*330 740  85    0
11145 10/12*340 742  85    0*350 741  85    0*359 740  85    0*366 737  85    0
11150 10/13*373 733  85    0*380 728  85    0*388 721  85    0*395 715  85    0
11155 10/14*403 706  85    0*412 694  85    0*422 681  85    0*433 665  85    0
11160 10/15*446 647  80    0*460 626  75    0*475 602  70    0*500 571  65    0
11165 10/16*530 522  65    0*  0   0   0    0*  0   0   0    0*  0   0   0    0
11170 HR

********************************************************************************