REESE LIBRARY 
 UNIVERSITY OF CALIFORNIA. 
 
 Received 
 
 Accessions No.^.K?.''-.. Shelf No. 
 
f- 
 
THE 
 
 LAW OF STORMS 
 
 CONSIDERED PRACTICALLY; 
 
 BEING 
 
 A DIGEST OF THE CIRCULAR THEORY OF STORMS, AND THE MODIFICATION OF THAT 
 
 THEORY AS DUE TO THE IN-GOING SPIRAL CIRCULATION OF THE WIND 
 
 IN A CYCLONE ; TOGETHER WITH A SUMMARY OF THE 
 
 RESULTS OF RECENT INVESTIGATION. 
 
 WITH NUMEROUS ILLUSTRATIONS. 
 
 BY W. H. ROSSER, 
 
 Author of the" Deviation of the Compass in Iron Ships Considered Practically," 
 " How to Find the Stars," " Stellar Navigation," &c. &c. 
 
 SECOND EDITION, WITH IMPORTANT CORRECTIONS AND ADDITIONS. 
 
 LONDON: 
 
 PUBLISHED BY NOEIE & WILSON, 
 PUBLISHER OF CHARTS AND NAUTICAL WORKS, 
 
 AT THE NAVIGATION WAREHOUSE AND NAVAL ACADEMY, 
 
 156 MINORIES. 
 
 1886. 
 
REID AND SON, PRINTERS, LEITH 
 
PREFACE TO THE FIRST EDITION. 
 
 THE progress of meteorology an<i a more systematized method of 
 observation having developed new views in respect to the nature and 
 origin of Storms, the subject has, during the last ten or dozen years, 
 been one of continual discussion and much controversy. This work has 
 been written to give, in brief, the history of the development of the 
 Law of Storms, to contrast the old " circular " with the later " spiral " 
 theory, and to show to what extent the old practical rules for storm- 
 sailing require modification. 
 
 W. H. K. 
 May 1876. 
 
 PREFACE TO THE SECOND EDITION. 
 
 WHEN the first edition of this work was published, the positive evidence 
 of the in-moving spiral circulation of the wind in a cyclone was scanty ; 
 recent research has thrown more light on the subject, and the result of 
 the latest observations in this department of meteorology have been 
 embodied in the second edition. The last part of the work is con- 
 sequently new, being chiefly due to the investigation of observers in 
 the tropical regions of the globe, but the views of no recent authorities 
 in meteorology have been omitted or neglected, to make the Law of 
 Storms as applied to navigation as complete as possible. 
 
 W. H. ROSSER, 
 
 NAUTICAL ACADEMY, 
 
 AMERICA SQUARE, MINOEIES, 
 
 LONDON, April 1886. 
 
THE LAW OF STORMS 
 
 COUSIDEEED PRACTICALLY. 
 
 Hurricane, Typhoon, and Cyclone * are but different names for 
 the same natural phenomenon a fierce tempest of wind and rain, with 
 thunder and lightning, accompanied by a heavy cross sea, and marked 
 by a very low barometer. The wind in such storms does not blow in 
 straight lines, but, according to the popular idea, it has a horizontal 
 rotation, as if on an axis : the centre of the storm is a calm area of 
 varying diameter, with the sea still rough, but the sun not unfrequently 
 shining by day or the stars by night ; immediately around this central 
 calm area is the dark bank of cloud where the hurricane is blowing 
 with its greatest strength ; and a ship that has entered the calm with 
 one wind leaves it with another blowing from an opposite, or nearly 
 opposite, direction, indicating that while there is a movement of hori- 
 zontal rotation, the storm-field is also subjected to a movement of 
 translation. 
 
 Such tempests are especially characteristic of, though not confined 
 to, the tropical regions of our globe ; they very frequently invade the 
 extra-tropical zones, and even occasionally originate in the latter, espe- 
 cially during the winter of the hemisphere ; they are subject to laws, as 
 are all physical phenomena ; and as it is of vital importance that a ship 
 should not be heedlessly navigated into or through the storm-field, which 
 tradition and experience have alike pronounced to be disastrous, practical 
 rules have been deduced from the Law of Storms, by the aid of which 
 the seaman may avoid the most dangerous parts of the storm, escape 
 therefrom when he happens to be caught in the vicinity, or occasionally 
 utilise the more moderate storm- winds when they are propitious for the 
 prosecution of the voyage. 
 
 * The term Hurricane comes through the Spanish huracan, which is said to be derived from a 
 West Indian native word imitative of rushing wind. Ty-foong is a Chinese word indicating a great 
 or mighty wind. Cyclone, as applied to the revolving gales of Indian seas, is said to have first been 
 used by Piddington in his eighteenth memoir on storms, which appeared in Jour. Asiatic Soc. of 
 Bengal, vol. xviii. 
 
2 HISTORY AND DEVELOPMENT 
 
 The literature relating to and bearing on storms and hurricanes is 
 voluminous. In 1865 Mr. ANDRE POEY, of Havana, who has himself 
 devoted much time and attention to meteorology, published a Catalogue 
 of 1008 works, tracts, memoirs, reports, and papers with a list of 
 upwards of 500 authors and writers all treating directly or indirectly 
 on hurricanes and cyclonic storms, tropical and extra-tropical. Many 
 of these productions date as far back as the 16th and 17th centuries; 
 but, as a rule, the writers of that period, though familiar with the 
 hurricanes of the West and East Indies, and the typhoons of the China 
 Sea, were not adepts in even the crude meteorological science of the 
 day : they conjectured much, while, on the other hand, they have given 
 most graphic and accurate descriptions of the storms to the perils of 
 which they were as seamen exposed, and to the destructive effects of 
 which they were eye-witnesses; of the veering of the wind through 
 several points of the compass, and of its mode of veering, they say little 
 or nothing, though they were not unacquainted with the fact that the 
 wind changed from one quarter to its opposite after an interval of 
 calm. Later on there were more skilful observers, who made some 
 shrewd guesses at truth while speculating on the phenomena that had 
 fallen under their notice. But it is only since the commencement of 
 the 19th century that anything approximating to correct ideas of the 
 tremendous tropical hurricane has been worked out ; the physical aspect, 
 character, and course of such storms attempted to be defined ; and their 
 seasons and range, within certain limits, ascertained. 
 
 It is probable that since 1865 until now POEY'S Catalogue has been 
 augmented by not less than 300 additional works, memoirs, and papers, 
 for meteorology has come to be a science of note, and the number of 
 observers and inquirers in the field has largely increased. Surely, 
 then, with such a mass of literature on a special department of the 
 science of meteorology, the Law of Storms must now, for all essential 
 purposes, be well understood, and the practical rules deduced from that 
 law well collated and known, for on the accuracy of such knowledge, 
 and the readiness with which it can be applied, depends the safety of 
 life and property where ships are so constantly, and in such numbers, 
 navigating the hurricane regions. It is proposed in these pages to show 
 that the law of storms requires reconsideration, and a more thorough 
 and critical examination of all the details on which it rests ; to show, 
 in fact, that there are doubts as to its being as well known as it was 
 thought to be, and this arising perhaps from an imperfect collating of 
 facts, perhaps from a too hasty generalization in respect to the facts, 
 or, again, from an erroneous conclusion that the physical characteristics 
 of all hurricanes and all parts of the storm-field are identical. It may 
 be as well, therefore, to commence by briefly tracing the history of the 
 
OP THE LAW OF S*OBfcl& I V E U S IT 
 
 ^ / 
 
 investigation of the law of storms; then, toeia $&. progress of the 
 development of that law in the past at the same time bringing together 
 the whole of the practical rules that have hitherto been laid down as 
 applicable to navigation for the avoidance of the inner and most 
 destructive part of the storm-field ; and finally, to draw special attention 
 to the new views on hurricanes, as resulting from numerous observations 
 lately, but carefully, made in India, Mauritius, and Reunion, in China 
 and Japan, and at Havana, observations which have caused notable 
 meteorologists to doubt the rotatory direction of these storm-winds, 
 at least in the sense until now supposed, and to dispute the strict 
 accuracy of the so-called law of storms, together with some of the 
 practical rules appertaining thereto as propounded by many eminent 
 men of science and writers during the last thirty or forty years, and 
 whose opinions still carry great weight. 
 
 History. COLUMBUS discovered America in 1492, and shortly after- 
 wards Spanish writers describe the West Indian hurricane : amongst 
 the earliest of these writers was FERNANDEZ DE OVIEDO, who, in his 
 Historia general y Natural de las Indias y Tierra Firme de la Mar 
 Oceano, gives an account of the Santo Domingo hurricanes of 1508 
 (Aug. 3) and 1509. Taking the storm of 1508, which lasted twenty-four 
 hours, he notes the destruction of houses, the loss of life, and also the 
 immense havoc done in different parts of the island j he remarks that the 
 wind commenced at North, driving ships from their moorings, and sub- 
 sequently changed suddenly to an opposite quarter, blowing then from 
 the South as violently and with as much fury as before from the North. 
 The Indians distinguished these excessively tempestuous storms as 
 huracanes. 
 
 WILLIAM DAMPIER, an English navigator and companion of some of 
 the old buccaneers, was an intelligent observer and careful narrator of 
 all he saw. He briefly describes the incidents and phenomena of the 
 typhoon in the China Seas with an accuracy scarcely to be surpassed : 
 
 " Tuffoons are a particular kind of violent Storms, blowing on the 
 Coast of Tonquin and the neighbouring Coasts in the Months of July, 
 August, and September. They commonly happen near the full or change 
 of the Moon, and are usually preceded by very fair weather, small 
 Winds, and a clear Sky. Those small Winds vere from the common 
 Trade of that time of the year, which is here at S. W. and shuffles about 
 to the N. and N.E. Before the Storm comes there appears a boding 
 Cloud in the N.E., which is very black near the Horizon, but towards 
 the upper edge it looks of a dark Copper Colour, and higher still it is 
 brighter, and afterwards it fades to a whitish glaring colour at the very 
 
4 HISTORY AND DEVELOPMENT 
 
 edge of the Cloud. This Cloud appears very amazing and ghastly, and 
 is sometimes seen 12 Hours before the Storm comes. When that Cloud 
 begins to move apace, you may expect the Wind presently. It comes 
 on fierce, and blows very violent at N.E. 12 Hours more or less. It is 
 also commonly accompanied with terrible claps of Thunder, large and 
 frequent flashes of Lightning, and excessive hard Rain. When the 
 Wind begins to abate it dies away suddenly, and falling flat calm, it 
 continues so an hour, more or less ; then the Wind comes about to the 
 S.W., and it blows and rains as fierce from thence as it did before at 
 N.E. and as long." (W. DAMPIER'S Voyages and Descriptions, vol. ii. 
 p. 36.) 
 
 DAMPIER being at St. John's Island* in July 1687, encountered 
 one of these tempests, and the following is his graphic account of what 
 occurred : 
 
 "It was now the time of the Year for the S.W. Monsoon, but the 
 Wind had been whiffling about from one part of the Compass to another 
 for two or three Days, and sometimes it would be quite calm. This 
 caused us to put to Sea, that we might have Sea-room at least ; for such 
 flattering Weather is commonly the forerunner of a Tempest. 
 
 " Accordingly we weighed Anchor, and set out ; yet we had very 
 little Wind all the next Night. But the Day ensuing, which was the 
 4th day of July, about Four a-Clock in the Afternoon, the Wind came 
 to the N.E. and freshned upon us, and the Sky looked very black in 
 that quarter, and the black Clouds began to rise apace and moved 
 towards us, having hung all the Morning in the Horizon. This made 
 us take in our Top-sails, and the Wind still increasing, about Nine 
 a-Clock we rift our Main-sail and Fore-sail ; at Ten we furl'd our Fore- 
 sail, keeping under a Main-sail and Mizen. At Eleven a-Clock we 
 furl'd our Main-sail, and ballasted our Mizen ; at which time it began 
 to rain, and by Twelve a-Clock at Night it blew exceeding hard, and 
 the Rain poured down as through a Sieve. It thundered and lightned 
 prodigiously, and the Sea seemed all of a Fire about us, for every Sea 
 that broke sparkled like Lightning. The violent Wind raised the Sea 
 presently to a great heighth, and it ran very short, and began to break 
 in on our Deck. One Sea struck away the Rails of our Head ; and our 
 Sheet Anchor, which was stowed with one Flook or bending of the Iron 
 over the Ships Gunnal, and lasht very well down to the Side, was 
 violently washt off, and had like to have struck a hole in our Bow, as it 
 lay beating against it. Then we were forced to put right before the Wind 
 
 * Changeum Island, S.W. of the entrance to Canton River. 
 
OF THE LAW OF STORMS. 
 
 to stow our Anchor again, which we did with much ado; but after- 
 wards we durst not adventure to bring our Ship to the Wind again, for 
 fear of foundring, for the turning the Ship either to or fro from the 
 Wind is dangerous in such violent Storms. The Fierceness of the 
 Weather continued till Four a-Clock that Morning ; in which time we 
 did cut away two Canoas that were towing astern. 
 
 " After Four a-Clock the Thunder and the Rain abated, and then 
 we saw a Corpus Sant at our Maintop-mast Head, on the very top of 
 the Truck of the Spindle. This sight rejoiced our Men exceedingly, 
 for the height of the Storm is commonly over when the Corpus Sant is 
 seen aloft ; but when they are seen lying on the Deck, it is generally 
 accounted a bad Sign. * * * * 
 
 " We continued scudding right before Wind and Sea from Two till 
 Seven a-Clock in the Morning, and then the Wind being much abated, 
 we set our Mizen again, and brought our Ship to the Wind, and lay 
 under a Mizen till Eleven. Then it fell flat calm, and it continued so 
 for about two Hours; but the Sky looked very black and rueful, 
 especially in the S.W., and the Sea tossed us about like an Egg-shell, 
 for want of Wind. About One a-Clock in the Afternoon the Wind 
 sprung up at S.W. out of the Quarter from whence we did expect it : 
 therefore we presently brailed up our Mizen, and wore our Ship : But 
 we had no sooner put our Ship before the Wind, but it blew a Storm 
 again, and rain'd very hard, though not so violently as the Night before ; 
 but the Wind was altogether as boisterous, and so continued till Ten or 
 Eleven a-Clock at Night. All which time we scuded and run before 
 the Wind very swift, though only with our bare Poles, that is, without 
 any Sail abroad. Afterwards the Wind died away by degrees, and 
 before Day we had but little Wind, and fine clear Weather. 
 
 " I was never in such a violent Storm in all my Life ; so said all the 
 Company." 
 
 It would seem probable that the old voyager was fated to experience 
 another storm at the change of the next monsoon, which, if not alto- 
 gether of full typhoon force, was very cyclonic in its changes ; he was 
 then at an anchorage at one of the Bashees : " We had yet the Winds 
 at S.W. and S.S. W., mostly fair Weather. In October we did expect the 
 Winds to shift to the N.E., and therefore we provided to sail (as soon as 
 the eastern Monsoon was settled) to cruize off of Manila. Accordingly 
 we provided a Stock of Provisions. * * * * 
 
 " About the 24th Day of September the Winds shifted about to the 
 East, and from thence to the N.E., fine fair Weather. The 25th it came 
 at N. and began to grow fresh, and the Sky began to be clouded, and the 
 Wind freshned on us. 
 
 " At Twelve a-Clock at night it blew a very fierce Storm. We were 
 
6 HISTORY AND DEVELOPMENT 
 
 then riding with our best Bower a-head ; and though our Yards and 
 Top-masts were down, yet we drove. This obliged us to let go our 
 Sheet Anchor, veering out a good Scope of Cable, which stopt us till 
 Ten or Eleven a-Clock the next Day ; then the Wind came on so fierce 
 that she drove again, with both Anchors a-head. The Wind was now 
 at N. by W., and we kept driving till Three or Four a-Clock in the 
 Afternoon : And it was well for us that there were no Islands, Rocks, or 
 Sands in our way, for if there had, we must have been driven upon 
 them. We used our utmost Endeavours to stop here, being loth to go 
 to Sea, because we had six of our Men ashore, who could not get off 
 now. At last we were driven out into deep Water, and then it was in 
 vain to wait any longer : Therefore we hove in our Sheet Cable, and 
 got up our Sheet Anchor, and cut away our best Bower (for to have 
 heav'd her up then would have gone near to have foundred us), and so 
 put to Sea. We had very violent Weather the Night ensuing, with 
 very hard Rain, and we were forced to scud with our bare Poles till 
 Three a-Clock in the Morning. Then the Wind slacken'd, and we 
 brought our Ship to under a Mizen, and lay with our Head to the 
 Westward. The 27th Day the Wind abated much, but it rained very 
 hard all Day, and the Night ensuing. The 28th Day the Wind came 
 about to the N.E., and it cleared up, and blew a hard Gale, but it stood 
 not there, for it shifted about to the Eastward, thence to the S.E., then 
 to the South, and at last settled at S.W., and then we had a moderate 
 Gale and fair Weather. 
 
 " It was the 29th Day when the Wind came to the S. W. Then we 
 made all the Sail we could for the Island again. The 30th Day we had 
 the Wind at West, and saw the Islands. * * * * This 
 last Storm put our Men quite out of heart ; for although it was not 
 altogether so fierce as that which we were in on the Coast of China, 
 which was still fresh in Memory, yet it wrought more powerfully, and 
 frighted them from their Design of cruizing before Manila, fearing 
 another Storm there." (See DAMPIER'S New Voyage round the World, 
 chap. xv. pp. 413-416, also pp. 437-439 j ed. 1697.) 
 
 It may not be amiss to notice that DAMPIER gives an excellent account 
 of the Monsoons of the Malabar and Coromandel coasts, of the period 
 of change of those winds, and of the storms then prevalent ; it is also 
 not improbable that he encountered a cyclone, May 1688, between the 
 Nicobars and Sumatra. Though he never experienced a hurricane 
 during his stay among the West Indian Islands, he describes one from 
 hearsay : how "the wind came on very fierce at N.E., and veering about 
 toN. and N.W. settled there, bringing with it very violent Rains;" 
 how it subsequently " fell flat calm, and the Rain ceased j " how, during 
 
OP THE LAW OP STORMS. 7 
 
 the calm, the men went down to the ships and found them high and dry, 
 and how they made haste back "because the Wind began to blow hard 
 at S.W., and it blew so violently before they recover'd the House, that 
 the Boughs of the Trees whipt them sufficiently before they got thither, 
 and it rained as hard as before ; " that " when the N.E. Gust raged, the 
 Sea ebb'd so prodigiously, or else was driven off the shore by the violence 
 of the Wind so far, that some Ships riding in the Harbour in 3 or 
 4 fathom Water, were a-ground, and lay so till the S.W. Gust came, 
 and then the Sea came rowling in again with such prodigious fury, that 
 it not only set them a-float, but dash'd many of them on the shore." 
 The narrative relates to Muskito Cove, St. John's Harbour, Antigua, 
 in August 1681 ; and he informs us that the island suffered as well as 
 the ships, houses being blown down and trees uprooted, so that all looked 
 like winter : " neither did the fury of this Storm light only here, for 
 Nevis and St. Christophers had their shares also ; but Mountsurat felt 
 little of it." He does not forget to mention that these violent storms 
 rage " chiefly among the Carribee Islands, though by relation Jamaica 
 is annoyed by them," and that " they are expected in July, August, or 
 September" After giving a long account of the Northers and the 
 Souths of the Gulf of Mexico and Carribean Sea, he concludes as 
 follows : " The Clouds that precede a Hurricane are different from the 
 North Banks, in this, that whereas the Clouds preceding Norths are 
 uniform and regular, of an exact blackness even from the Horizon to 
 the upper edge of it, and that as streight and even as a Line stretched 
 out, on the contrary, the Hurricane Clouds tower up their Heads, 
 pressing forwards as if they all strove for precedency, yet so linked one 
 within another that all move alike. Besides, the edges of these Clouds 
 are guilded with various and affrighting Colours ; the very edge of all 
 seems to be of a pale fire colour, next that of a dull yellow, and nearer 
 the Body of the Cloud of a Copper Colour, and the Body of the Cloud, 
 which is very thick, appears extraordinary Black; and altogether it 
 looks very terrible and amazing, even beyond expression. Tho' I have 
 never been in any Hurricane in the West Indies, yet I have seen the 
 very Image of them in the East Indies, and the effects have been the 
 very same ; and for my part, I know no difference between a Hurricane 
 among the Carribee Islands in the West Indies and a Tuffoon on the 
 Coast of China in the East Indies, but only the Name : And I am apt 
 to believe that both Words have one signification, which is a violent 
 Storm" (DAMPIER'S Voyages and Descriptions, vol. ii. part iii.) 
 
 During the 17th and 18th centuries descriptions of the West Indian 
 hurricanes become numerous : the writers are chiefly English, Spanish, 
 and French ; and in various " dissertations on winds, " " theories of the 
 winds, " and " causes of hurricanes " appear many " new hypotheses," 
 
\ 
 8 HISTORY AND DEVELOPMENT 
 
 such as the effect of lunar, magnetic, or electric influence, etc. The veering 
 of the wind in a hurricane came, however, to be a recognised fact. A 
 Capt. LANGFORD has a paper in the Philosophical Transactions, 1698, 
 on West Indian hurricanes and their prognostics, in which he repeatedly 
 refers to them as " whirlwinds, " states how the winds veer, but evidently 
 misapprehended their progression ; and Sir GILBERT BLANE, describing 
 (Edinburgh Phil. Trans. 1788) the Barbadoes hurricane of Oct. 10, 
 1780, observes that "the wind blew all round the compass, a circum- 
 stance which distinguishes the hurricane from all other gales within the 
 tropics." 
 
 We now arrive at the 19th century, prior to which it cannot be said 
 that there was any investigation (properly so to speak) into the physical 
 character of these tropical storms, nor any generalization of facts. 
 
 In 1801 Colonel JAMES CAPPER published his Observations on Winds 
 and Monsoons ; in which work, speaking of the " tremendous gales, or 
 rather hurricanes," that blow, after the change of the monsoons, on the 
 Malabar and Coromandel coasts and in the Bay of Bengal, he states 
 that " all the circumstances, properly considered, clearly manifest the 
 nature of these winds, or rather positively prove them to be whirlwinds ; " 
 he indicates their probable size, their localities, and their seasons, 
 referring at the same time to the hurricane season at Mauritius and the 
 West Indies. But what is more significant than all, he indistinctly sees 
 that his investigations might have a practical application : " It would 
 not perhaps be a matter of great difficulty to ascertain the situation of 
 a ship in a whirlwind by observing the strength and changes of the wind. 
 If the changes are sudden and the wind violent, in all probability the 
 ship must be near the centre or vortex of the whirlwind ; whereas, if 
 the wind blows a great length of time from the same point, and the 
 changes are gradual, it may be reasonably supposed the ship is near the 
 extremity of it." 
 
 HORSBURGH, formerly so well known to seamen by his Directory, 
 published, in 1805, Memoirs comprising the Navigation to and from 
 China by the China Sea, in which he gives a long account of the 
 Ty-foongs ; he speaks of their " rotary " motion, describes the veering 
 of the wind according to their direction, notes the characteristics of their 
 approach, and advises the navigator how to mano3uvre when overtaken 
 in such storms. 
 
 ROMME in his Tableau des Vents, published in 1806, describes the 
 hurricanes and whirlwinds of the West Indies and Mexico, the tempests 
 of the Mozambique Channel and the Bay of Bengal, and the Tyfoons 
 of China, etc.; he notes the veering of the wind, and the whirlwinds 
 
OF THE LAW OF STORMS. 
 
 (tourbillons) of those regions, with their characteristichlglr-55a, cloudy 
 sky, and heavy rain. 
 
 Professor FARRAR, of Cambridge University, U. S., when describing 
 the storm that passed over Boston on Sept. 23, 1815, though unable 
 "to find the centre or limits of this tempest," estimates it at about 
 200 miles in diameter ; he remarks that " it was very violent at places 
 separated by a considerable interval from each other, while the inter- 
 mediate region suffered much less ; " and comes to the conclusion that it 
 was " a moving vortex, and not the rushing forward of the great body 
 of the atmosphere. " 
 
 About this period every storm of greater violence than usual had its 
 historian ; and speculations on the cause and physical character of the 
 phenomena were as conflicting as numerous. There were writers on the 
 subject in England, France, Germany, and the United States : some 
 describe, with considerable accuracy, the prognostics of hurricanes, and 
 their seasons ; others notice their progression in particular directions ; 
 later on, a few infer, like Professor E. MITCHELL in his Proximate Cause 
 of Certain Winds (1831), that "the phenomena of winds and storms are 
 the result of a vortex or gyratory movement * * * in the region 
 of the atmosphere where they prevail ; " the great and rapid oscillations 
 of the barometer during a hurricane had come to be a matter of frequent 
 and special remark ; while of theories respecting the origin of these 
 tempests, the electrical one, as in the case of tornadoes and waterspouts, 
 was most in favour. 
 
 As yet, however, nothing very definite or satisfactory had been done 
 towards elucidating the real character of hurricanes. That could only 
 be accomplished, as we now fully appreciate, by a close investigation of 
 numerous simultaneous observations extending over a wide area : this 
 was clearly perceived by the late WILLIAM C. REDFIELD, a naval archi- 
 tect in the United States, who published his first paper Remarks on the 
 Prevailing Storms of the Atlantic Coast of the North American States in 
 1831 ; but if report is correct, he had been engaged on the study of the 
 subject since 1821 a period of ten years. The conclusions at which he 
 arrived were the product of continuous observation combined with 
 experiment : when a hurricane occurred he collected whatever infor- 
 mation was to be obtained respecting it in its passage along the coast 
 and through the country, and for the oceanic area he procured extracts 
 from the log-books of ships that had been involved in it ; the mass of 
 facts was at first a source of much thought and speculation ; eventually, 
 having plotted on a chart of large scale the direction of the winds ob- 
 served at given positions on the same day and about the same hour in 
 fact, a series of nearly simultaneous observations made far and wide 
 
10 HISTORY AND DEVELOPMENT 
 
 he adopted a tentative process, and found that by placing on this chart 
 a transparency (tracing paper) marked in concentric circles, the majority 
 of the wind arrows (for a given day) were closely coincident with these 
 
 circles (see Fig. 1) ; trial 
 after trial gave the same 
 result, and he then became 
 convinced that the mass of 
 air constituting the hurri- 
 cane had a rotatory move- 
 ment around a central 
 calm. The experiment, 
 similarly tried at another 
 spot for the following day, 
 and so on for successive 
 days at different spots, 
 showed also that the storm- 
 field was not stationary, but 
 that it moved bodily for- 
 ^^^^^ ward. It was on the basis 
 
 of these empirical investigations that REDFIELD announced, in 1831, 
 that the winds constituting a hurricane had a rotatory movement 
 (i.e. blowing around a common centre), while the storm itself had a pro- 
 gressive motion ; and this was the foundation of the " Law of Storms" 
 
 REDFIELD'S views are fully embodied in a paper, published in 1833, 
 and entitled Observations on the Hurricanes and Storms of the West 
 Indies and the Coast of the United States, a portion of which it may not 
 be amiss to introduce here : 
 
 " From a careful attention to the progress and phenomena of some 
 of the more violent storms which have visited the Western Atlantic, 
 I have found that they exhibit certain characteristics of great uniformity. 
 This appears not only in the terminate course which these storms are 
 found to pursue, but in the direction of wind and succession of changes 
 which they exhibit while they continue in action. The same general 
 characteristics appear also to pertain, in some degree, to many of the 
 more common variations and vicissitudes of winds and weather, at 
 least in the temperate latitudes. The following points I consider as 
 established : 
 
 " 1. The storms of greatest severity often originate in the tropical 
 latitudes, and, not unfrequently, to the Eastward of the West India 
 Islands ; in the tropical regions they are distinguished by the name of 
 hurricanes. 
 
 " 2. These storms cover, at the same moment of time, an extent of 
 
OF THE LAW OF STORMS. 11 
 
 contiguous surface, the diameter of which may vary, in different storms, 
 from 100 to 500 miles, and in some cases they have been much more 
 extensive. They act with diminished violence towards the exterior, 
 and with increased energy towards the interior, of the space which they 
 occupy. 
 
 " 3. While in the tropical latitudes, or South of the parallel of 30 N., 
 these storms pursue their course, or are drifted by the natural atmo- 
 spheric current of the region towards the West, on a track which inclines 
 gradually to the Northward, till it approaches the latitude of 30 N. 
 In the vicinity of this parallel their course is changed somewhat abruptly 
 to the Northward and Eastward, and the track continues to incline 
 gradually to the East, towards which point, after leaving the lower 
 latitudes, they are found to progress with an accelerated velocity. 
 
 " The rate at which these storms are found thus to advance in their 
 course varies in different cases, but may be estimated at from 12 to 
 30 miles an hour. The extent to which their course is finally pursued 
 remains unknown ; but it is probable that, as they proceed, they become 
 gradually extended in their dimensions, and weakened in their action, 
 till they cease to command any peculiar notice. One of the hurricanes 
 of August 1830 has been traced in its daily progress from near the 
 Carribee Islands to the coast of Florida and the Carolinas, and thence 
 to the Banks of Newfoundland, a distance of more than 3000 miles, 
 which was passed over by the storm in about six days. The duration 
 of the most violent portion of this gale, at the different points over 
 which it passed, was about 12 hours, but its entire duration was, in 
 many places, more than twice that period. Another hurricane which 
 occurred in the same month passed from near the Windward Islands 
 on a more Eastern but similar route, and has also been traced in its 
 daily stages, by means of the journals and reports of voyages, near 
 2500 miles. The hurricane of August 1831, which desolated the Island 
 of Barbadoes on the 10th of that month, the daily progress of which 
 has also been ascertained, passed in nearly a direct course to the 
 Northern shores of the G-ulf of Mexico and New Orleans, where it 
 arrived on the 1 6th of the same month, having passed over a distance 
 of 2300 statute miles in six days after leaving Barbadoes. Many cases 
 of like character might be adduced. 
 
 " 4. The duration of the storm at any place within its track depends 
 upon its extent and the rate of its progressive velocity, as these circum- 
 stances are found to determine the time which is required for the storm 
 to pass over any given locality falling within its route. Storms of 
 smaller extent or dimensions are usually found to move from one place 
 to another with greater rapidity than larger storms. 
 
 " 5, The course thus pursued by the storm is found to be entirely 
 
12 HISTORY AND DEVELOPMENT 
 
 independent of the direction of wind which it may exhibit at the different 
 points over which it passes the wind in all such storms being found 
 to blow after the manner of a whirlwind, around a common centre or 
 vortex, during their entire progress, and in a determinate direction or 
 course of rotation, which is from right to left (or in the direction from 
 West to South) horizontally. The direction of the wind therefore, for 
 the most part, does not coincide with the course of the storm. 
 
 " 6. In the lower latitudes, while drifting to the "Westward, the 
 direction of the wind at the commencement, or under the most advanced 
 portion of these storms, is from a Northern quarter, usually at some 
 point from North-east to North-west, and during the latter part of the 
 gale it blows from a Southern quarter of the horizon, at all places where 
 the whole effect of the gale is experienced. 
 
 " 7. After reaching the more Northern latitudes, and while pursuing 
 their course to the Northward and Eastward, these storms commence 
 with the wind from an Eastern or Southern quarter, and terminate 
 with the wind from a Western quarter, as will appear more distinctly 
 under the three following heads, the latter portion of the storm being 
 usually attended with broken or clear weather. 
 
 " 8. On the outer portion of the track, North of the parallel of 30 N., 
 or within that portion of it which lies furthest from the American coast, 
 these storms exhibit at their commencement a Southerly wind, which, as 
 the storm comes over, veers gradually to the Westward, in which quarter 
 it is found to terminate. 
 
 "9. In the same latitudes, but along the central portion of the track, 
 the first force of the wind is from a point near to South-east, but after 
 blowing for a certain period it changes suddenly, and usually, after a 
 short intermission, to a point nearly or directly opposite to that from 
 which it has previously been blowing, from which opposite quarter it 
 blows with equal violence till the storm has passed over or has abated. 
 This sudden change of a South-easterly wind to an opposite direction 
 does not occur towards either margin of the storm's track, but only on its 
 more central portion, and takes effect in regular progression along this 
 central part of the route, from the South-west towards the North-east, 
 in an order of time which is exactly coincident with the progress of the 
 storm in the same direction. It is under this portion of the storm that 
 we notice the greatest fall of the barometer, and the mercury usually 
 begins to rise a short time previous to the change of wind. In this part 
 of the track the storm is known as a South-easter, and is usually attended 
 with rain previous to the change of wind, and perhaps for a short time 
 after. 
 
 "10. On that portion of the track which is nearest the American 
 coast, or which is furthest inland if the storm reaches the continent, the 
 
OF THE LAW OF STORMS. 13 
 
 wind commences from a more Eastern or North-eastern point of the 
 horizon, and afterwards veers more or less gradually, by North, to a 
 North-western or Westerly quarter, where it finally terminates. Here 
 also the first part of the storm is usually, but not always, attended with 
 rain, and its latter or Western portion with fair weather. The first or 
 foul- weather portion of the storm is, on this part of its track, recognised 
 as a North-easter. 
 
 " It should be noted, however, that near the latitude of 30 N., and 
 on the shores of Carolina, where the storm enters obliquely upon the 
 coast, while its track is rapidly changing from a Northwardly to an 
 Eastwardly direction, the wind on the central track of the storm will 
 commence from an Eastern or North-eastern point of the compass, 
 and will gradually become South-easterly as the storm approaches its 
 height. 
 
 "11. A full and just consideration of the facts which have been 
 stated will show conclusively that the portion of the atmosphere which 
 composes for the time being the great body of the storm, whirls or blows, 
 as above stated, in a horizontal circuit around a vertical or somewhat 
 inclined axis of rotation which is carried onward with the storm ; that 
 the course or direction of this circuit of rotation is from right to left ; 
 and that the storm operates nearly in the same manner as a tornado or 
 whirlwind of smaller dimensions the chief difference being in the more 
 disc-like form of the whirling body and the magnitude of the scale of 
 operation.* This view of the subject, when fully comprehended, affords 
 a satisfactory solution of the otherwise inexplicable phenomena of storms, 
 and will also be found to accord entirely with the fact which appears in 
 the above statement, that in the phases or changes which pertain to a 
 storm, the wind, on one margin of its track, veers with the sun, or from 
 left to right, while under the opposite margin of the same storm it veers 
 against the sun, or from right to left ; for this peculiarity necessarily 
 attends the progressive action of any whirlwind which operates hori- 
 zontally. 
 
 "12. Owing to the centrifugal action of these rotative storms, the 
 barometer, whether in the higher or lower latitudes, always sinks while 
 under the first portion or moiety of the storm on every part of its track, 
 excepting perhaps its extreme outward margin, and commonly affords 
 us the earliest and surest indication of the approaching tempest. The 
 mercury in the barometer always rises again during the passage of the 
 
 * It is to be understood that the diameter of the whirlwind which constitutes the storm is com- 
 mensurate with the width of the track over which the storm passes. The main body of the storm is 
 supposed to move in the form of an extensive disc, whirling around its own centre as it advances in 
 its regular track with this difference, that the rotative movement is far more rapid in the interior 
 portions of the whirling body than towards its exterior limits. 
 
14 HISTORY AND DEVELOPMENT 
 
 last portion of the gale, and commonly attains the maximum of its 
 elevations on the entire departure of the storm. 
 
 " The great value of the barometer to navigators is becoming well 
 understood, and its practical utility might be greatly increased by hourly 
 entries of the precise height of the mercurial column, in a table pre- 
 pared for the purpose. Its movements, unless carefully recorded, often 
 escape notice or recollection ; which may easily happen at those times 
 when a distinct knowledge of its latest variations might prove to be of 
 the greatest importance. 
 
 " In the foregoing statements our design has been to designate in a 
 summary manner the principal movements which, in these regions at 
 least, constitute a storm ; and we do not attempt to notice the various 
 irregularities and subordinate or incidental movements and phenomena 
 of the atmosphere with which a storm may chance to be connected, or 
 which may necessarily result from such violent movements in a fluid 
 which is so tenuous and elastic in its character. It may be remarked in 
 general, that the most active or violent storms are usually the most 
 regular and uniform in the development of those characteristic move- 
 ments which we have already described ; it is also probable that the 
 vortex or rotative axis of a violent gale or hurricane oscillates in its 
 course with considerable rapidity, in a moving circuit of moderate 
 extent, near the centre of the hurricane ; and such an eccentric move- 
 ment of the vortex may, for aught we know, be essential to the con- 
 tinued activity or force of the hurricane. Such a movement will fully 
 account for the violent flaws or gusts of wind, and the intervening lulls 
 or remissions which are so often experienced towards the heart of a 
 storm or hurricane, when in open sea ; but of its existence we have no 
 positive evidence. 
 
 " It has been stated, ' that while one vessel has been lying-to in a 
 heavy gale of wind, another not more than 30 leagues distant has at 
 the very same time been in another gale equally heavy, and lying-to 
 with the wind in quite an opposite direction.' 
 
 "This statement is obviously to be understood as applicable to 
 two vessels falling under the two opposite sides or portions of the 
 same storm, where the wind, in its regular circuit of rotation, must of 
 course blow from the opposite quarters of the horizon. We will sup- 
 pose one of the vessels to be at A, and the other at B, in the annexed 
 figure. The storm, in pursuing its course from W. towards N., will 
 strike the first-mentioned vessel in the direction which is shown by 
 the wind arrows at the point c, which, if the position be in the tem- 
 perate latitudes, North of 30, will be from Eastward. Now it is 
 obvious that, as the storm advances in its course North-eastward, this 
 vessel, if nearly stationary, will intersect the body of the gale on the 
 
OF THE LAW OF STORMS. 15 
 
 line c A d. As the storm advances the wind must also veer to the 
 Northward, as shown by the arrows, being at N.E. when the vessel is 
 
 brought under the point 
 A; and near the close or 
 departure of the storm by 
 its further progress East- 
 ward, the wind will have 
 further veered to the direc- 
 tion shown at d, which, 
 with due allowance for the 
 progressive motion of the 
 storm, we will set down at 
 N.N.W. The other vessel, 
 as is equally obvious, will 
 first take the wind from 
 the Southward, as shown 
 at e, in which quarter it 
 will blow with no great 
 variation till, by the advance of the storm, the ship is brought under 
 the point B. The barometer, which had previously been falling, will 
 now commence rising, and the wind, veering more Westerly, will at the 
 departure of the storm be found in the direction shown at /, which, 
 after the allowance already referred to, may be stated at W.N.W. 
 Such, substantially, are the facts commonly reported by vessels which 
 fall under the lateral portions of the Atlantic storms ; and it is readily 
 seen that the opposite winds, which are exhibited on the two different 
 intersections of the storm as above described, will very naturally be 
 mistaken for two separate and distinct gales. 
 
 " The phases^ of the wind in these gales are, however, in all cases 
 modified, more or less, by the course or changing position of the vessel 
 exposed to its action. For example : A ship on taking the gale say 
 at E.S.E. at the point h on the figure, and lying-to with her head to the 
 Northward may by that means be brought to intersect the storm on 
 the line h i, and at the point i would suddenly be taken aback with the 
 wind, say at N.N.W., as in the case of the Jamaica homeward-bound 
 fleet in 1782; and the barometer, which reaches its lowest depression 
 under the central portion of the storm, would about this period be found 
 to have commenced rising with some degree of rapidity. 
 
 " A further reference to the figure will show that a ship, which may 
 be at the point G- during the passage of the gale, would be exposed to a 
 heavy swell from the Southward and Westward ; but, being beyond the 
 organized limits of the storm, may remain entirely unaffected by the 
 violence of the wind, which at the same time may be raging with 
 
16 HISTORY AND DEVELOPMENT 
 
 destructive fury at the distance of a few leagues. The writer has 
 knowledge of many such examples. 
 
 " It has been suggested that the ' port tack is the proper one to 
 lie-to on, as the wind will then be found to draw aft;' but this will 
 frequently prove erroneous, as the wind may draw either way on either 
 tack, according to the position and course of the ship, in the storm, and 
 the extent and rate of progress of the latter. In the case of the fleet 
 which encountered the gale of 1782, it was probably the best course to 
 carry sail to the Northward at the very commencement of the gale, and 
 as far and as long as possible. By this means the fleet might perhaps 
 have been drawn as far Northward as the point A on the figure, and the 
 change of wind to the Northward and Westward would have been 
 rendered more gradual. The chief difficulty and danger is when the 
 direction of the wind, at the first setting in of the gale, is found to be 
 nearly at right angles with the known course of the storms in the region 
 where the gale is encountered, and it is then desirable to pursue such 
 a course as to avoid, if possible, falling into the heart of the storm. 
 
 " It frequently happens that a storm, during the first part of its 
 progress over a given point, fails to take effect upon the surface, while 
 it exhibits its full activity at a greater altitude. This commonly happens 
 when this portion of the storm arrives from, or has recently blown over, 
 a more elevated country, or is passing or blowing from the land to the 
 sea. On land the most violent effects are usually felt from those storms 
 which enter and blow from the open ocean upon the shores of an island 
 or continent. Upon the latter, under such circumstances, the first part 
 of the gale is usually the most severe, and that coast of an island upon 
 which a storm first enters or blows also suffers most from the early 
 part of the gale; but its later or receding part often acts with the 
 greatest fury upon the opposite side of the island, which had previously 
 derived some degree of shelter from the intermediate elevations and 
 other obstacles opposed to the force of the wind, the benefit of which is 
 now lost by its counter-direction from the open ocean. Owing to similar 
 causes, the force of the storm is sometimes very unequal at different 
 places, situated in nearly the same part of its track ; and such inequality, 
 as we have before intimated, necessarily pertains to two places one of 
 which is near the centre, and the other towards the margin of the route." 
 
 REDFIELD'S remarks are accompanied by the tracks of eleven "West 
 Indian hurricanes, and he observes further : 
 
 " The dimensions of these storms appear to gradually expand during 
 their course. * * * The track of most hurricanes appears to form 
 part of an elliptical or parabolic circuit. We are also struck with the 
 fact that the vortex of the curve is uniformly found on or near the 30th 
 degree of latitude. 
 
OF THE LAW OF STORMS. 17 
 
 " The typhoons and storms of the China Sea and Eastern coast of 
 Asia appear to be similar in character to the hurricanes of the West 
 Indies and the storms of this (American) coast when prevailing in the 
 same latitudes. There is reason to believe that the great circuit of wind, 
 of which the Trade Winds form an integral part, are nearly uniform in 
 all the great oceanic basins ; and that the course of these circuits, and of 
 the stormy gyrations which they may contain, is in the Southern Hemi- 
 sphere, in a counter-direction to those North of the equator, producing a 
 corresponding difference in the general phases of storms and winds in 
 the two hemispheres."* 
 
 The Law of Storms as thus developed by REDFIELD was not, how- 
 ever, permitted to pass unchallenged \ but his chief opponents were his 
 own countrymen, and to these it will be necessary to refer in the sequel. 
 
 While REDFIELD was busy in America there were active enquirers 
 in the field of research in Europe ; and the first important work published 
 on the subject in England was by Lieut.-Col. W. REID in 1838 An 
 Attempt to develop the Laws of Storms by means of Facts, arranged 
 according to Place and Time, etc. a third edition of which appeared in 
 1850. 
 
 As Governor of the Bermudas, Col. (afterwards Sir W.) REID was 
 well placed for personal observations and investigation, as the islands lie 
 in the track of hurricanes. He not only confirmed REDFIELD'S views 
 that those storms in the Northern Hemisphere are vast whirlwinds with 
 a progressive motion, revolving by a fixed law but, by collecting infor- 
 mation concerning the Mauritius gales, he proved that they revolve 
 in opposite ways on opposite sides of the Equator, thus also corroborat- 
 ing what REDFIELD had already announced as probable. It may be 
 said, without any undervaluing of the labours of others, that most of 
 the really useful practical rules in storm-sailing are due to REID, and 
 that he never lost sight of the practical application of the subject to navi- 
 gation. Besides reports and papers in various Journals, he published in 
 1849 a work on the Progress of tht Development of the Law of Storms 
 and of Variable Winds. 
 
 HENRY PIDDINGTON, formerly President of the Calcutta Marine 
 Courts, and so well known to seamen by the Sailor's Horn-Book for the 
 Law of Storms, appeared in the field in 1840. He was a careful 
 observer, and an indefatigable collector of facts respecting hurricanes. 
 In addition to twenty-four Memoirs published in the Jour. Asiatic Soc. 
 of Bengal, he wrote several pamphlets and short papers. He was a 
 
 * REDFIELD, although he never published a separate work on the Law of Storms and its applica- 
 tion, was a frequent contributor to various Magazines, and to the Proceedings of the American 
 Society for the Advancement of Science. These memoirs and papers amount in number to thirty- 
 three; his last observations were on the "Cyclones of the Western Pacific," written for Admiral 
 PERRY'S Japan Expedition shortly before his death, in 1857. 
 
 B 
 
18 HISTORY AND DEVELOPMENT 
 
 follower of REDFIELD and REID, but was more popular than either 
 perhaps from collating his material better, and being on the whole more 
 discursive. 
 
 Another observer and writer must not be omitted. ALEXANDER 
 THOM, Surgeon, when stationed with his regiment at Mauritius, gathered 
 the materials for an Inquiry into the Nature and Course of Storms in the 
 Indian Ocean south of the Equator, which was published in 1845. 
 While agreeing in the main with REDFIELD and REID on the law of 
 storms, and the practical rules laid down by the latter, he entirely 
 differs from PIDDINGTON in his theory. It may, however, be stated 
 that, on the whole, the four writers are more practical than theoretical. 
 It is unnecessary to refer at present to the views of later meteorologists. 
 
 REDFIELD undoubtedly arrived at his results by an independent 
 course of investigation ; but another was at work at the same time in 
 the same direction. Professor H. W. DOVE, of Berlin, in an article on 
 Barometric Minima, published in 1828, had come to the conclusion that 
 hurricanes, accompanied by great depressions of the barometric column, 
 are vast whirlwinds moving in the Northern Hemisphere from S.W. to 
 N.E., and on this basis claimed priority of discovery of the Law of 
 Storms. 
 
 From a practical point of view, however, REDFIELD was unquestion- 
 ably first in the field. On the other hand, to DOVE belongs the merit 
 of having been the first to expound and elucidate the rotatory theory, on 
 the principle that it is the simple and necessary consequence of the earths 
 rotation : this he fully developed in his important work, Das Gesetz der 
 Sturme (the Law of Storms), extracts from which in so far as they 
 immediately relate to and have a direct practical bearing on the sub- 
 ject may be appropriately condensed into a few pages : 
 
 The earth is a globe turning on its axis from West to East, and the 
 velocity of rotation at different places on the surface varies as the 
 diameter of the parallels on which the places are situated ; thus this 
 velocity increases from zero at the poles to its maximum at the Equator.* 
 Now the air, even when apparently motionless, partakes of the velocity 
 of rotation of the part of the globe over which it rests : hence, if a 
 difference of temperature, or any other cause, gives it a tendency to 
 move in a direction along a parallel, its motion cannot be affected by 
 the earth's rotation, because the points on the earth's surface towards 
 which it is moving have exactly the same velocity of rotation as those 
 which it has left. If, however, through any cause the air is set in 
 motion from the poles towards the equator, it passes from places with 
 less, to others with greater, rotatory velocity ; and consequently having 
 
 * The rotatory velocity of a place on the Equator is 15 geog. miles per minute ; in lat. 45* it is 
 10 J miles ; in lat. 51 about 9J miles ; in lat. 60* about 7 miles ; and in lat. 70* about 5 miles. 
 
UNIVERSITY 
 
 OF THE LAW OF STORMS. 19 
 
 less velocity towards the East than the places at which it arrives, it 
 appears to be moving in the contrary direction, viz. from East to West. 
 The deviation of the wind from its direction at starting, supposing 
 its rate of progression to remain unaltered, will be greater, the greater 
 the difference between the velocity of rotation of the point whence the 
 wind set out and that where it is observed, or, in other words, the 
 greater the difference of latitude between the two places. From this it 
 results that 
 
 1. In the Northern Hemisphere winds, which at starting are North, 
 change gradually, through N.E., towards East. 
 
 2. In the Southern Hemisphere winds, which at starting are South, 
 change gradually, through S.E., towards East. 
 
 3. In the Northern Hemisphere a southerly wind, in its progression, 
 gradually veers, through S.W., towards West. 
 
 4. In the Southern Hemisphere a northerly wind, in its progression, 
 gradually veers through N.W. towards West. 
 
 A West wind in both hemispheres will interfere with fresh equatorial 
 currents, and bring them to a state of relative rest. If the tendency 
 towards the pole continues, the same thing will always be repeated until 
 fresh polar currents change the West wind in the Northern Hemisphere, 
 through N.W. into North, and in the Southern Hemisphere, through 
 S.W. into South. This gives 
 
 For the Northern Hemisphere, the change through S., S.W,W,N.W,N. 
 For the Southern Hemisphere, the change through N.,N.W.,W., S.W., S. 
 
 And hence the following results may be deduced : 
 
 (A) In the Northern Hemisphere, when polar and equatorial currents 
 
 alternately succeed each other, the wind veers in general round 
 the compass in the direction 
 
 a> > s -> S.W, W, N.W, N, N.E, E, S.E, S. 3fr > 
 and the change takes place oftener between S. and W. and be- 
 tween N. and E. than between W. and N. and between E. and S. 
 
 (B) In the Southern Hemisphere, when polar and equatorial currents 
 
 alternately succeed each other, the wind veers in general round 
 the compass in the direction 
 
 a> > S, S.E, E, N.E, N, N.W, W, S.W, S. Jfr > 
 and the change takes place oftener between N. and W. and be- 
 tween S. and E. than between W. and S. and between E. and N. 
 
 This is DOVE'S " Law of Gyration " (Gesetz der Drehung), or the 
 Law of the Rotation of the Winds. 
 
20 HISTORY AND DEVELOPMENT 
 
 According to this law, the veering of the wind-vane may merely 
 indicate the existence of a steady current of air ; and it is a misappre- 
 hension of this fact that causes the existing, or constantly recurring, 
 confusion with reference to questions bearing on the theory of the winds. 
 The essential difference between the veering of the wind-vane as pro- 
 duced by a steady breeze and that resulting from a centripetal current, 
 or a whirling motion with an advancing centre, is, however, that in the 
 first case (i.e. with a steady breeze) the rotation is always in the same 
 direction, but in the case of the second (i.e. in a whirlwind or hurricane) 
 it veers on both sides of the track, and in opposite directions. 
 
 If therefore, as regards the Northern Hemisphere, the veering in the 
 direction S., W., N., E., be called "with the sun," or direct, and the 
 veering in the direction S., E., N., W., be called " against the sun," or 
 retrograde, as is customary; and similarly, in the Southern Hemisphere, 
 the veering through S., E., N., W., be called "with the sun,"* and the 
 veering through S., W., N., E., be called "against the sun,"* it follows 
 that 
 
 Steady winds turn the vane only in a direct sense, that is, with the 
 sun. 
 
 Whirlwinds, or centripetal gales, if they have a progressive motion, 
 turn the vane either with or against the sun, according to the side 
 on which the centre passes the place of observation. Finally 
 
 When steady winds from different directions approach each other, 
 rotations ensue in both ways : direct when, on the west side of 
 the compass, a wind is more northerly than that which preceded 
 it, retrograde, when it is more southerly. On the east side, on 
 the contrary, the gyration is direct when the displacing wind is 
 more southerly than that which preceded it, retrograde, when it 
 is more northerly. Winds from directly opposite points may 
 check each other and produce a calm, and thereby give rise to the 
 same phenomenon as is observed at the centre of a series of 
 centripetal currents or of a whirlwind, viz. winds diametrically 
 opposed separated by a calm. 
 
 Thus the same appearances may arise under totally different con- 
 ditions, and it is only by means of a careful investigation of all the 
 phases of a phenomenon that a decision can be pronounced on any one 
 
 * These expressions dift'er from the seaman's view : the movement from left to right, as being 
 " with the sun," originated in the Northern Hemisphere, beyond the tropic, where the sun culminates 
 in the South, and hence, looking in that direction, the sun rises on the left and sets on the right 
 hand i.e. left to right, " with the sun ; " this idea the seaman carries with him over every part of 
 the globe. But left to right, " with the sun," is erroneous in the Southern Hemisphere, where an 
 observer sees the sun rise on the right hand, culminate in the North, and set on the left hand. The 
 expression in the text is therefore correct, though not in the seaman's sense. 
 
OF THE LAW OF STORMS. 21 
 
 phase which presents itself in a particular instance. In these researches 
 the barometer is one of the chief guides. 
 
 The following brief remarks (pp. 21-29) are a general summary of 
 the result of DOVE'S investigations into the Law of the Rotation of the 
 Winds and the Law of Storms*: 
 
 (1.) All steady winds are modified by the earth's rotation in such 
 manner that equatorial currents receive a westerly deflection, 
 and polar currents an easterly deflection. 
 
 The N.E. and the S.E. Trade winds are steady polar currents. 
 
 The Monsoons are alternations of a polar and equatorial 
 current, according to the season of the year; therefore they 
 are N.E. and S.W. in the Northern Hemisphere, and S.E. and 
 N.W. in the Southern Hemisphere. 
 
 (2.) Bodies of air which, from a state of rest, are set in motion in the 
 
 direction of the meridian, turn the wind-vane as follows : 
 The polar current in the northern hemisphere, from N. to E. 
 The polar current in the southern hemisphere, from S. to E. 
 The equatorial current in the northern hemisphere, from S. to W. 
 The equatorial current in the southern hemisphere, from N. toW. 
 
 Hence, generally, in the Northern Hemisphere 
 
 Winds between N. and E. are the polar current, and 
 Those between E. and S. mark the transition of the polar 
 
 into an equatorial current : 
 
 Winds between S. and W. are the equatorial current, and 
 Those between W. and N. mark the transition of the 
 
 equatorial into a polar current. 
 
 In the Southern Hemisphere 
 
 Winds between S. and E. are the polar current, and 
 Those between E. and N. mark the transition of the polar 
 
 into an equatorial current : 
 
 Winds between N". and W. are the equatorial current, and 
 Those between W. and S. mark the transition of the equa- 
 torial into a polar current. 
 
 Thus a complete rotation in the Northern Hemisphere is 
 ]H> > S., W., N., E., S., )]> > with the sun; 
 
 And in the Southern Hemisphere 
 
 ) > S., E., K, W., S., )^> > with the sun; 
 
 * " Pas Gesets der Stiirme," pp. 187-192. 
 
22 HISTORY AND DEVELOPMENT 
 
 (3.) The course of a steady wind may be obstructed from following 
 the deflection impressed on it by the rotation of the earth : 
 
 (a) By a wind blowing constantly in a direction perpendicular 
 
 to its primitive : such are the West Indian hurricanes, 
 which, for this reason, advance first from S.E. to N.W., 
 while those of the Southern Hemisphere advance from 
 N.E. to S.W. 
 
 ( b) By a less deflected aerial current : such is the origin of 
 
 typhoons during the S.W. monsoon, which, further to 
 the eastward, is bounded by the South monsoon. Of 
 the typhoons, however, that have a progressive motion 
 from W. to E., it is probable that they result from the 
 heavy air of the Trade-wind region situated to the east- 
 ward rushing laterally through the rarefied air of the 
 region of the S.W. monsoon, and thus producing a whirl- 
 wind. 
 
 (c) By a mechanical impediment; as in the instance of a 
 storm (described by PIDDINGTON) which occurred in the 
 Bay of Bengal, June 1839.* 
 
 In these cases, when the storm is an equatorial one, whirlwinds 
 are produced in the Northern Hemisphere in a direction contrary 
 to that in which the hands of a watch move ; and in the Southern 
 Hemisphere, in the same direction as the hands of a watch move. 
 
 Hence in the Northern Hemisphere the following shiftings of 
 the wind-vane occur : 
 
 (a) If the centre of the storm passes westward of the place 
 
 of observation 
 
 The rotation will be ^> > S., W., N., E., S. ^> > with 
 the sun ; 
 
 (b) If the centre of the storm passes eastward of the place 
 
 The rotation will be )^> > S., E., K, W., S. >J|> > 
 
 against the sun. 
 
 But in the Southern Hemisphere, on the contrary, when the 
 storm is an equatorial one : 
 
 (a) If the centre of the storm passes westward of the place of 
 
 observation 
 
 The rotation will be )]. > S., E., N., W., S. )^> > 
 with the sun ; 
 
 * When a strong monsoon wind, " blowing in with tolerable steadiness from one quarter of the 
 compass," was diverted from its course by the mountains of Arraean, and became "a hurricane, i.e. 
 a violent wind blowing in a circle or vortex of greater or less diameter." 
 
OF THE LAW OP STORMS. 23 
 
 (b) If it passes eastward of the place 
 
 The rotation will be J^ > S., W., K, E., S. J^ > 
 against the sun. 
 
 Thus, in both hemispheres, an Equatorial whirlwind storm 
 passing on the west side of the place of observation produces 
 normal rotations, i.e. such as are in accordance with the law 
 of gyration; but its passage on the east side of the same 
 place produces anomalous rotations, i.e. such as are contrary 
 to the law of gyration. In Polar whirlwind storms the 
 effect is precisely opposite to that just described : the 
 rotation is normal when the centre passes to the eastward, 
 and anomalous when it passes to the westward, of the place 
 of observation. 
 
 The old rule, that " anomalous shif tings of the wind-vane 
 occur in stormy weather," is in this manner justified. It 
 cannot, however, be ascertained from those shif tings whether 
 an equatorial whirlwind passes on one side, or a polar one 
 on the other; this depends on the initial point of the 
 shifting, and, as a consequence of the rotation, can never 
 amount to more than half a circle. 
 
 (4.) The wind- vane may shift from one direction to that immediately 
 opposite to it 
 
 (a) When steady winds oppose each other, i.e. fight with each 
 other, as the seaman says ; and 
 
 (b) When the centre of a whirlwind storm passes over the 
 
 place of observation. 
 
 (5.) A wind may be part of a whirlwind storm even when the wind- 
 vane does not shift at the place of observation. This happens 
 when the place is so situated as to be only on the outer edge 
 of the whirlwind. On one side the storm then appears to 
 retrograde, and on the other to advance : if, for instance, 
 a whirlwind (revolving contrary to the motion of watch hands) 
 is advancing from S.W. to N.E., on the north-west side of 
 such a storm, a N.E. wind is experienced in the south-westerly 
 regions earlier than in the north-easterly, whilst on the south- 
 east side of the storm a S.W. wind actually appears to be 
 advancing, and so it enters the north-easterly regions at a 
 later period. The first is the well-known case observed 
 (in 1742) by FRANKLIN during an eclipse of the moon.* 
 
 * This was the hurricane of Oct. 21, 1742. FRANKLIN, at Philadelphia, was unable to observe the 
 eclipse of the moon that occurred that evening, but it was seen at Boston. The N.W. side of the 
 storm, with its N.E. wind, readied Philadelphia about 7 p.m., and subsequently (at 11 p.m.) appeared 
 at Boston, about 400 miles N.E. of the former place. 
 
24 HISTORY AND DEVELOPMENT 
 
 Those who, founding their theory on this and analogous 
 instances, divide the winds into positive (by aspiration) and 
 negative (by impulsion), call one side of such a storm positive, 
 and the other negative. 
 
 (6.) In the temperate zones storms are caused, also, by two oppo- 
 sitely directed currents of air in juxtaposition one pressing 
 laterally on the other ; then, during the transition, the shift- 
 ing of the wind- vane may be either with or against the sun, 
 but always according to the direction of one or other of the 
 currents impinging on each other. 
 
 (7.) Local phenomena, as land and sea breezes, counter currents 
 through valleys, deflections of the Trade winds by the coast, 
 trombs,* etc., affect the wind- vane according to local conditions, 
 and these may be such as either to produce a veering of the 
 vane in accordance with, or contrary to, the law of rotation. 
 In open regions this occurs periodically every day, at the 
 season when the general air currents (Trades) are less pre- 
 dominant, as is the case also in the region of the " variables " 
 between the Trades, at the time of the change of the 
 monsoons, and generally in summer, when the ascending 
 current (courant ascendant) weakens the force of the horizontal 
 currents. Indeed it is not impossible that, setting aside local 
 influences, one cause of a periodical shifting of the wind-vane 
 originates in the daily course of the sun : for example, if the 
 locality where the daily maximum of temperature occurs forms 
 a point of attraction for bodies of air in its vicinity, then will 
 the direction of the wind during the morning be westerly, and 
 during the afternoon, easterly, and thus the wind- vane will 
 turn in a direction contrary to the movement of the magnetic 
 needle. 
 
 Now whirlwinds not being confined to a definite locality, the pro- 
 bability that any given place may be situated on the east side of their 
 course is just as great as that it should be on the west side ; the same is 
 also the case where whirlwinds owe their origin to local causes. The 
 predominance of the rotation of the wind in a definite direction (with 
 the sun) is consequently a phenomenon unconnected with the whirling 
 motion of the storm-winds, but due only to the influence of the earth's 
 rotation on progressive steady winds. 
 
 * It has been observed that trombs, or small local whirlwinds, being due to local causes and not 
 being affected by the earth's rotation, turn indiscriminately, some, one way and some the other, and 
 depend on no rule producing uniformity of motion. 
 
OF THE LAW OP STORMS. 25 
 
 Now three things are possible : 
 
 1. Either all rotations of the wind-vane through great arcs are caused 
 
 by whirlwinds ; and in this case 
 
 (a) If these whirlwinds occur arbitrarily, now at one place and 
 then at another, there is no predominating direction in the 
 shifting of the vane ; or, 
 
 (6) If the origin of these whirlwinds is local, and if, in their 
 progress they pursue paths more or less constant, then 
 there are, at certain places in each hemisphere, predominant 
 shif tings of the wind-vane in the direction S., E., N., W., 
 and at others in the direction S., W., N., E. 
 
 2. Or, all shif tings of the wind- vane have their origin in alternations or 
 
 displacements of steady meridional air currents, in accordance with 
 the principles of Hadley's theory of the Trade-winds, and in 
 consequence of opposing currents pushing through each other ; in 
 this case the rotation is in the direction S., E., N., W., in the 
 northern hemisphere, and S., W., N., E. in the southern 
 hemisphere, i.e. in both cases the wind veers with the sun. The 
 backing of the wind cannot, under these conditions, exceed a 
 quadrant of the circle. 
 
 3. Or, lastly, the shiftings of the wind-vane arise in both ways viz. 
 
 from the alternation and displacement of meridional air currents, 
 and from a progressive whirlwind; consequently, in each hemi- 
 sphere, rotations of both kinds must occur, but those "with the 
 sun" will predominate. In both hemispheres, the first cause 
 produces only rotations with the sun, and the second produces just 
 as many rotations with the sun as against the sun. Hence (however 
 frequent whirlwinds may be) the rotations that are direct must 
 always be in excess of those that are retrograde. 
 
 Research has shown that whirlwind storms (hurricanes or cyclones), 
 originate at certain places, and pursue certain directions ; for instance 
 
 (a) WEST INDIA HURRICANES (Aracan or Huiran-vucan of the 
 coast of Mexico, Vuthan of Patagonia) commence on the 
 inner edge of the N.E. Trade wind, and even within the 
 Trade wind itself, especially in the latter part of summer 
 and in autumn; and their whirling motion is caused 
 originally by the S.E. Trade encountering the N.E. Trade, 
 or by portions of the upper Trade wind descending from 
 above. They progress from S.E. to N.W. in the torrid 
 zone, then, recurving at right angles at the outer edge of 
 the Trade wind, they move from S.W. to N.E. ; during the 
 
26 HISTORY AND DEVELOPMENT 
 
 progression, the whirlwind, revolving in a direction 
 opposite to that in which the hands of a watch move, 
 expands very considerably. In the region of the S.E. 
 Trade wind of the Atlantic, whirlwind storms occur with 
 less frequency. 
 
 (b) TYPHOONS in the northern Indian Ocean and China Sea occur 
 most frequently in autumn, but are also violent at the 
 commencement of the S.W. monsoon. The direction in 
 which they advance is rather more from E. to W. than 
 from N.E. to S.W., especially on the south coast of China. 
 In the southern half of the Indian Ocean they are likewise 
 very violent, progressing there from N.E. to S.W., and 
 recurving at right angles at the outer edge of the Trade 
 wind, whence they advance from N.W. to S.E. The rota- 
 tion of the air in the whirlwind storms which occur in the 
 Northern Hemisphere is in the opposite direction to that in 
 which the hands of a watch move, and in the direction 
 with watch hands in the Southern Hemisphere. The causes 
 which produce these effects lie in the juxtaposition of the 
 regions of the monsoons and Trade winds, as well as in the 
 displacement of one monsoon by the other. 
 
 Under certain local conditions, a whirlwind storm may be so modified 
 as to assume at particular spots the form of a steady wind, although, 
 when its course is observed over a considerable distance, it shows itself 
 to be a real whirlwind. This is the case, for instance, according to 
 REDFIELD, with the Northers of the Gulf of Mexico, which, especially 
 from September to March, at Yera Cruz, reach their greatest height 
 four hours after they commence, and then blow for 48 hours with 
 undiminished fury. REDFIELD explains this long duration of the storm 
 from one direction, by the whirlwind, when advancing from the east- 
 ward, being arrested by the high land, and as a consequence flattened, 
 so that its circular form is partially changed into a rectilinear one. The 
 N.E. and N.N.E. storms of the Pacific Ocean, on the coasts of Nicaragua 
 and Guatemala, known by the names Papagallo and Tehuantepec, are, 
 according to KEDFIELD, the fine-weather side of a whirlwind storm ; 
 while the S.W. gales in August and September, which are called 
 Tapayaguas, represent the other side. 
 
 It may not be wholly irrelevant to give, in this place, DOVE'S views 
 on the behaviour and movements of meteorological instruments in 
 connection with the " Law of Gyration :" Calculating the mean of all 
 the readings of the barometer, thermometer, and hygrometer, observed 
 during each change of the wind, taken individually, after the elimination 
 
OF THE LAW OF STORMS. 27 
 
 of the periodical changes that is, if we determine the mean distribution 
 of pressure, temperature, and moisture round the compass, or, in other 
 words, construct a barometrical) thermal, and atmospherical wind-rose 
 it will be found that this wind-rose has two poles of pressure and heat, 
 that in fact there are two points, nearly opposite to each other, at one of 
 which it is coldest, and the barometer highest ; and at the other it is 
 hottest, and the barometer lowest. From the maximum of pressure to 
 its minimum, as well as from the maximum of heat to its minimum, 
 the barometrical and thermal mean corresponding to different winds 
 decreases uninterruptedly. The former point is situated nearly at N.E., 
 and the other nearly at S.W. Proceeding from S.W., through West, 
 to N.E., the mean height of the thermometer decreases, whilst the mean 
 height of the barometer increases; but from N.E., through East to 
 S.W., the mean height of the thermometer increases, whilst the mean 
 height of the barometer decreases. What is perceptible in the mean 
 height of the thermometer and barometer due to the winds must also 
 appear in the mean movements of those instruments at the time of the 
 transition of one wind into another, and this just as well on the 
 hypothesis of a varying as on that of a uniform rotatory velocity. 
 Since, however, the elasticity of aqueous vapour, so far as regards its 
 distribution around the compass, is in close connection with the thermal 
 wind-rose, but the pressure of dry air is closely connected with the 
 barometric one, it follows that the changes in the pressure of dry air 
 and of the barometer are precisely in an inverse ratio to the changes in 
 the temperature of the air and the elasticity of the vapour suspended 
 in it. Assuming then, as a necessary consequence of these theoretical 
 considerations, that the N.W. point is the equivalent in the Southern 
 Hemisphere of the S.W. point in the Northern, and the S.E. point in 
 the former corresponds to the N.E. point in the latter, we have the 
 following 
 
 MEAN MOVEMENTS OF METEOROLOGICAL INSTRUMENTS : 
 In the Northern Hemisphere. In the Southern Hemisphere. 
 
 1. The Barometer falls with East, S.E., 1. The Barometer falls with East, N.E., 
 
 and South winds ; changes from and North winds ; changes from 
 
 falling into rising with a S.W. falling into rising with a N.W. 
 
 wind; rises with West, N.W., and wind; rises with West, S.W., and 
 
 North winds ; and changes from South winds ; and changes from 
 
 rising into falling with a N.E. wind. rising into falling with a S.E. wind. 
 
 2. The Thermometer rises with East, 2. The Thermometer rises with East, 
 
 S.E., and South winds; changes N.E., and North winds; changes 
 
 from rising into falling with a S. W. from rising into falling with a N. W. 
 
 wind; falls with West, N.W., and wind; falls with West, S.W., and 
 
 North winds ; and changes from South winds ; and changes from 
 
 falling into rising with a N.E. wind. falling into rising with a S.E. wind, 
 
28 HISTORY AND DEVELOPMENT 
 
 3. The Elasticity of Aqueous Vapour in- 3. The Elasticity of Aqueous Vapour in- 
 
 creases with East, S.E., and South creases with East, N.E., and North 
 winds ; its increase changes into winds ; its increase changes into 
 decrease with a S. W. wind ; it de- decrease with a N. W. wind ; it de- 
 creases with West, N. W., and North creases with West, S. W., and South 
 winds ; and its decrease changes winds ; and its decrease changes 
 into increase with a N. E. wind. into increase with a S. E. wind. 
 
 4. The Pressure of Dry Air decreases 4. The Pressure of Dry Air decreases 
 
 with East, S.E., and South winds ; with East, N.E., and North winds ; 
 
 its decrease changes into increase its decrease changes into increase 
 
 with a S. W. wind ; it increases with with a N. W. wind ; it increases with 
 
 West, N.W., and North winds ; West, S.W., and South winds ; and 
 
 and its increase changes into de- its increase changes into decrease 
 
 crease with a N.E. wind. with a S.E. wind. 
 
 Thus there is this in common between the two hemispheres : the 
 movements of meteorological instruments are the same with East winds 
 in the NORTHERN HEMISPHERE as they are with East winds in the 
 SOUTHERN ; and the same is the case with West winds. The difference 
 between the two hemispheres is only quantitative with N.W., N.E., 
 S.W., and S.E. winds, and, on the other hand, it is qualitative with 
 North and South winds; i.e. the mean changes of meteorological in- 
 struments are greatest in the Northern Hemisphere with N.W. and 
 S.E. winds, and least (owing to the compensation of the opposite move- 
 ments) with N.E. and S.W. winds; in the Southern Hemisphere (owing 
 to the compensation of opposite movements) they are least with N.W. 
 and S.E. winds, and, on the contrary, they are greatest with N.E. and 
 S.W. winds. The changes with North winds in the Northern Hemi- 
 sphere are, however, different from the changes that occur with North 
 winds in the Southern Hemisphere ; under like climatic conditions they 
 are, however, according to their amount, the same in both hemispheres : 
 if, for instance, there be a rise in an instrument with a North wind in 
 the Northern Hemisphere, it will sink with a North wind in the 
 Southern, and vice versa. The same occurs with South winds. 
 
 In conclusion, the character of approaching weather as indicated by 
 meteorological instruments, and the rules appertaining thereto, will be 
 most simple in the zone of the Trade- winds, because in that region the 
 rotation of the earth (as shown by the simple circumstance of a constant 
 direction in the initial and terminal points of the air current) does not 
 produce a regular shifting of the wind-vane, but its constant deflection, 
 that is, a steady wind which is the Trade ; and the storms have only 
 one form, that of the whirlwind. In the region of the monsoons they 
 will be somewhat more complicated, because there the annual alternation 
 of two air currents produces one annual shifting of the vane, being 
 periodical winds ; and the storms, the form of which is likewise whirl- 
 ing, have however a less constant direction, and in fact must deviate 
 in a certain degree from the prevalent one in order to produce a whirl- 
 
OF THE LAW OF STORMS. 29 
 
 wind. Lastly, they must be most complicated in the temperate zone, 
 where the law of rotation is most completely developed; but storms 
 occur in all forms hitherto known. The temperate zone has, however, 
 this advantage over the torrid zone, that whirlwinds produce but modified 
 effects as compared with the fearful violence of similar storms in the 
 tropics, where they are so destructive. DOVE. 
 
 It is necessary, however, to mention here that DOVE'S weather wind- 
 roses, as given on pp. 27-28, were calculated half-a-century ago on the 
 assumption that what held good for Western Europe was true for the 
 entire Northern Hemisphere, while those for the Southern Hemisphere 
 were formed on the assumption that a southerly wind, trending from the 
 pole towards the equator, corresponds to a northerly wind in northern 
 latitudes ; hence it is not surprising that later observations, carried on 
 over a more extended area, show that these wind-roses have only a partial 
 value. It is due, however, to DOVE to state that he did not suppose 
 they would be consistent for all seasons : but rather that the changes 
 would be more decided in winter than in summer. 
 
 Barometer. Of all meteorological instruments the barometer is 
 the most valuable to the seaman, even though its mean height for the 
 geographical position is unknown to him. 
 
 Beyond the tropics, and in the higher latitudes, the oscillations of the 
 barometer are incessant and irregular, ranging (rising and falling) 
 through a space of 1J to 2 inches, or even 3 inches on extraordinary 
 occasions; and thus its indications fail at times to convey the same 
 definite information that they do between the tropics, unless interpreted 
 by the aid of the thermometer and hygrometer, in conjunction with 
 various atmospheric and oceanic appearances. 
 
 Between the tropics the oscillations of the barometer are also incessant, 
 but they are regular. The periodical range due to the season does not 
 exceed a few tenths ; but the most marked of the periodical oscillations 
 is that which occurs daily, and this is so regular that Humboldt asserted 
 that the hour of the day might be known within fifteen to seventeen 
 minutes (on the average) by the height of the mercury. It shows every 
 day two maxima and two minima, as follows, according to the locality : at 
 some time between 9 and 11 a.m. the mercury begins to fall, and con- 
 tinues to do so until between 3 to 5 p.m., when it reaches the lowest 
 point of depression; it then rises until between 9 to 11 p.m., when it 
 reaches the highest point of elevation ; after that it recommences to fall, 
 falling until between 3 to 5 a.m., from which time it again rises until 
 between 9 to 11 a.m. The amount of range, again according to locality, 
 varies from -06 to '12 of an inch. A daily oscillation is not confined to 
 

 30 HISTORY AND DEVELOPMENT 
 
 the tropics ; it is equally existent in the higher latitudes, where however 
 it is much smaller than in the vicinity of the Equator, but is considerably 
 masked by the frequent fluctuations to which the atmosphere is at all 
 times subjected. The daily oscillation within the tropics is also very 
 generally masked to a ship sailing from north to south, or from south to 
 north, owing to her continual change of place, so that the barometer may 
 appear to be tolerably steady : hence in equatorial and tropical regions 
 any deviation from the usual mean such as rising when it ought to fall, 
 or falling when it ought to rise, or a change during a few hours of two 
 or three tenths of an inch is a cautionary warning that the usual and 
 regular state of the tropical atmosphere is disturbed, and that a hurricane 
 has already commenced at a distance from the ship, in which case the 
 barometer will continue to fall at a rate depending on the distance at 
 which the ship is situated from the central calm as the hurricane 
 passes in her direction. On the outer limits of the storm the barometer 
 may be at its normal height, or even a little above it, say 30 or 30-1 
 inches ; it falls more and more rapidly 
 as the centre of the storm approaches, 
 while within the calm area it has been 
 found as low as between 27 and 28 
 inches, so that the barometric curve 
 for the entire storm-field would present 
 something of the form of Fig. 3. 
 
 Prognostications of unfavourable weather, and especially of an 
 approaching hurricane, vary considerably, but between the tropics the 
 barometer is an unerring monitor \ and a rapid fall, after a few hours 
 irregular oscillation, even when atmospheric and oceanic appearances are 
 absent, is conclusive, and involves important consequences. 
 
 So far, then, the first stage in the history of the Law of Storms was 
 the general recognition that hurricanes or cyclones are progressive whirl- 
 winds, revolving around a calm centre, and that, whatever other 
 prognostications might herald their approach and mark their continuance, 
 the barometer is the most valuable monitor. 
 
 But these views did not pass wholly unchallenged. REDFIELD, when 
 discussing the facts and observations he had accumulated, must neces- 
 sarily theorise on the cause of hurricanes ; and in this he was unfortunate. 
 He had two opponents in America, who, though they did not entertain 
 precisely similar opinions on the subject, were yet, in the main, advocates 
 of an "in-blowing theory" as opposed to the "circular theory" of 
 storms maintaining at the same time the progressive motion. These 
 opponents were Dr. ROBERT HARE and JAMES ESPY : the former had 
 written on gales and storms as early as 1822, and continued to oppose 
 
pSE LIBft^, 
 
 * or rf 'r ~\ 
 
 EBS1TY jj 
 
 OF THE LAW OF STO^Itet Cd/ lerrtDMtA. 31 
 
 REDFIELD'S views, and make occasional strictures on those of ESPY, until 
 1854; ESPY, as a writer on storms, dates from 1836 to 1857, and his 
 chief works are, the Philosophy of Storms, and Reports on Meteorology 
 to the Burgeon-General of the United States Army. 
 
 It cannot be denied that ESPY'S views on, and theory of, storms, 
 hurricanes, and tornadoes were favourably received by a large body of 
 supporters. He had noticed the extraordinary fall of the barometer 
 which accompanies these phenomena ; then, collecting facts and obser- 
 vations, as well as following the traces impressed on the ground in the 
 track of hurricanes, he concluded that in those storms the air rushes 
 directly from the exterior towards a central space ; so that, if the wind 
 from one side blows east, it blows from the west on the other side, while 
 the centre shifts its place. He recognised, as a cause of this " in- 
 blowing," an ascending column of air, the temperature of which does 
 not vary, because the precipitation of its moisture restores to it the 
 temperature it loses by expansion : this ascending column gives rise to 
 a wind on the surface of the earth ; and above, to the formation of a 
 cumulus, which is immediately resolved into hail or rain. ESPY'S theory 
 explained in full is as follows : 
 
 " When the air in any locality acquires a higher temperature or a 
 higher dew-point than that of the surrounding regions, it is specifically 
 lighter, and will ascend. In ascending it comes under less pressure, and 
 expands ; in expanding from diminished pressure, it grows colder about 
 1J Fahr. for every hundred yards of ascent. In cooling as low as the 
 dew-point (which it will do when it rises as many hundred yards as the 
 dew-point at the time is below the temperature of the air in degrees of 
 Fahrenheit) it will begin to condense its vapour into water or cloud. In 
 condensing its vapour, it will evolve latent caloric; this evolution of 
 latent caloric will prevent the air from cooling so fast in its further 
 ascent as it did in ascending below the base of the cloud now forming ; 
 the current of air, however, will continue to ascend and grow colder 
 about half as much as it would do if it had no vapour in it to condense ; 
 and when it has risen high enough to have condensed by the cold of 
 expansion from diminished pressure one-hundredth of its weight of 
 vapour, it will be about 48 less cold than it would have been if it had 
 had no vapour to condense nor latent caloric to give out that is, it will 
 be about 48 warmer than the surrounding air at the same height ; it 
 will therefore (without making any allowance for the higher dew-point 
 of the ascending current) be about one-tenth lighter than the surrounding 
 colder air, and of course it will continue to ascend to the top of the 
 atmosphere, spreading out in all directions above as it ascends, over- 
 lapping the air in all the surrounding regions in the vicinity of the storm, 
 and fall still more under the storm-cloud by the outspreading of air 
 
32 HISTORY AND DEVELOPMENT 
 
 above thus leaving less ponderable matter near the centre of the up- 
 moving column to press on the barometer below. 
 
 " The barometer thus standing below the mean under the cloud in 
 the central region, and above the mean on the outside of the cloud, the 
 air will blow on all sides from without inward under the cloud ; the 
 air, on coming under the cloud, being subjected to less pressure, will 
 ascend and carry up the vapour it contains with it, and as it ascends 
 will become colder by expansion from constantly diminished pressure, 
 and will begin to condense its vapour into the cloud at the height 
 indicated before, and thus the process of cloud-forming will go on. 
 
 " Now it is known that the upper current of air in the United States 
 moves constantly, from a known cause, towards the eastward probably 
 a little to the south of east ; and as the upward-moving column con- 
 taining the cloud is chiefly in this upper current of air, it follows 
 that the storm cloud must move in the same direction; and over 
 whatever region the storm cloud appears, to that region will the wind 
 blow below : thus the wind must set in with a storm from some eastern 
 direction, and, as the storm cloud passes, the wind must change to some 
 western direction, and blow from that quarter till the end of the storm." 
 
 Thus ESPY'S theory is not wholly dissimilar from that of BRANDE, 
 who wrote on the great European storm of 1782. It may be mentioned 
 that Prof. BACHE, the late Superintendent of the United States Coast 
 Survey, was also a supporter of ESPY'S " in-blowing " theory. 
 
 Now KEDFIELD had been unfortunate in his first supposition that " the 
 leading storms of the Northern and Western Atlantic and of the American 
 coast originate in detached and gyratory portions of the northern 
 margin of the Trade winds, occasioned by the oblique obstruction which 
 is opposed by the islands to the direct progress of this part of the Trade, 
 or to these causes combined ; " and when his " circular " theory was 
 threatened by an opponent's " in-blowing " one, he made more mistakes 
 and many gratuitous assertions ; among the most remarkable may be 
 mentioned that by which he explains the rotation of storms in a 
 particular direction, on the assumption that the winds blow in horizontal 
 circuits round the great oceanic basins, and his repudiation of the 
 generally acknowledged theory of the Trade winds, mainly, as it would 
 appear, because that theory was " founded on the alleged effects of 
 calorific rarefaction in the equatorial region." From his stand-point 
 there was no ascending current of air at the equatorial regions overflowing 
 and passing off towards the temperate and polar regions, for rarefied air 
 could not ascend except by means of " vorticular action;" and this 
 seemed his best argument against the " in-blowing or centripetal theory " 
 of ESPY and HARE, 
 
OF THE LAW OF STORMS. 33 
 
 It may here be observed that HARE, HEID and PIDDINGTON but 
 especially HARE and PIDDINGTON ascribed the origin of hurricanes to 
 an electrical or electro-magnetic disturbance of the atmosphere ; this is 
 not singular, when we remember that not many years back every obscure 
 natural phenomenon, especially in meteorology, was forthwith solved by 
 referring it to an electrical cause : it is scarcely necessary to say that 
 such origin for those storms is now regarded as untenable ; but there is 
 no doubt that in most hurricanes electricity is developed to an enormous 
 extent. PIDDINGTON also suggested that cyclones were originally formed 
 above the earth, and that they descended, now horizontally (as horizontal 
 discs), and now inclined at an angle, and that they both dilated and 
 contracted in their progress. 
 
 DOVE'S views have already been given in full (p. 18-26); he arrived 
 at conclusions widely differing from those of ESPY, and strongly 
 supported the " circular or rotary theory " as enunciated by REDFIELD, 
 and developed by REID, PIDDINGTON and THOM. He admits of no 
 centripetal theory, no incurving of the winds, but, on the contrary, 
 maintains that they blow at the circumference in a direction perpen- 
 dicular to the radii of the circle, while their progressive motion may be 
 curvilinear, as in the West Indian and Mauritius hurricanes, or recti- 
 linear, as in some of the Indian and China cyclones. In explaining the 
 cause of their combined rotatory and progressive motion, he takes the 
 West Indian hurricane as typical of all others, and proceeds to show 
 how it originates at the inner boundary of the N.E. trade-wind owing 
 to the intrusion of the upper counter-current of the Trade into the lower 
 current, or even through a part of the S.E. trade traversing the equator, 
 entering the northern hemisphere, and producing similar results ; as for 
 example, 
 
 Take the line a b (Fig. 4) to represent, for the Northern hemisphere, 
 a series of molecules of air parallel with the equator, and which are set 
 in motion towards the North, in the direction a c ; if the space d b h 
 were empty or void, these molecules would tend towards g k, because 
 their progression takes place from larger parallel circles towards those 
 which are smaller. But if the space d b h be filled with air at rest, the 
 
 molecules at b must encounter, 
 on their way towards d, others 
 having less velocity of rotation, 
 the effect of which will be to 
 diminish their velocity towards 
 the East, and consequently the 
 air at b, instead of moving to- 
 wards h, would tend towards./! 
 The molecules between a and b 
 
34 HISTORY AND DEVELOPMENT 
 
 are also affected but differently and to a smaller extent moving more 
 towards g. It follows that if a b be a mass of air pressing forward 
 from south to north, the direction of the storm- wind will be much more 
 from south on its east than on its west side, where it will be more 
 westerly, and thus a tendency to whirl (or rotate) originates in the 
 direction S., E., N., "W. This tendency to whirl would not exist if there 
 were no opposing mass in the space dbh; but, on the other hand, it will 
 augment in proportion to the opposition of the two forces, and the 
 westerly deflection of the storm is checked ; and the storm itself will 
 rotate the more violently, the more it retains the original direction of 
 its course. 
 
 In the zone of the Trade-winds the space d b h is filled with air 
 moving from N.E. towards S.W. : and it is there that the opposition 
 will be greatest. The air at b may be checked to such an extent that 
 it will preserve, unaltered, its direction towards d, while that at a has a 
 tendency in the direction of g. The storm will therefore rotate with 
 greatest force in the torrid zone, but at the same time preserve its 
 dimensions unchanged. As soon, however, as it enters the temperate 
 zone, it comes to a region where the space d b his already full of air 
 moving from S.W. to N.E. The resistance which the air at b had 
 previously experienced is now suddenly diminished, or wholly ceases, 
 and instead of taking the direction b d it turns sharply in that of b h. 
 The storm thus bending round suddenly at right angles to its previous 
 
 direction, rapidly expands its 
 dimension, since the differences 
 in motion between the points 
 a and b no longer exist. 
 
 The hurricanes of the 
 Southern hemisphere originate 
 from the same cause, but the 
 rotation takes place in the 
 opposite way (Fig. 5), while 
 the change of direction on the borders of the tropic is analogous to 
 that which takes place in the Northern hemisphere. 
 
 It may be remarked, in passing, that DOVE is a strong supporter of 
 the ascending current in the region of the equatorial calms. 
 
 Before proceeding further it may be as well to compare the three 
 Systems the centripetal or "in-blowing," the centrifugal or "out- 
 blowing," and the " circular or rotatory," to note where there is simi- 
 larity, and where dissimilarity, whether they are stationary or have 
 progression in any particular direction. 
 
 The perfectly centripetal or " in-blowing " system is seen in Fig. 6, 
 where the winds blow in all directions towards the centre, in such manner 
 
OF THE LAW OF STORMS. 35 
 
 that they are northerly on the north side, easterly on the east side, 
 southerly on the south side, and westerly on the west side. This would 
 
 be the same for both hemispheres. At the centre there is a calm area. 
 There is no rotatory motion of the winds, though there may be pro- 
 gression in a definite direction, which would give an apparent veering 
 of the wind either with or against the sun's motion. 
 
 The perfectly centrifugal or " out-blowing " system is seen in Fig 7 : 
 in it the winds blow in all directions from the centre, and consequently 
 directly opposite to those of the " in-blowing " system : thus, they are 
 southerly on the north side, westerly on the east side, northerly on the 
 south side, and easterly on the west side, for both hemispheres with a 
 central calm area. In this case, also, there is no rotatory motion, but 
 should there be progression in a definite direction, there must also be 
 an apparent veering of the wind with or against the sun's motion. 
 
 In the circular and rotatory system the wind, according to the popular 
 idea, rotates around a central calm area, in a direction from right to left 
 in the northern hemisphere, but from left to right in the southern hemi- 
 
 sphere or, to use a familiar illustration, con- 
 trary to the motion of the hands of a watch in 
 the northern, but coincident therewith in the 
 southern, hemisphere, and which is conse- 
 quently contrary to the apparent course of the 
 sun in each hemisphere (see Fig. 8). Hence it 
 follows that, in the northern hemisphere the 
 winds must be easterly in the northern quad- 
 rant, southerly in the eastern quadrant, 
 
 in the western quadrant ; while, for similar 
 quadrants in the southern hemisphere, the 
 direction of the winds must (since the rotation is reversed) be the 
 reverse of those just given i.e., westerly in the northern quadrant, 
 northerly in the eastern quadrant, easterly in the southern quadrant, and 
 southerly in the western quadrant. And further, in each hemisphere, 
 the winds must be westerly on the equatorial side of the storm. 
 
36 
 
 HISTORY AND DEVELOPMENT 
 
 One of the earliest axioms supposed to be established in connection 
 with the rotatory theory of storms was, that, in each hemisphere, the 
 centre of the storm always bore eight points from the wind, reckoned with 
 the sun; such that in the N. hemisphere, when facing the wind, the 
 centre would be square to the right ; in the S. hemisphere, square to 
 the left. 
 
 So also, as a consequence of the circular motion of the wind round 
 a centre, combined with a bodily progressive movement, other two 
 important axioms were thought to be established, as if appertaining 
 exclusively to rotatory storms viz., that on the passage of such a storm 
 over a place or ship 
 
 . In the N. hemisphere the veering of the wind in the right hand 
 semicircle must coincide with the apparent course of the sun, while in 
 the left, hand semicircle it would be against the apparent course of the 
 sun.* 
 
 In the S. hemisphere the veering of the wind in the right hand 
 semicircle must be contrary to (against) the apparent course of the sun, 
 while in the left hand semicircle it would coincide with the apparent 
 course of the sun.* 
 
 Such, in brief, is the " Law of Storms " on the rotatory theory. 
 
 It has already been shown (p. 20) that, for the N. hemisphere, 
 veering in the direction E., S., W., N., E. is with the sun; and W., S., 
 E., N., W. is against the sun; and, 
 
 For the S. hemisphere, veering through E., N"., W., S., E. is with 
 the sun ; and W., N., E., S., W. is against the sun. 
 
 A glance at Figs. 6, 7, and 8, in conjunction with the more explicit 
 statements in the following table, will show that, for the three systems 
 of winds the centripetal, centrifugal, and circular though, for any 
 given side of the area within which the wind is blowing, the direction 
 of the wind is totally different, the veering, as with or against the sun, 
 is precisely similar in all, and not confined to the circular or rotatory 
 storm. 
 
 N. HEMISPHERE. 
 
 Centripetal. 
 
 Centrifugal. 
 
 Circular. 
 
 
 Progres. motion to N". W. 
 On right hand* 
 On left hand 
 Change of wind at centre 
 
 Wind veers 
 
 N.-N.E.-E. 
 
 w.-s.w.-s. 
 
 N.W.toS.E. 
 
 Wind veers 
 S.-S.W.-W. 
 
 E.-KE.-K 
 S.E. toN.W. 
 
 Wind veers 
 
 E.-S.E.-S. 
 K-N.W.-W. 
 N.E. toS.W. 
 
 with sun 
 against sun. 
 
 Progres. motion to N.E. 
 On right hand 
 On left hand 
 Change of wind at centre 
 
 Wind veers i Wind veers 
 E.-S.E.-S. W.-KW.-K 
 N.-KW.-W. S.-S.E.-E. 
 N.E. toS.W.S.W. toKE. 
 
 Wind veers 
 S.-S.W.-W. 
 
 E.-N.E.-N. 
 S.E. toN.W. 
 
 with sun 
 against sun. 
 
 * Right and left in this case have reference to the track, as if looking towards the direction in 
 which the storm is bodily progressing. 
 
OF THE LAW OP STORMS. 
 
 37 
 
 S. HEMISPHERE. 
 
 Centripetal. 
 
 Centrifugal. 
 
 Circular. 
 
 Progres. motion to S. W. 
 On right hand 
 On left hand 
 Change of wind at centre 
 
 Wind veers 
 W.-N.W.-N. 
 S.-S.E.-E. 
 S.W. toN.E. 
 
 Wind veers 
 E.-S.E.-S. 
 N.-N.W.-W. 
 N.E. toS.W. 
 
 Wind veers 
 S.-S.W.-W. 
 
 E.-N.E.-N. 
 S.E. to N.W. 
 
 against sun 
 with sun. 
 
 Progres. motion to S. E. 
 On right hand 
 On left hand 
 Change of wind at centre 
 
 Wind veers 
 S.-S.W.-W. 
 
 E.-N.E.-N. 
 S.E. toN.W. 
 
 Wind veers 
 
 N.-N.E.-E. 
 
 w.-s.w.-s. 
 
 N.W. to S.E. 
 
 Wind veers 
 E.-S.E.-S. 
 
 N.-N.W.-W. 
 N.E. toS.W. 
 
 against sun 
 with sun. 
 
 The most marked characteristic in the three systems is the change 
 in the direction of the wind as the centre passes over a place ; in each 
 it is from a given point to another directly opposite; but on the 
 circular theory the change must always occur through two opposite 
 directions, each of which is perpendicular to the movement of transla- 
 tion, for example, in the N. hemisphere, with the storm moving to 
 N.W.-ward, the change at the centre will be from N.E. to S.W. 
 
 All these points of similarity and difference must be borne in mind 
 in connection with the new views that are held by eminent meteorolo- 
 gists in respect to hurricanes. 
 
 The laws and rules briefly indicated on pp. 35-36, and respecting 
 which it has been said that "nothing can be more definite and 
 unambiguous," have also been very generally accepted in their hard 
 and fast condition by the majority of French authors ; and in a work 
 recently published the following illustration of the effect of the com- 
 bined motion of translation and rotation is to be found : 
 
 Take the case of a storm, in the northern hemisphere, moving 
 towards the N.W. at the rate of 8 miles an hour, and in which the 
 rotatory velocity of the winds is 80 miles : if tangents be drawn to the 
 circle such that they represent the velocity of rotation (80 miles) as one 
 component, and straight lines of one-tenth that length (8 miles) be 
 drawn parallel with the track of the storm, as the other component ; on 
 completing the parallelograms of which these are the sides, the diagonals 
 of the parallelograms will give the resultant winds both as regards 
 direction and velocity. On such a projection the winds at the different 
 parts of the storm will be, in respect to direction, 
 
 At S.W. part, wind N.W. 
 
 S. 
 S.E. 
 
 E. 
 N.E. 
 
 N. 
 N.W. 
 
 W. 
 
 W.^S. instead of West. 
 
 S.W.4S. instead of S.W. 
 
 S. J E. instead of South. 
 
 S.E. 
 
 E.S. instead of East. 
 
 .E.E. instead of N.E. 
 
 N.E. instead of North. 
 
 Now, it is not a little curious that, on the supposition of the winds 
 having the direction just given, they would be slightly centrifugal on 
 
38 HISTORY AND DEVELOPMENT 
 
 the advancing semicircle of the storm, but slightly centripetal on the 
 receding half. 
 
 Again, following out the author's illustration, the sum of the two 
 components (80 + 8), viz., 88 miles, would be the velocity of the wind 
 at the N.E. part of the storm (c in 
 fig. 9) ; and the difference of the com- 
 ponents (80 - 8), i.e. 72 miles, would 
 be the velocity of the wind at its S.W. 
 part (a in fig. 9) : and thus, from the 
 minimum at a the velocity would 
 increase alike on the advancing and 
 receding semicircles, i. e. in the directions 
 a d c and a b c, until it attained its 
 maximum at c. This is not incon- 
 sistent with what must naturally occur in all storm-fields of wide area, 
 when the wind is circling around, with a greater or less convergence 
 towards, a central area of low pressure. 
 
 It may be well, at this point, to inquire whether the founders of the 
 rotatory theory of storms laid down hard and fast laws and rules for the 
 guidance of mariners, such as their followers have done *? There is no 
 doubt that they did, but almost invariably with many subsequent quali- 
 fications, and much explanation as to apparent anomalies ; so that, far 
 from everything in connection with the theory and practice being 
 " definite and unambiguous," the reverse is the case. It may indeed be 
 said that they gave but a very qualified adhesion to much that has been 
 enunciated for the last 30 or 40 years as the true principle of the law 
 of storms. 
 
 REDFIELD, writing in the American Journal of Science, in 1846, 
 says : " When, in 1830, I first attempted to establish by direct evidence 
 the rotative character of gales or tempests, I had only to encounter the 
 then prevailing idea of a general rectilinear movement in these winds. 
 Hence I have deemed it sufficient to describe the rotation in general 
 terms, not doubting that on different sides of a rotatory storm, as in 
 common rains or sluggish storms, might be found any course of wind, 
 from the rotative to the rectilinear, together with varying conditions as 
 regards clouds and rain. 
 
 " But I have never been able to conceive that the wind in violent 
 storms moves only in circles. On the contrary, a vortical movement, 
 approaching to that which may be seen in all lesser vortices, aerial or 
 aqueous, appears to be an essential element of their violent and long 
 continued action, of their increased energy towards the centre or axis, 
 and of the accompanying rain. In conformity with this view the storm 
 
OF THE LAW OF STORMS 39 
 
 figure on my chart of the storm of 1830 was directed to be engraved in 
 spiral or involute lines, but this point was yielded for the convenience 
 of the engraver. * * * * 
 
 "The common idea of rotation in circles, however, is sufficiently 
 correct for practical purposes, and for the construction of diagrams, 
 whether for the use of mariners, or for determining between a rotative 
 and general rectilinear wind, on one hand, or the lately alleged centri- 
 petal winds on the other. The degree of vorticular inclination in violent 
 storms must be subject, locally, to great variations; but it is not 
 probable that on an average of the different sides, it ever comes near to 
 forty-five degrees from the tangent of a circle, and that such average 
 inclination ever exceeds two points of the compass, may well be doubted." 
 
 REID, equally with REDFIELD, saw the weak points of a purely 
 circular theory, and hence, in his Progress of the Development of the Law 
 of Storms, he writes : 
 
 " On the approach of a revolving storm, the weather is sometimes 
 calm, or the wind is unsteady, or it may draw towards the storm ; and 
 when this occurs it may be difficult to determine the exact point of the 
 compass from whence the true whirlwind sets in. In like manner, it 
 may not always be easy to determine the exact point at which a gale 
 ends." 
 
 And reverting to the subject, he says, 
 
 "But a circle will only correctly represent a whirlwind when 
 stationary. In the progressive whirlwind the figure would become 
 cycloidal, and the degree of curvature would depend upon the rate of 
 progress of the storm. While attempting to lay down a storm's track a 
 wrong judgment may be formed, if data be rejected on account of its not 
 conforming to a figure exactly circular." Then, using the track of the 
 Charles Heddle reversed, as for the northern hemisphere, he says " at the 
 point a the wind would be north j at the point b (due east of the point a) 
 it would be S.W. : circles, nevertheless, have been found the most con- 
 venient figures in endeavouring to explain the nature of storms." 
 
 PIDDINGTON devotes eight pages of his Horn-Book to discussing the 
 " probability of the incurving of the winds in a cyclone," and the " flat- 
 tening in of the circle of a cyclone " quoting at the same time, with 
 approbation, the extract from REDFIELD, given above. It appears to 
 him " not improbable that, while at some point in the whole zone of a 
 cyclone the wind may be blowing in a true circle, it may on the inner or 
 central side of that part be a series of mcurving spirals, and that on the 
 outer part there may be a centrifugal (i.e., a curving off from the centre) 
 tendency." He also expresses the view, as a probability or possibility, 
 
40 
 
 HISTORY AND DEVELOPMENT 
 
 " that when cyclones are increasing in diameter or dilating, the spiral is 
 a diverging one, or that the arrows (of the wind) curve outwardly, and 
 when it contracts the spiral is a converging or incurving one : " and in 
 another place he says, "it would seem also that this incurving takes 
 place sometimes at a distance from the centre." PIDDINGTON is generally 
 full of instances, and, according to his usual practice, adduces examples 
 illustrating the remarks quoted above; he particularises ships' logs 
 which indicated an alternation or shifting of the wind through four, 
 five and more points, and infers that one vessel " by the alternating of 
 the winds must have drifted over a succession of curves." It is also by 
 " an excessive incurving of the winds towards the centre " that he finds 
 " an explanation of how some ships (when disabled) 
 may be blown and drifted round and round, with- 
 out drifting out of the fatal centre : " and he con- 
 cludes his observations as follows, "perhaps a 
 further proof, or rather an indication of this in- 
 curving may be found in the remarkable instances 
 of ships being surrounded or having their decks 
 covered during the passage of the calm centre of 
 cyclones, in the neighbourhood of the land, with 
 land and aquatic birds, butterflies, horse-flies, &c. ; 
 now within a cyclone these animals must be 
 incapable of doing more than keeping themselves 
 
 in the air ; they must therefore, we may 
 
 suppose, be gradually carried inwards by the in- 
 curving tendency of the winds, and at the centre 
 are kept there because they cannot fly out of it ; 
 
 if the wind blew in true circles they would be scattered all 
 
 over the body of the cyclone." 
 
 It has been seen that BEDFIELD and PIDDINGTON agree in this that 
 the incurving of the winds in a hurricane 
 may amount to fully two points, if not 
 more. No diagrams have been given to 
 show the character of this incurving, as 
 they understood it so that some have 
 taken it to be of a spiral kind, as repre- 
 sented in Fig. 10 ; while others have 
 supposed it to approximate to the form 
 of Fig. 11 which is really an incurved 
 spiral (see dotted lines) but circular as 
 the centre is approached. There is no doubt that REID, equally with 
 REDFIELD and PIDDINGTON, wavered in opinion as to the precise form of 
 
OP THE LAW OP STORMS. 
 
 a cyclone ; but nevertheless all three gave very positive rules, especially 
 as to the direction of the winds on the different sides of a hurricane, 
 and as to the bearing of the centre, taking the wind as a guide : it is 
 true that cautionary paragraphs constantly occur in their writings, such 
 as very considerably modify the positive character of their previous rules ; 
 but subsequent authors, and the devotees of the circular theory, have 
 persistently lost sight of the cautions, and only given, in a very dogmatic 
 manner, the " definite and unambiguous " rules, forgetful that the 
 backing and filling is of their very essence: for as a shipmaster has 
 written " if the centre always bore eight points from the direction of the 
 wind ; if the wind gradually increased in force as we near the centre ; 
 if the wind veered gradually in all parts of the storm ; and if the centre 
 were the only dangerous part of it, then the avoiding of a hurricane 
 would be very simple." 
 
 To sum up, though REDFIELD, REID and PIDDINQTON at first con- 
 jectured a purely circular and rotatory action for the winds within the 
 area of the storm-field, they came at last to modify this view, as a 
 deduction from the observations submitted to them ; and they conceded 
 this much that at least in some hurricanes, if not in all, there must be 
 a slight incurvature of the winds ; but they had no distinct perception to 
 what extent this deviation existed, and, consequently, as the practical 
 rules deduced from the law of storms were already in the hands of those 
 for whose benefit they were written, they were careful to note the 
 anomalies while leaving the rules unmodified. Their followers elaborated 
 the one and neglected the other. 
 
 THOM, as before observed, was also a supporter of the rotatory theory 
 of hurricanes, and he has much to say on the varied aspects of these 
 storms as they occur in the Southern Indian Ocean : but in respect of the 
 manoeuvring of ships when endeavouring to steer clear of, or when con- 
 tending with, their fury, he merely suggests the course to be pursued, and 
 his remarks are not to be taken as arbitrary rules, " for it is obvious that 
 general rules cannot be made applicable to every case." 
 
 It is, however, worthy of note that THOM was one of the first to indi- 
 cate that the hurricanes of the Southern Indian Ocean had their origin 
 in the region of calms or " variables " between the N. W. monsoon and 
 the S.E. trade. His observations are as follow : 
 
 We have two great winds blowing, in contrary directions, from one 
 side to the other of the Indian Ocean, between the equator and the 
 tropic, the one depending on fixed and permanent laws, and the other, 
 periodically called into existence by agencies of a temporary duration, 
 exerting an almost antagonist force to the former. As the solar action 
 takes place in a gradual manner, the recession of the trade-wind, and 
 
42 HISTORY AND DEVELOPMENT 
 
 establishment of the monsoon, follow in a corresponding degree : the sky 
 is serene and clear, indicating an increasing temperature in the upper 
 regions of the atmosphere, and a rising dew point. But, as soon as the 
 sun begins to leave the tropic, and the heat to decline, the changes con- 
 tingent on the subsidence of the monsoon take place in a sudden and 
 unequal manner, and it would seem often give rise to those fearful conflicts 
 of elementary forces which form the subject of our present inquiry. 
 
 "Between the two winds in question, about lat. 10 to 12 S., there 
 is an intervening space of several degrees of variable and light winds, or 
 calms, distinguished by an unsettled state of the atmosphere, exactly 
 similar to that which occurs between the S.E. and N.E. trade- winds in 
 the Atlantic. In approaching this tract, the S.E. trade deflects to South, 
 and the westerly monsoon to N.N.W. or North. At the meeting of 
 these currents of air, or rather in the neutral interval, sudden squalls, 
 lulls, thunder-storms, and heavy rains, are usually experienced. This is 
 to be expected from the very different conditions of the two winds, for 
 they are not only diametrically opposite in their direction, but as 
 dissimilar in reference to their respective temperature, humidity, and 
 specific gravity, as it is well possible for two neighbouring masses of air 
 to be. * * * * 
 
 "During the presence of the sun in the southern hemisphere, a 
 constant struggle between the two opposite currents of atmosphere may 
 naturally be expected to take place along the interval which separates 
 them. Not only theory, but observation establishes this fact. On the 
 borders of the two winds, ships meet unsettled weather, and variable 
 winds, at one moment are in a calm, and the next, perhaps, laid on 
 their beam-ends by a squall. It is in this situation that thunder-storms 
 and whirlwinds abound. To this tract we wish to direct attention, as 
 being not only the locality where rotatory storms are first formed, but 
 along which they move during the greater part of their course, till they 
 acquire an intestine and specific action, involving the neighbouring 
 currents of the atmosphere, which enable them to advance through the 
 trade-wind to its opposite limits. 
 
 " Two currents of atmosphere in motion from opposite points, on ap- 
 proaching the neutral columns of air which intervene between them, 
 cannot be supposed to lose their momentum, or to have their velocity 
 suddenly arrested, or even materially diminished ; on the contrary, we 
 must suppose that their motion will be diverted into a new channel. 
 From a knowledge of the forces which propel the currents towards one 
 another, we are prepared to expect that the S.E. trade will have a 
 constant tendency, in a greater or less degree, according to accidental 
 circumstances, to deflect towards the equator, and blow across the line of 
 
OP THE LAW OP STORMS. 43 
 
 " variables " into the region of the monsoon. This deviation will be 
 favoured by any temporary diminution of the antagonist force of the 
 opposite current. Suppose, for instance, a sudden elevation of tempera- 
 ture, and consequent vertical current at some part of the monsoon, a 
 considerable barometrical depression must follow, and a rush of the 
 surrounding atmosphere would take place towards the point of diminished 
 pressure, but with greatest force from that side where the heavier and 
 cool S.E. trade-wind was pressing onwards at the earth's surface. This 
 last current, on entering the 
 hotter locality of the monsoon, 
 would quickly acquire its 
 temperature and relative de- 
 gree of moisture, and con- 
 sequently its direction. A 
 portion of the trade-wind, in 
 undergoing this change, would 
 curve from its natural course 
 S.E. to South, till it mixed 
 with the monsoon, and then 
 become a S.W., West, and 
 N.W. wind (see Fig. 12). 
 
 " Now, the new and periodical agencies which communicate a N. W. 
 direction to the monsoon, evidently cause it to blow from the equator, 
 so that at the point where it meets the trade-wind its course would be 
 North, and from its greater degree of rarefaction would ascend and pass 
 over it. But, in doing so, a portion of it must gradually mix with the 
 colder S.E. wind, lose some of its heat, and partake of the character and 
 assume the direction of the latter, in the course of which it would be 
 found veering from N.W. to North and N.E. 
 
 "Thus, by the natural deflection of the two winds in opposite 
 directions, a constant tendency would exist in the atmosphere interposing 
 between them to assume a circular motion. The curve of the trade-wind 
 from East by S.E., to S. and S.W., to unite with the monsoon, amounts 
 at least to half a circle, while the inclination of the latter from the 
 equator, or in a reverse direction to the former, and to an equal extent, 
 would form the opposite half of a completely connected circle. * * * 
 
 " In every case where a circular storm is clearly determined, a branch 
 of the S.E. trade is found to cross the line of "variables" and enter the 
 monsoon, while the latter recrosses into the former, each gradually de- 
 flecting till it appears to be part of the other, and both revolve around a 
 common centre (see Fig. 1 2). It has been shown that the disc of atmosphere 
 in a state of rotation is progressive, and, at places over which the centre 
 
44 HISTORY AND DEVELOPMENT 
 
 passes, its first approach is detected by the increasing violence of the S.E. 
 trade, and its departure by a N. W. wind ; so that throughout the whole 
 progress of the storm the relative positions of the two winds are preserved. 
 Even the peculiar characteristics of each, as already described, are easily 
 recognised at that part of the circle to which they belong. There is 
 another circumstance worthy of notice in reference to the mutual action 
 of the two currents when involved in a circular motion, which is, that 
 during the early stages the N.W. predominates over the S.E., but, as 
 the hurricane moves towards the tropic, it is exactly the reverse. In 
 lat. 20 and 22 S., the S.E. trade often prevails with tempestuous 
 violence for several days before the vortex arrives; but when the N.W. 
 or west part of the storm sets in, although it blows with equal severity 
 for ten or twelve hours, it seldom continues much longer. * * * 
 
 "Whether the opinions which have been advanced regarding the 
 formation of rotatory gales approach the truth, must in some measure 
 depend on future observation ; one point, however, seems to be satis- 
 factorily ascertained, that they are generally developed in the interval 
 between, what we may be allowed to call, their parent winds, and in 
 intimate relation with both, in such a manner that each would convey 
 additional impulse to the revolving mass at the opposite sides of the 
 circle. In this manner a storm may be supposed to receive accessions of 
 impetus at the exterior of the circle, by the conversion of the rectilinear 
 force into a circular motion, and by enlarging the sphere of its action 
 it would impinge more extensively on neighbouring winds, and receive 
 a continuous support. Even if the contiguity of two winds is not 
 required to form a rotatory motion in the atmosphere, it appears to be 
 absolutely necessary for the continued existence of such an action. 
 There seems to be no other cause adequate to explain the duration of 
 hurricanes for fifteen or twenty days, while travelling along a distance 
 of 3000 miles : and finding that they arise between regular but opposite 
 currents, continue their progression in the same relative position, and 
 only terminate or lose their specific action at that point where the two 
 winds cease to be regular, or near the extra-tropical " variables," it 
 is consonant to reason to trace a connection of cause and effect between 
 them." 
 
 From THOM'S stand-point there is no hurricane until the winds shall 
 have assumed a rotatory, or approximately rotatory, character ; he, how- 
 ever, admits that even within the area of the storm-field the winds are 
 often fitful and irregular, both as regards direction and force, and 
 speaks of " alternate lulls and fierce spiral gusts in the vicinity of the 
 vortex;" but his happy exposition of the origin of such storms "from 
 a constant struggle between two opposite currents " of air clearly in- 
 
OF THE LAW OF STORMS. 45 
 
 dicates that on the borders of the hurricane there must be a considerable 
 space where the " parent winds," that " convey additional impulse to 
 the revolving mass," are stormy and tempestuous to such a degree as 
 to render it " difficult," as REID says, " to determine the exact point 
 from whence the true whirlwind sets in;" and THOM'S diagram, a 
 portion of which is given on page 43, is very suggestive on this point. 
 
 THOM'S explanation of the origin of the cyclones of the Southern 
 Indian Ocean was written in 1845, but at that time it did not re- 
 ceive much attention. Some years afterwards, when the Mauritius 
 Meteorological Society was in full activity, and the storms of that 
 region had become a constant subject of discussion among its members, 
 Mr. MELDRUM, the able secretary of the Society, was enabled by fuller 
 investigation to corroborate and develop the views and opinions enun- 
 ciated by THOM in respect to the locality where cyclones are generated ; 
 and in 1861 he wrote, 
 
 "The more the Hurricanes of the Southern Indian Ocean are 
 studied, the more evident does it become, that they owe their origin 
 and continued existence, for several days, to the vibrating and conflict- 
 ing action of the equatorial Westerly monsoon and the S.E. trade-wind, 
 when the sun has southern declination. Between these two systems of 
 winds there is a belt of maximum heat, and consequently of atmospheric 
 rarefaction, with light and variable winds ; and towards this space of 
 diminished pressure the monsoon and trade-wind flow from opposite 
 directions, bringing with them the vapours which have been generated 
 on either side. This belt of variables is inclined obliquely across the 
 ocean, extending further south on its western side, towards the heated 
 land ; and it also has, within certain limits, a vibratory motion from 
 north to south, and vice versd, along its whole extent, the trade and 
 monsoon reigning alternately in the same localities. The monsoon, 
 however, is the aggressor, being backed and pressed upon by a much 
 greater extent of cold and heavy air than its opponent ; and it is during 
 its irruptions into the trade-wind region that hurricanes take place. 
 
 " It might be expected that as the line of demarcation between the 
 two winds has a great extent in longitude, so several hurricanes (or 
 cyclones) might be formed between them at the same time, but at con- 
 siderable distances apart ; and observation shows that this is the case. 
 Several instances occurred in 1860, particularly in February. During 
 the last ten days of that month, the equatorial borders of the trade, 
 along its whole extent, were in a state of stormy gyration, and several 
 distinct cyclones raged at the same time. The same thing has again 
 occurred this year (1861)." 
 
46 HISTORY AND DEVELOPMENT 
 
 The foregoing remarks on the history of the investigation which 
 resulted in the establishment of the circular theory of storms are suffi- 
 ciently explicit. The outcome of the investigation was the development 
 of that theory on the basis of the wind being more or less tangential to 
 the circle, with the centre of the storm eight points (90 degrees) to the 
 right of the wind's direction in the Northern Hemisphere, but to the left 
 in the Southern Hemisphere. 
 
 This part of the subject may be aptly closed by introducing the cele- 
 brated case of the brig Charles Heddle, which gave such impetus to the 
 purely circular theory of hurricanes; after which, and before proceeding to 
 discuss the more recent views of meteorologists in respect to these storms, 
 it will be most appropri- 
 ate to collect into a few 
 pages what have hitherto 
 been considered the prac- 
 tical rules deduced from 
 the "law of storms." 
 
 The Charles Heddle, 
 (Capt. Finck) was a 
 Mauritius brig, engaged 
 in the cattle' 
 trade, and 
 having been 
 
 IV^BnEHRHBHBr 
 
 built as a slaver, 
 was considered a 
 very fast sailer. 
 She left Mauri- 
 tius towards the 
 end of February, 
 
 1845 (other vessels having started a day 
 or two earlier), and after clearing the 
 island, took a northerly course. Towards 
 evening of the day of sailing she appears 
 to have fallen into a heavy gale, which 
 eventually proved to be the outer verge of a hurricane 
 advancing from the N.E.-ward. Whether she put her 
 helm up in order "to run out of the hurricane" is not 
 clearly stated, but for Jive successive days she was apparently scudding 
 right before the wind, as shown by the following extract from her 
 log-book, and from which PIDDINGTON deduced the remarkable track 
 given in Fig. 13. (c shows the storm's centre.) 
 
OF THE LAW OF STORMS. 
 
 47 
 
 Log of the Brig CHARLES HEDDLE, of Mauritius, Captain FINCK ; (kept in 
 French), copied ly Captain ROYER, Master Attendant at that Port, 
 and translated by HENRY PIDDINGTON. Nautical Time. 
 
 H. 
 
 K. 
 
 F. 
 
 Courses. 
 
 Winds. 
 
 Remarks. 
 
 P.M. 
 1 
 
 2 
 3 
 
 5 
 5 
 
 K 
 
 4 
 4 
 
 NEbN 
 
 ESE 
 
 Variable 
 to S "E 
 
 Friday 21st to Saturday 22nd February, 1845. 
 Horizon very low (tres rapproche), thick 
 weather all round. Heavy sea, smart breeze, 
 
 4 
 
 o 
 
 5 
 
 4 
 
 
 l/U O Jy 
 
 under the large sails ; pumped every two hours. 
 
 5 
 
 6 
 
 
 
 
 Sea and wind gradually increasing, vessel 
 
 6 
 
 6 
 
 
 
 
 labouring greatly, weather squally and threaten- 
 
 7 
 
 7 
 
 
 
 
 ing all round; the squalls very heavy. At 9.30 
 
 8 
 
 7 
 
 4 
 
 
 
 p. M. the mainyard went in two in the slings, 
 
 9 
 
 7 
 
 
 
 
 clued up and furled main-topsail, unbent main- 
 
 10 
 
 7 
 
 
 
 
 sail, and secured the pieces of the mainyard on 
 
 11 
 
 6 
 
 
 
 
 the booms. In jib and mizen ; scudding under 
 
 12 
 
 6 
 
 
 
 
 the foresail, fore-topsail and fore-topmast stay- 
 
 A.M. 
 
 
 
 
 
 sail, to wait for daylight; heavy squalls and 
 
 1 
 
 6 
 
 4 
 
 North 
 
 SE 
 
 sea. Down main-topsail yard, and struck top- 
 
 2 
 
 6 
 
 4 
 
 
 SSE 
 
 gallant-mast. Noon, in close-reef fore-topsail. 
 
 3 
 
 4 
 
 6 
 6 
 
 4 
 4 
 
 
 
 The gale begins to make itself be felt ; scudding 
 under foresail, and fore-topsail. 
 
 5 
 
 6 
 
 
 
 
 
 6 
 
 6 
 
 
 
 
 
 7 
 
 7 
 
 
 
 
 
 9 
 
 10 
 
 8 
 
 
 
 
 Latitude by Account 16 42' S. 
 
 11 
 
 8 
 
 
 
 
 Longitude Account 57 45' E. 
 
 12 
 
 8 
 
 
 
 
 
 P.M. 
 
 
 
 
 
 
 1 
 
 *13 
 
 ... 
 
 WNW 
 
 ESE 
 
 Saturday 22nd to Sunday 23rd February. 
 
 2 
 3 
 
 4 
 
 13 
 12 
 11 
 
 ... 
 
 NW 
 NNW 
 North 
 
 SE 
 SSE 
 South 
 
 Very bad weather; frightful sea; blowing 
 very hard with incessant rain ; vessel taking in 
 seas over the quarter while scudding under the 
 
 5 
 
 11 
 
 
 
 
 foresail and close reefed fore-topsail. Pumped 
 
 6 
 
 11 
 
 
 
 
 every hour ; vessel labouring greatly from the 
 
 8 
 
 12 
 12 
 
 
 NNE 
 
 
 seas which swept over us. At 2 p. M. perceiving 
 that the head rope of the foresail had given way, 
 
 9 
 10 
 
 12 
 12 
 
 ... 
 
 NE 
 
 
 sent two hands to cut away the earings, and let 
 it come on deck ; saved the sail. The fore-top- 
 
 11 
 12 
 
 12 
 12 
 
 
 
 
 mast staysail halyards having given way hoisted 
 the sail by a tackle. Gale at its height, scud- 
 
 A.M. 
 
 
 
 
 
 ding right before the wind, as it continually 
 
 1 
 2 
 3 
 
 12 
 12 
 12 
 
 ... 
 
 ENE 
 East 
 
 
 veered round the compass ; pumps attended to ; 
 vessel labouring excessively. It being impos- 
 sible to clue up the fore-topsail without risking 
 
 4 
 
 12 
 
 
 
 
 severe damage, we resolved to run our chance 
 
 5 
 
 12 
 
 1 9 
 
 
 
 
 of what might happen. 
 
 7 
 
 Lu 
 
 12 
 
 
 ESE 
 
 
 N. B. No position is given on this day. H.P. 
 
 8 
 
 12 
 
 
 
 
 
 9 
 
 12 
 
 
 
 
 
 10 
 
 12 
 
 
 SE 
 
 
 * About is marked in the log. 
 
 11 
 
 12 
 
 ... 
 
 SSE 
 
 
 t These last winds and courses are so marked in the 
 
 12 
 
 12 ... 
 
 South 
 
 S~\\T 
 
 North 
 
 N"C1 
 
 log. I presume they mean to designate the changes be- 
 tween Noon and 1 A.M. on the next day, as a memoran- 
 
 
 
 
 W 
 
 E 
 
 ENE 
 
 dum of a gradual veering. H.P. 
 
 
 
 
 
 East 
 
 
 
 
 
 
 ESEt 
 
 
HISTORY AND DEVELOPMENT 
 
 Log of the Brig CHARLES HEDDLE continued. 
 
 H. 
 
 K. 
 
 F. 
 
 Courses. 
 
 Winds. 
 
 Remarks. 
 
 P.M. 
 
 
 
 
 
 Sunday 23rd to Monday 24th February. 
 
 1 
 2 
 3 
 
 12 
 12 
 12 
 
 ... 
 
 S W 
 
 ws w 
 
 NE 
 ENE 
 
 Weather always the same with a frightful sea. 
 Shipping from time to time very heavy seas ; 
 one filled the whole deck fore and aft with two 
 
 4 
 
 12 
 
 
 
 
 feet of water ; the larboard waist-board carried 
 
 5 
 
 6 
 
 , 7 
 8 
 9 
 
 12 
 12 
 12 
 12 
 12 
 
 ... 
 
 West 
 WNW 
 
 East 
 ESE 
 
 away ; much water going down the hatchways 
 and cabin scuttle, though all secured by tarpau- 
 lins. 4 P. M. fore-topsail blew away ; scudding 
 under bare poles ; the new fore-topmast staysail 
 giving way, saved it ; two men at the helm, 
 
 10 
 11 
 12 
 
 10 
 10 
 10 
 
 ... 
 
 N W 
 NNW 
 
 SE 
 SSE 
 
 vessel labouring greatly; storm always at the 
 same height ; winds veering round the compass 
 from hour to hour, and even in half an hour. * 
 
 A.M. 
 j 
 
 ift 
 
 
 
 
 Brought all the crew aft into the cabin to be 
 
 1 
 2 
 
 JLU 
 
 10 
 
 
 
 
 at hand, closed up the fore-scuttle. 
 
 3 
 
 11 
 
 ... 
 
 North 
 
 South 
 
 N.B. No position given on this day. H.P. 
 
 4 
 
 11 
 
 
 
 
 
 5 
 6 
 
 11 
 11 
 
 
 NNE 
 
 SS W 
 
 * The expression is "faisant le tour du compos d'heure 
 en Jieure et meme une demi heure," of which the literal 
 
 7 
 
 11 
 
 ... 
 
 NE 
 
 S W 
 
 translation would be, " going round the compass from 
 
 8 
 
 11 
 
 
 ENE 
 
 w s w 
 
 hour to hour and even in half an hour." What is meant 
 
 9 
 
 11 
 
 
 
 
 is evidently (by the log) that the wind was going round 
 the compass and changing every hour or every half hour. 
 
 10 
 
 11 
 
 
 
 
 H.P. 
 
 11 
 
 11 
 
 
 East 
 
 West 
 
 
 12 
 
 11 
 
 
 
 
 . 
 
 P.M. 
 
 
 
 
 
 Monday 24th to Tuesday 25th February. 
 
 1 
 
 2 
 3 
 
 11 
 
 12 
 12 
 
 
 ESE 
 
 WNW 
 
 The gale always at the same degree of strength, 
 but the squalls a little heavier, pumps always 
 in hand, vessel making water. All the cabins 
 
 4 
 5 
 6 
 
 12 
 12 
 11 
 
 ... 
 
 SE 
 
 N W 
 
 below wet, the provisions in the great cabin 
 also wet, the vessel making water through every 
 seam in the deck without exception ; baled the 
 
 7 
 
 11 
 
 
 
 
 water out of the cabin by buckets. 
 
 8 
 9 
 
 11 
 11 
 
 ... 
 
 SSE 
 South 
 
 NNW 
 North 
 
 Shipped several seas which went over all. 
 
 10 
 
 11 
 
 
 
 
 At two in the morning the vessel broached-to, 
 
 11 
 12 
 
 11 
 11 
 
 
 
 
 the water two feet deep on the deck; we re- 
 mained in this dangerous position for about ten 
 
 A.M. 
 
 
 
 
 
 minutes, when she righted. We broached-to 
 
 1 
 
 11 
 
 
 ss w 
 
 NNE 
 
 again several times from the speed of the vessel ; t 
 
 2 
 
 11 
 
 
 s w 
 
 NE 
 
 cleared the scuppers. At 10, shipped a sea in 
 
 3 
 
 4 
 
 11 
 11 
 
 ... 
 
 ws w 
 
 ENE 
 
 the fore rigging which carried away jib and fly- 
 ing jib-boom. Cut away the wreck to clear the 
 
 5 
 
 11 
 
 
 
 
 bowspit. 
 
 6 
 
 11 
 
 ... 
 
 West 
 
 East 
 
 Latitude by a doubtful Observation 16 18' S. 
 
 7 
 8 
 
 11 
 11 
 
 ... 
 
 WNW 
 
 NTIT 
 
 ESE 
 
 Longitude Chronometer 53 2' 30" E. 
 
 9 
 
 11 
 
 ... 
 
 W 
 
 S E 
 
 t The words are "par la, vitesse du bailment." No 
 
 11 
 
 11 
 
 
 NNW 
 
 SSE 
 
 doubt the difficulty of steering her is here implied. H.P. 
 
 12 
 
 11 
 
 
 
 
 
OF THE LAW OF STORMS. 
 
 Log of the Brig CHARLES HEDDLE. continued. 
 
 49 
 
 H. 
 
 K. 
 
 F. 
 
 Courses. 
 
 Winds. 
 
 Remarks. 
 
 P.M. 
 1 
 
 11 
 
 
 North 
 
 South 
 
 Tuesday 25th to Wednesday 26th February. 
 
 2 
 
 11 
 
 ... 
 
 NNE 
 
 SS W 
 
 The gale always at the same strength with- 
 
 3 
 
 11 
 
 
 
 
 out the least intermission, heavy sea and rain. 
 
 4 
 5 
 
 10 
 10 
 
 ... 
 
 NE 
 ENE 
 
 S W 
 
 wsw 
 
 The tiller ropes gave way ; changed them ; then 
 the bolts also of the tiller having given way, 
 
 6 
 
 10 
 
 ... 
 
 East 
 
 West 
 
 drove in preventer ones. 
 
 7 
 8 
 9 
 10 
 11 
 
 10 
 10 
 10 
 10 
 10 
 
 ... 
 
 ESE 
 
 SE 
 SSE 
 South 
 
 WNW 
 
 NW 
 NNW 
 North 
 
 Pumped ship every hour. The trusses of the 
 fore-yard gave way, replaced them ; scudding 
 under bare poles. The sea frightful, vessel 
 making much water through the deck. 
 
 12 
 
 10 
 
 
 S S W 
 
 NNE 
 
 Crew worn out with fatigue. The sun ap- 
 
 A.M. 
 1 
 
 10 
 
 
 SW 
 
 NE 
 
 peared indistinctly at noon, whereby we ob- 
 tained an indifferent latitude and longitude. 
 
 2 
 
 11 
 
 ... 
 
 WSW 
 
 ENE 
 
 
 3 
 
 10 
 
 ... 
 
 West 
 
 East 
 
 
 4 
 
 10 
 
 ... 
 
 WNW 
 
 ESE 
 
 
 5 
 
 10 
 
 ... 
 
 N W 
 
 SE 
 
 
 6 
 
 10 
 
 ... 
 
 NNW 
 
 SSE 
 
 
 7 
 
 10 
 
 ... 
 
 North 
 
 South 
 
 
 8 
 
 10 
 
 ... 
 
 NNE 
 
 SS W 
 
 
 9 
 
 10 
 
 ... 
 
 NE 
 
 SW 
 
 
 10 
 
 10 
 
 ... 
 
 ENE 
 
 wsw 
 
 Latitude by indifferent Observation 18 2' S. 
 
 11 
 
 10 
 
 ... 
 
 East 
 
 West 
 
 Longitude, ditto, ditto ... 51 2' 30" E. 
 
 12 
 
 10 
 
 ... 
 
 ESE 
 
 WNW 
 
 
 P.M. 
 
 
 
 
 
 Wednesday 26th to Thursday 27th February. 
 
 1 
 
 2 
 3 
 4 
 5 
 
 10 
 10 
 
 9 
 9 
 9 
 
 
 SSE 
 South 
 
 N W 
 NNW 
 
 North 
 
 The horizon always obscure though some- 
 times clearing a little, but the squalls and sea 
 always heavy ; pumped every hour. Two men 
 at the helm. Always under bare poles. At 
 
 6 
 7 
 8 
 9 
 
 9 
 10 
 10 
 10 
 
 ... 
 
 SS W 
 WSW 
 West 
 NW 
 
 NNE 
 ENE 
 East 
 S F 
 
 10 p.m. clearing up a little, and we saw some 
 stars, but the sea and wind always heavy. 
 Bent fore-topmast staysail, and fore and aft 
 mainsail with two reefs in it. Bent another 
 
 10 
 
 10 
 
 
 vv 
 
 NNW 
 
 O I * 
 
 SSE 
 
 fore-topmast staysail on the forestay to balance 
 
 11 
 12 
 
 AM 
 
 10 
 10 
 
 ... 
 
 North 
 
 NNE 
 
 South 
 SS W 
 
 the vessel's sails. * 
 Scudding always according to the veering of 
 the wind. 
 
 . JM. 
 
 1 
 
 2 
 3 
 
 4 
 
 10 
 10 
 10 
 10 
 
 ... 
 
 NE 
 ENE 
 East 
 ESE 
 
 S W 
 WSW 
 West 
 WNW 
 
 Seeing that we had sustained much damage, 
 and that we were nearer to Mauritius than to 
 any other place, the Captain resolved to return 
 there, not considering the vessel in a state to 
 
 5 
 
 10 
 
 
 SE 
 
 NW 
 
 continue her voyage. 
 
 6 
 
 10 
 
 
 SSE 
 
 NNW 
 
 Latitude Observation 20 12' S. 
 
 7 
 
 10 
 
 ... 
 
 South 
 
 North 
 
 Longitude Chronometer ... 52 24' E. 
 
 3 
 
 10 
 
 
 SS W 
 
 NN "E 
 
 
 9 
 
 10 
 
 ... 
 
 O W 
 
 S W 
 
 1.1 -Cj 
 
 NE 
 
 * i.e. When sail should be made, having lost the jib-boom. 
 
 10 
 
 10 
 
 ... 
 
 wsw 
 
 ENE 
 
 
 11 
 
 10 
 
 ... 
 
 West 
 
 East 
 
 
 12 
 
 10 
 
 
 
 
 
50 HISTORY AND DEVELOPMENT OP THE LAW OF STORMS. 
 
 Log of the Brig CHARLES HEDDLE. continued. 
 
 H. 
 
 K. 
 
 F. 
 
 Courses. 
 
 Winds. 
 
 Remarks. 
 
 P.M. 
 
 1 
 
 7 
 
 
 SE 
 
 ENE 
 
 Thursday 27th to Friday 28th February. 
 
 2 
 3 
 
 7 
 7 
 
 ... 
 
 SSE 
 ESE 
 
 Variable 
 toNE 
 
 The weather becoming fine, bent the foresail 
 and spare fore-topsail, took the main-topsail 
 
 4 
 
 7 
 
 
 
 
 yard for a mainyard, and let the reefs out of 
 
 5 
 
 7 
 
 
 
 
 the fore and aft mainsail. 
 
 6 
 
 7 
 
 6 
 6 
 
 4 
 4 
 
 
 
 Cloudy still, and lightning in all quarters. 
 Fine, and sea smooth with a pleasant breeze. 
 
 8 
 
 6 
 
 4 
 
 
 
 
 9 
 
 6 
 
 
 
 
 
 10 
 
 6 
 
 
 
 
 Latitude Observation ... 23 19' S. 
 
 11 
 
 6 
 
 
 
 
 Longitude Chronometer ... 54 22' 28" 
 
 12 
 
 6 
 
 
 
 
 At end of day. 
 
 It must be noticed that PIDDINQTON was indefatigable when he got hold of a log- 
 book or a set of log-books that would illustrate the law of cyclones ; and in respect 
 to the log of the Charles Heddle he must have spent much time and labour far 
 more than it was worth in determining in a variety of ways such results as it would 
 give ; but his main and most valuable achievement was the diagram of the ship's 
 course from day to day, as shown in Fig. 13, p. 46, which, no doubt, is approxi- 
 mately correct. The brig's course and distance by log was about N. 38 E. 102 
 miles, while the course and distance by D.R. and Obs. combined was S. 55 W. 370 
 miles ; which would give a current or storm- wave running in the direction of S. 51 
 W., about 468 miles in five days, or at the average rate of about 4 miles per hour. 
 But the log, on the face of it, and by comparing the courses, winds, and distance 
 with the remarks, gives no sign of strict accuracy, notwithstanding PIDDINGTON'S 
 eulogy : it may be tried in a variety of ways, with an infinity of results none satis- 
 factory: the only fact that comes broadly out is, that in about the space of 120 
 hours, when scudding (according to the account) before the wind, the brig made five 
 complete circuits of the compass the 1st in 31 hours, the 2nd in 35 hours, the 3rd 
 in 21 hours, the 4th in 17 hours, and the 5th in 16 hours whether strictly before 
 the wind, all the time, is however open to question. But one thing is certain, and 
 which alone would confute the log, if the winds in the storm have circular motion, 
 together with a general tendency (progression) towards the S.W.-ward, then (also of 
 necessity) the resultant of the southerly and westerly courses must be in excess of 
 that in any other direction, supposing a vessel to be running before the wind and 
 this too independently of any current. Log-books, kept in the ordinary way, are 
 but a poor basis on which to found or establish a theory. 
 
THE CIRCULAR THEORY OF STORMS. 51 
 
 THE CIRCULAR THEORY OP STORMS, AND 
 PRACTICAL RULES BASED THEREON. 
 
 From whatever cause a steady wind is so obstructed as to produce a 
 Hurricane, Typhoon or Cyclone, the wind then assumes a rotatory 
 motion over an area varying from 20 or 30 to some hundreds of miles 
 in diameter, revolving the more rapidly the nearer the centre or vortex, 
 up to a certain distance, within which there is a calm ; meanwhile the 
 centre of rotation, together with the body of the storm, has a progressive 
 motion varying from 2 to 20 or 40 miles per hour, with occasional inter- 
 mitting intervals, during which it appears to be almost stationary or 
 even retrograde, and the impetuosity of the wind itself around the 
 circle may continue undiminished. Such storms have also been desig- 
 nated Revolving Storms. 
 
 flotation of the Wind- Vane in Hurricanes or Cyclones. In the same 
 hemisphere the wind always revolves in the same way, and that is con- 
 trary to the apparent course of the sun ; hence this direction is opposite 
 in opposite hemispheres, being from right to left in the northern, but 
 from left to right in the southern hemisphere. Or, to use the familiar 
 illustration of the hands of a watch laid face upwards, the rotation of 
 the wind during a hurricane, in the Northern Hemisphere, is in a 
 direction contrary to that in which the hands move (Fig. 16, p. 53), 
 while in the Southern Hemisphere the rotation is in the same direction 
 as that of the hands of a watch (Fig. 17, p. 53.) 
 
 Progressive Movement. The habitual tracks of hurricanes, if they 
 have any, is unknown ; but being prevalent alike in the three oceans the 
 Atlantic, Indian and Pacific they follow at certain seasons tolerably 
 definite routes, so that there must be few navigators who have never 
 crossed the path of one. They are however rarely found within 6 to 
 8 of the equator, but they rage with most fearful violence in the tropical 
 regions, and, as a rule, take a curved route approximating to a parabola. 
 
 The hurricane is impelled to the west in low latitudes because the 
 tendency of the two currents of air (polar and equatorial) is in those 
 regions to the westward along the surface; the equatorial current is 
 much less so, and diminishing until actually altered to easterly, near the 
 Tropic, after which its preserved equatorial rotatory force becomes more 
 and more evident, while the westwardly tendency of the polar current 
 diminishes; as a consequence, near the Tropic, the whirlwind ceases to 
 move westward it recurves and then in its polarwise progression 
 moves towards the eastward. 
 
52 THE CIRCULAR THEORY OF STORMS, AND 
 
 In the Northern Hemisphere, commencing about Lat. 10 to 20 N. 
 the hurricane first advances from E.S.E., progressing towards W.N.W. 
 with a northerly tendency as it approaches the northern tropic, where its 
 course for some distance is nearly due North ; somewhere in the vicinity 
 of the parallels of 25 
 to 32 N. it recurves, 
 and invading the 
 extra-tropical regions, 
 continues its pro- 
 gression, advancing 
 from S.W. towards 
 the N.E. increasing 
 in diameter, while the 
 wind, though still of 
 fearful violence, has 
 somewhat diminished 
 in force as compared 
 with its rotatory velo- 
 city while within the 
 tropics, and so it 
 moves onward, until 
 it finally breaks up (Fig. 14). Such is especially the character of the 
 West India hurricanes, which range between the parallels of 10 and 
 50 N., and between the meridians of 30 and 103 W., though the 
 majority of them are 
 included between 55 
 and 85* W. Most of 
 the storms that occur 
 on the western side of 
 the North Pacific and 
 in the vicinity of Japan 
 affect the same form, 
 but those off" the west 
 coast of Mexico and 
 Lower California 
 merely take a N.W.-ly 
 direction of compara- 
 tively small extent. 
 The cyclones of the 
 Bay of Bengal, and the 
 Arabian Sea, as also 
 the typhoons of China, take now a straight, and now a curvilinear 
 course, due to local circumstances and the interference of the land. 
 
PRACTICAL RULES BASED THEREON. 
 
 53 
 
 In the Southern Hemisphere, the cyclone, commencing in the equa- 
 torial region, first advances from E.N.E., progressing towards the 
 W.S.W. with a more and more southerly tendency in approaching 
 the southern tropic, in the vicinity of which its course is nearly due 
 South; it then commences to recurve, and continues its progressive 
 movement, advancing from N.W. towards S.E., expanding in dimen- 
 sions and probably diminishing in force as it traverses the extra- 
 tropical regions, until it finally disappears (Fig. 15). Cyclones occur 
 in the Southern Indian Ocean between the N.W. coast of Australia and 
 the northern end of Madagascar, and range from the parallels of 6 to 
 40 S. ; but they are unknown in the tropical region of the South 
 Atlantic. A few have been experienced in the South Pacific, but 
 hitherto there have been few good records of them. 
 
 Deductions from the Order of Rotation. Bearing in mind the 
 direction of the rotation of the wind in each hemisphere, viz., against 
 watch hands in the Northern, but with watch hands in the Southern, 
 Hemisphere certain portions of the storm will be characterized by 
 certain hurricane winds ; and 
 considering the storm to be 
 divided into four distinct 
 parts, it will be found that, 
 in the Northern Hemisphere, 
 on the northern margin the 
 wind must be easterly; on 
 the eastern margin, southerly; 
 on the southern margin, west- 
 erly ; and on the western 
 margin, northerly ; and thus 
 each quarter of the cyclone 
 is distinguished by its appro- 
 priate wind. (See Fig. 16.) 
 
 The relation of the winds 
 to the margins in the South- 
 ern Hemisphere will be ex- 
 actly the reverse of their 
 relation in the Northern, 
 owing to the reversion of the 
 rotation ; thus, it is the 
 southern margin of the 
 storm, south of the equator, 
 that exhibits an easterly ; 
 the western margin, a southerly ; the northern margin, a westerly ; and 
 the eastern margin, a northerly wind. (See Fig. 17.) 
 
54 THE CIRCULAR THEORY OF STORMS, AND 
 
 Bearing of the Storm's Centre. Also, as each portion of the hurricane 
 has its appropriate wind, there results, according to the Law of the 
 Rotation of Revolving Gales, a very simple Rule for determining the 
 bearing of the centre of the storm from the ship viz., Look to the wind's 
 eye, and set its bearing by compass, the EIGHTH point to the RIGHT 
 thereof, when in the Northern Hemisphere but to the LEFT of the wind's 
 direction when in the Southern Hemisphere will be the bearing of the 
 storm's centre. 
 
 Hence, in the NORTHERN HEMISPHERE, from an easterly wind which 
 characterizes the northern margin of the storm, its centre will bear 
 south ; from a northerly wind the centre will bear east ; from a 
 westerly wind it will bear north; and from a southerly wind, west. 
 (See Fig. 16.) 
 
 But, in the SOUTHERN HEMISPHERE, from an easterly wind, the 
 centre of the storm bears north; from a southerly wind the centre 
 bears east ; from a westerly wind, south ; and from a northerly wind, 
 west. (See Fig. 17.) 
 
 It is well also to note that it is characteristic of the hurricanes of 
 both hemispheres that westerly winds are found on their equatorial side, 
 and easterly winds on their polar side, throughout their entire course, 
 after recurvature no less than before recurvature. 
 
 N.B. In Figs. 16 and 17, the line Q Q represents the Equator, 
 and the arrows are to be taken as flying with the wind. These arrows 
 (with the letters between them) show the direction of the wind for each 
 hemisphere, according to the circular theory; the letters outside the 
 arrows represent the points of the geographical horizon ; is the storm's 
 centre; on this basis the two diagrams (16 and 17) fully illustrate the 
 preceding paragraphs viz., the wind blowing against watch hands in 
 the Northern Hemisphere, but with watch hands in the Southern Hemi- 
 sphere ; and also, when facing the wind, the centre or area of lowest 
 pressure is eight points to the right in the N. Hemisphere, but eight points 
 to the left in the S. Hemisphere. Ex. In Fig. 16 on the N.W. side of 
 the storm the wind will be N.E. and (when facing it) the centre will 
 bear S.E., since it represents the N. Hemisphere; but, on the N.W. 
 side of a cyclone in the S. Hemisphere (see Fig. 17) the wind will be 
 S.W. and the centre bear S.E. ; thus Fig. 16 is a storm-compass for the 
 Northern Hemisphere, and Fig. 17 a storm-compass for the Southern 
 Hemisphere. 
 
 As it is especially important to avoid the centre of the storm, the 
 following Table shows at a glance (according to the circular theory) its 
 relative bearing in each hemisphere : 
 
PRACTICAL RULES BASED THEREON. 
 
 55 
 
 IN THE N. ] 
 
 HEMISPHERE. 
 
 IN THE S. I 
 
 EEMISPHERE. 
 
 If the wind be 
 
 The centre of the 
 Storm will bear 
 from the Ship. 
 
 If the wind be 
 
 The centre of the 
 Storm will bear 
 from the Ship. 
 
 North 
 N. by E. . . 
 N.N.E. . . . 
 
 East. 
 E. by S. 
 E.S.E. 
 
 North . 
 N. by E. 
 N.N E. . 
 
 West. 
 W. by N. 
 WN W 
 
 N.E. byN.. . 
 N.E. . . 
 
 S.E. by E. 
 S.E. 
 
 N.E. by N. . 
 
 N.E. . 
 
 N.W. by W. 
 N W ' 
 
 N.E. by E. . . 
 E.N.E. . . . 
 E. by N. . . 
 East .... 
 E. by S. . . . 
 E.S.E. . . . 
 S.E. by E. . . 
 S.E. . . 
 
 S.E. by S. 
 S.S.E. 
 S. by E. 
 South. 
 S. by W. 
 S.S.W. 
 S.W. by S. 
 S.W. 
 
 N.E. by E. 
 E.N.E. . 
 E. by N. 
 East . . 
 E. by S. . 
 E.S.E. . 
 S.E. by E. 
 S E . . 
 
 N.W. byN. 
 
 N.N.W. 
 N. by W. 
 North. 
 N. by E. 
 
 N.N.E. 
 N.E. byN. 
 N E " 
 
 S.E. byS. . . 
 S.S.E. . . . 
 
 S.W. by W. 
 WS W. 
 
 S.E. by S. 
 
 S S E 
 
 N.E. byE. 
 E N E 
 
 S. by E. . . . 
 South . 
 S. by W. . . 
 S.S.W. . . . 
 S.W. by S. . . 
 S.W 
 
 W. by S. 
 West. 
 W. by N. 
 W.KW. 
 N.W. by W. 
 N.W. 
 
 S. by E. . 
 South . 
 S. by W. 
 S.S.W. . 
 S.W. by S. 
 S.W 
 
 E. by N. 
 East. 
 E. by S. 
 E.S.E. 
 S.E. byE. 
 S E. ' 
 
 S.W. by W. . 
 W.S.W. . . . 
 
 N.W. by N. 
 
 N.N.W. 
 
 S.W. by W. . 
 W S W . . 
 
 S.E. by S. 
 S S E 
 
 W. by S. . . 
 West .... 
 W. by N. . . 
 W.N W 
 
 N. by W. 
 North. 
 
 N. by E. 
 
 N N E 
 
 W. by S. 
 West . . 
 W. by N. 
 
 w N'W 
 
 S. by E. 
 South. 
 S. by W. 
 S S W 
 
 N.W. byW. . 
 
 N W 
 
 N.E. by N. 
 NE ' 
 
 N.W. by W . 
 
 N W 
 
 S.W. byS. 
 S W 
 
 N.W. by N. . 
 N.N.W. . . 
 
 N.E. by E. 
 E.N.E. 
 
 N.W. by N. . 
 N.N.W. . . . 
 
 S.W. by W. 
 W.S.W. 
 
 N. by W. . . 
 
 E. by N. 
 
 N. by W. . . 
 
 W. by S. 
 
 Note. It will, however, be shown presently that this Table may not be strictly correct in all 
 cases, for instance, when the storm- winds have an incurvature towards the area of lowest pressure ; 
 but it is generally correct for the front of the storm. 
 
 The Storm's Path, or Axis Line of the Storm's Progression. The 
 line along which the centre of the hurricane progressively moves is 
 variously called the storm-path, the storm's track, and the axis line ; 
 or, not inaptly, by the French, the trajectory as indicative of a 
 parabolic track in space. (See the dotted line in Figs. 14 and 15, 
 p. 52.) 
 
 A ship, in front of the storm, sailing, or lying-to, on the axis line 
 would have no change of wind until the calm centre had passed over 
 her, after which the wind would spring up suddenly with great fury 
 
56 
 
 THE CIRCULAR THEORY OF STORMS, AND 
 
 from the opposite direction to that by which she had entered the 
 centre. 
 
 In extra-tropical regions a cyclonic storm does not always develop a 
 central calm, as is almost invariably the case within the tropics. 
 
 When approaching the centre, and 
 being off the axis line, the changes in the 
 wind will necessarily become more and 
 more rapid, and the sea more and more 
 confused and heaped up. 
 
 For any given fixed point in the storm- 
 field, the' slower the progressive move- 
 ment, the less rapid will be the change in 
 the direction of the wind. And should 
 the storm-area become temporarily sta- 
 tionary, or nearly so, as is sometimes the 
 case, the hauling of the wind, due to the 
 ship's progressive motion (if it has any), 
 may be contrary to the usual order. 
 
 The axis line, or track of the hurricane, divides the storm-field into 
 two parts, and, looking towards the direction in which the storm is 
 moving, the part to the right is the right-hand semicircle, and that to 
 the left is the left-hand semicircle. Another line, transverse to the 
 axis line and passing through the calm centre, 
 divides the storm-field into four quadrants, 
 each of which has its characteristic pre- 
 dominant wind, as seen in Fig. 18, and as 
 already explained on pp. 53-54. 
 
 French authors, speaking of the two sides 
 of the storm-field, have applied the term 
 "dangerous" (dangereux) to the semicircle 
 in which the rotatory and progressive 
 motion of the winds is approximately in 
 the same direction ; and the term " manage- 
 able " (maniable), or perhaps in this case better 
 expressed as "navigable," to the semicircle 
 in which the two motions are approximately 
 opposed to each other. These two sides, in 
 Fig. 19, are respectively designated by the 
 letters D and N, and a glance at the long 
 curved arrows, one north and the other south 
 of the equator, which represent for each hemi- 
 sphere the storm's track, shows that, in both 
 
PRACTICAL RULES BASED THEREON. 57 
 
 hemispheres, the dangerous semicircle is inside or to eastward of the 
 parabolic curve, and the navigable semicircle outside or to westward ; 
 or, in relation to right and left of the track, in opposite hemispheres 
 the positions are reversed, thus 
 
 In the Northern Hemisphere, the dangerous semicircle is to the 
 right, and the navigable semicircle to the left, of the track : but, 
 
 In the Southern Hemisphere, the dangerous semicircle is to the 
 left, and the navigable semicircle to the right, of the track. 
 
 In both cases, however, the dangerous semicircle is on that side of 
 the storm towards which the curvature of the path takes place. 
 
 The terms dangerous and navigable are not wholly inappropriate; 
 for, when in the dangerous semicircle and running before the wind, the 
 tendency is ever towards the axis line and the centre of the storm ; but 
 if in the navigable semicircle, except with bad management and being 
 on the wrong tack when in front of the storm, the tendency is ever 
 away from the centre. 
 
 On the basis just indicated the winds for each semicircle are as 
 follows : 
 
 In the NORTHERN HEMISPHERE : 
 
 Storm moving towards N.W. 
 
 In dangerous semicircle, winds from N. E. to S.W. round by E. and S. 
 In navigable semicircle, winds from N.E. to S.W. round by N. and W. 
 
 Storm recurving to Northward 
 
 In dangerous semicircle, winds from East to West round by S. 
 In the navigable semicircle, winds from East to West round by N. 
 
 Storm moving towards N.E. 
 
 In dangerous semicircle, winds from S. E. to N.W. round by S. and W. 
 In navigable semicircle, winds from S.E. to N.W. round by E. and N. 
 
 In the SOUTHERN HEMISPHERE : 
 
 Storm moving towards S.W. 
 
 In dangerous semicircle, winds from S.E. to N.W. round by E. and N. 
 In navigable semicircle, winds from S.E. to N.W. round by S. and W. 
 
 Storm recurving to Southward 
 
 In dangerous semicircle, winds from East to West round by N. 
 In navigable semicircle, winds from East to West round by S. 
 
 Storm moving towards S.E. 
 
 In dangerous semicircle, winds from N.E. to S.W. round by N. and W. 
 In navigable semicircle, winds from N.E. to S.W. round by E. and S. 
 
58 
 
 THE CIRCULAR THEORY OF STORMS, AND 
 
 Position of Ship recognised by the veering or by the backing of the 
 Winds. It may be further stated in elucidation of the hauling of the 
 wind with or against watch hands that for a ship hove-to 
 
 (1.) In the NORTHERN HEMISPHERE, and in the dangerous (right- 
 hand) semicircle, the winds must necessarily veer, in their pro- 
 gressive motion over the ship, but in the navigable (left-hand) 
 semicircle they will back : 
 
 (2.) In the SOUTHERN HEMISPHERE, and in the dangerous (left-hand) 
 semicircle, the winds will back, but in the navigable (right- 
 hand) semicircle they will veer : 
 
 But better still is the consideration of the storm-field, as divided into 
 four quadrants, as in Figs. 20 and 21 ; from which it can readily be 
 
 seen that the anterior quadrants (a c), that is, those in front of the ad- 
 vancing centre, must, in each hemisphere, be more dangerous than the 
 posterior (b d); while of the two anterior quadrants, since, in one of 
 them the wind blows towards, and in the other away from, the axis 
 line 
 
 The advancing quadrant (Fig. 20 a) of the right-hand semicircle in 
 the Northern Hemisphere, but the advancing quadrant (Fig. 21 a) of 
 the left-hand semicircle in the Southern Hemisphere, must be the MOST 
 
PRACTICAL RULES BASED THEREON. 59 
 
 DANGEROUS in which, if a ship be put before the wind, she must 
 inevitably approach nearer and nearer to the storm's track and centre ; 
 but running, be it observed, is a manoeuvre only justifiable in the case 
 of the vessel being on the verge of the storm, and so near the axis line 
 that she can speedily cross it, and thus be brought into the safer of the 
 two advancing quadrants. The posterior quadrants can never come 
 down on a vessel, but, with plenty of sea-room, may be entered, and the 
 winds turned to advantage, giving due attention at the same time to 
 the indications of the barometer, and the state of the weather and sea. 
 
 Briefly stated, the storm paths over the Atlantic Ocean are charac- 
 terized by certain winds that are most dangerous to vessels falling in 
 with them according to the locality as follows : 
 
 (1.) Before recurving off the West India Islands the most danger- 
 ous hurricane winds are those between E.N.E. and N.N.E. 
 
 (2.) About the period of recurving, vessels off the Bahamas and the 
 coast of Florida find East, E.S.E., and S.E. winds most 
 dangerous. 
 
 (3.) After recurving, vessels off the coast of the United States find 
 E.S.E. to S.S.E. and South winds most dangerous. 
 
 In the Southern Indian Ocean, if the cyclone be moving towards 
 the S. W. the most dangerous winds are those between South and East ; 
 but when moving towards the S.E., winds between East and North 
 are the most dangerous. At the period of recurving, winds between 
 N.E. and S.E. are most dangerous. 
 
 So that in both hemispheres winds inclining to westerly are compara- 
 tively safe, while such as have easting in them require caution, skill and 
 judgment in manoeuvring the ship, under whatever circumstances she 
 encounters them whether by running into, or being overtaken by, a hur- 
 ricane, and especially in the latter case, because if these winds change 
 but little, and the barometer is falling rapidly, the ship is directly in front 
 of the storm, and probably not far from the storm's progressive path. 
 
 Use of the Barometer. The words "fair," "change," "rain," <fcc., 
 as marked on most barometers, have ceased to have any significance for 
 the meteorologist. It is of more importance to know the mean pressure 
 of the atmosphere for the locality and season ; and to note the baro- 
 metric oscillations as they occur. 
 
 It was a trite observation of PIDDINGTON'S that " he who watches 
 his Barometer watches his ship." This invaluable instrument, if well 
 understood and watched, invariably announces the approach of a revolv- 
 ing storm, shows whether the vessel is plunging into the vortex, or if 
 she be receding from it and hence, by carefully noticing its indications, 
 the disastrous consequences of a hurricane may, to a great extent, be 
 
60 THE CIRCULAR THEORY OF STORMS, AND 
 
 avoided, for the laws of its oscillations are appreciably characteristic 
 and very distinctly marked. The barometer often stands unusually 
 high before the commencement of a hurricane, and frequently (if not 
 always) just around the storm area; and conceiving the hurricane to 
 be divided, as before, into two parts by a diameter at right angles to 
 its track, it may be noted that in both hemispheres 
 
 (a) The barometer always falls during the passage of the advancing 
 semicircle of a revolving storm ; and again 
 
 (6) The barometer always rises during the passage of the receding 
 semicircle of a revolving storm : 
 
 hence, in cases of manoeuvring to take advantage of hurricane winds and 
 to keep just within the verge of the storm, the barometer is of signal 
 benefit ; it should be kept as high as possible without losing the wind. 
 When sailing from the vortex the barometer will rise ; when sailing 
 towards the vortex it will fall ; and if, after having fallen considerably, 
 it rises at first rapidly, and then gradually such rising indicates that 
 the first (or dangerous) half of the storm has passed away. 
 
 The following observations have already been given, but they may 
 be repeated with advantage. 
 
 Beyond the tropics, and in the higher latitudes, the oscillations of the 
 barometer are incessant and irregular, ranging (rising and falling) 
 through a space of 1J to 2 inches, or even 3 inches on extraordinary 
 occasions; and thus its indications fail at times to convey the same 
 definite information that they do between the tropics, unless interpreted 
 by the aid of the thermometer and hygrometer, in conjunction with 
 various atmospheric and oceanic appearances. 
 
 Between the tropics the oscillations of the barometer are also incessant, 
 but they are regular. The periodical range due to the season does not 
 exceed a few tenths ; but the most marked of the periodical oscillations 
 is that which occurs daily. It shows every day two maxima and two 
 minima, as follows, according to the locality; at some time between 
 9 and 11 a.m. the mercury begins to fall, and continues to do so until 
 between 3 to 5 p.m., when it reaches the lowest point of depression ; 
 it then rises until between 9 to 11 p.m., when it reaches the highest 
 point of elevation; after that it recommences to fall, falling until 
 between 3 to 5 a.m., from which time it again rises until between 
 9 to 11 a. m. The amount of range, again according to locality, varies 
 from -06 to *12 of an inch. Any disturbance of the regularity of the 
 diurnal oscillation within the tropics is significant of approaching danger. 
 
 Tack on which to heave-to. When involved in a hurricane, in the 
 Northern Hemisphere, if the ship is on the starboard tack her head is 
 
i 
 PRACTICAL RULES BASED TH 
 
 directed from the centre ; but if on the port tack, her head is towards 
 the centre. In the Southern Hemisphere, as a necessary consequence 
 of the rotation of the wind, the conditions are reversed, and thus the 
 starboard tack leads towards, and the port tack from, the centre of the 
 storm. But nevertheless, as a general rule, since it is advisable to 
 heave a ship to so that she shall come up as the wind shifts, the rule 
 will be, in both hemispheres 
 
 (a) When in the right-hand semicircle heave-to on the starboard 
 tack; but 
 
 (b) When in the left-hand semicircle heave-to on the port tack. 
 
 Running is never recommended, except as a necessity, owing to the 
 danger of broaching to if the wind is brought on the quarter; but if 
 compelled to run 
 
 (a) Keep the wind well on the starboard quarter in the northern 
 hemisphere ; 
 
 (b) Keep the wind well on the port quarter in the southern hemi- 
 sphere. 
 
 Heaving-to will always be more or less dangerous when in front of 
 the cyclone and near the axis line, but having reason to believe the ship 
 to be in that position, avoid heaving-to under the following circumstances, 
 unless compelled to do so : 
 
 When in the Northern Hemisphere 
 
 Storm moving N.W.-ly, winds between N.N.E. and E.N.E. 
 
 Storm recurving, winds between E.N.E. and E.S.E. 
 
 Storm moving N.E.-ly, winds between E.S.E. and S.S.E. 
 When in the Southern Hemisphere 
 
 Storm moving S.W.-ly, winds between S.S.E. and E.S.E. 
 
 Storm recurving, winds between E.S.E. and E.N.E. 
 
 Storm moving S.E.-ly, winds between E.N.E. and N.N.E. 
 General Rule. The elder cyclonologists were very cautious about 
 giving rules for manoeuvring. Their rules may be summarized as 
 follows heave-to for the first hour to two (when overtaken by a 
 cyclone) to determine your position in the storm, by noting the changes 
 of wind and state of barometer ; then, if it be deemed safe to stand on, 
 do so ; should bad weather increase, with a falling barometer, heave-to 
 again, and so continue, until new changes indicate that standing on 
 may once more be attempted ; if, having done so, there is no improve- 
 ment, exercise your own judgment as to heaving-to a third time. 
 
 PIDDINGTON was wise in his generation, but far too severe a critic, 
 for it is often easy to see, after the event, how misadventure has 
 
62 THE CIRCULAR THEORY OF STORMS, AND 
 
 occurred; but none of the cyclonologists ever wrote more practical 
 directions than he did, when in later days he penned the following : 
 ALL POSITIVE rules for manoeuvring are sheer nonsense, and only tend to 
 mislead : every ship in every cyclone must have its own peculiar manage- 
 ment, dependent on the four great elements of the problem, which are (1) 
 the ship and her sea-room; (2) the track of the cyclone; (3) its rate of 
 travelling ; and (4) the ship's run and drift : while on the part of the 
 commander, caution and watchfulness are essentially requisite. 
 
 The meteorological signs tJiat herald the approach of, and that usually 
 accompany, a hurricane or cyclone have been carefully collected from 
 various sources by W. R. BIRT, and a summary of them is given in 
 that author's "Handbook of the Law of Storms." In enumerating and 
 collating them he has classified them under several distinct and charac- 
 teristic headings, as follows : 
 
 " Meteorological signs recognised by the feelings 
 
 1. A sultry, oppressive state of the atmosphere. 
 
 2. A calm. 
 
 " Meteorological signs recognised by the eye 
 
 3. A remarkably clear state of the atmosphere, so that the stars 
 may be seen to rise and set with nearly the same distinctness as 
 the sun and moon. 
 
 4. A peculiar white appearance in the zenith, more or less of a 
 circular form. 
 
 5. A remarkably red or fiery appearance of the sky. 
 
 (This is not unfrequently of such intensity as to tinge all surrounding 
 objects with a deep crimson ; and when this is observed, there can be no 
 question that the violent portion of the cyclone is not far from the vessel. 
 When this red light is seen at night, the impression on the seaman's mind 
 is, that 'day has broken before its time.') 
 
 6. A peculiar colouring of the clouds, more especially towards 
 an olive green tinge. 
 
 (This is generally the precursor of a most violent and terrific hurricane. ) 
 
 7. A thick, hazy appearance in that quarter of the horizon in 
 which the cyclone is raging. 
 
 8. A remarkable and peculiar appearance of the heavenly bodies. 
 When shining through a haze, they are said to shine with a pale 
 sickly light, and are not unfrequently surrounded by rings of light, 
 or halos. Some observers describe the stars as ' looking big with 
 burrs about them.' Others speak of their dancing, and generally, 
 they have been noticed as being remarkably bright and twinkling. 
 
PRACTICAL RULES BASED THEREON. 63 
 
 9. The sun on some occasions has exhibited a blue appearance, 
 and white objects have been seen of a decided light-blue colour. 
 The sun has also been observed of a pale and somewhat similar 
 appearance to that of the full moon. 
 
 10. A dense heavy bank of cloud in the direction of the hurri- 
 cane. 
 
 11. A peculiar appalling appearance in this bank, more particu- 
 larly as if it were a solid wall drawing down upon and closing 
 around the ship. 
 
 1 2. A darting forward of portions of this bank, as if torn into 
 rags and shreds by some violent force, and driven before, not borne 
 by, the wind. (When this indication is distinctly recognised, a 
 run of about two hours towards the centre will involve the vessel in 
 a destructive hurricane.) 
 
 13. A peculiar motion exhibited by small bodies, as branches of 
 trees, when agitated by the wind, consisting of a sort of whirling, 
 not a bending forward, as if bent by a stream of air. 
 
 14. Lightning of a remarkably columnar character, shooting up 
 in stalks from the horizon with a dull glare ; also like flashes from 
 a gun, and sparks from a flint and steel. 
 
 " Meteorological signs recognised by the ear 
 
 15. A distant roar (probably of the hurricane itself), as of wind 
 rushing through a hollow vault. 
 
 16. A peculiar moaning of the wind, indicative of the close 
 proximity of the violent portion of the hurricane. 
 
 "The meteorological phenomena accompanying a revolving storm 
 may also be enumerated thus : 
 
 1. A very rapid motion of the air, constituting the hurricane, 
 and increasing in velocity as the centre is approached. 
 
 2. A fitful variation of intensity, in the force of the wind, which 
 sometimes blows with fearful violence, carrying away everything 
 that opposes its progress, then sinking to a gentle breeze, or even 
 lulling to a calm, but almost immediately afterwards springing up 
 with greater violence than before. The hurricane winds are nearly, 
 if not entirely, without exception puny, violent, and blowing in 
 gusts. 
 
 3. An immense condensation of aqueous vapour, forming large 
 banks of cloud, which precipitate torrents of rain. The condensa- 
 tion appears to be so exceedingly rapid that large quantities of 
 
64 THE CIRCULAR THEORY OF STORMS, AND 
 
 electricity are generally developed, giving rise to incessant flashes 
 of lightning. 
 
 4. A general darkness and gloominess within the area of the 
 cyclone, relieved only by the fitful glare of the lightning, or the 
 appearance of the imperfect circle of light near the centre or axis 
 of the storm. 
 
 5. A separation of the clouds in or near the centre of the hurri- 
 cane, so as to produce in the immediate neighbourhood of the axis 
 a clear sky, through which the sun and stars are often seen with 
 great brilliancy. 
 
 6. A calm in the centre of the cyclone. 
 
 " Indications of approaching or existing hurricanes manifested by 
 the ocean, of especial utility to vessels at anchor in roadsteads, &c. : 
 
 1. A swell produced by the storm-wave rolling in upon the 
 shore, at first of a gentle character. The direction of this swell 
 will pretty surely indicate the bearing of the storm, and its changes 
 will point out, in some localities, the course the hurricane may be 
 pursuing. 
 
 2. A swell rolling in, without cJianging its direction, may be 
 regarded as indicative of a hurricane approaching the shore. The 
 same phenomenon met with at sea (the ship's course being taken 
 into account) will indicate the bearing down of the cyclone on the 
 vessel. 
 
 3. A dirty green appearance of the ocean ; on some occasions 
 its assuming a muddy or brown colour, on others its being remark- 
 ably clear; its temperature increasing, and its smelling stronger 
 than at other times, are all indications of the proximity of a 
 cyclone." 
 
 The Storm Wave. The rotation of the wind in a hurricane violently 
 agitates the surface of the ocean in the direction of the wind, producing 
 a swell or STORM WAVE, the undulations of which roll to an extraordi- 
 nary distance, behind, before, and on each side of, the storm. The wind, 
 however, is incessantly changing ; and thence there results a series of 
 undulations rolling from the margin of the storm, both in advance and 
 regression of the storm itself, and the succession of undulations en- 
 countering each other produces, in the area of intersection, cross seas 
 more or less dangerous, according as they are met in advance of, or 
 behind, the hurricane. It is easy to see that such a series of undulations 
 must always fringe the storm's wake, and be found to the right and left 
 of the path over which the cyclone has advanced ; while the long rolling 
 
PRACTICAL RULES BASED THEREON. 65 
 
 swell generally precedes it, and thus indicates the approach of the gale 
 many hours if not a day before it is encountered by the ship. 
 
 In the NORTHERN HEMISPHERE it is on the left side of the storm's 
 wake, in the SOUTHERN HEMISPHERE on the right side of the storm's 
 path, that the cross pyramidal sea is found ; and when a ship meets with 
 such a sea, it may be taken as a tolerably certain indication that a rota- 
 tory gale has passed over the locality. Generally the storm-waves un- 
 dulate in a direction which, the further they are from the vortex, is more 
 at right-angles with the direction of the wind ; in a heavy gale they 
 merely proceed in the direction of the gale itself. 
 
 The area that rotatory storms have been estimated to cover varies 
 from 30 or 40 to 1000 miles; but while the diameter, so long as the 
 storm is within the tropics, expands but very gradually, it suddenly 
 increases in dimensions with remarkable rapidity after recurving. The 
 rate at which they travel on their onward course also varies greatly, not 
 only in different parts of the globe, but even in the same locality, and 
 at the same season. Often, however, the velocity with which the vortex 
 of the hurricane progresses is greater as the storm recurves on approach- 
 ing the outer edge of the trade-winds, but this is not always the case. 
 
 The velocity of translation according to REID varies from 3 to 43 miles 
 an hour; according to PIDDINGTON the mean rate is from 16 to 20 
 miles an hour : the estimate of REDFIELD varies from 9J to 43 miles 
 for the hurricanes of the West Indies and those of the American coast. 
 THOM, in relation to the cyclones of the Southern Indian Ocean, gives a 
 rate of from 2 to 10 miles an hour ; but BRIDET, with some attempt at 
 precision, estimates that between 
 
 lat. 5 to 10 S. the progressive velocity varies from 1 to 5 miles, 
 15 to 25 S. 5 to 10 
 
 and in the higher latitudes, from 12 to 18 miles per hour. 
 
 According to REDFIELD, the hurricane of August 1853 traversed 
 7276 English miles in about 12 days, with a mean progressive velocity 
 of 26 miles per hour ; but after it had arrived at the Banks of 
 Newfoundland, this velocity was increased to about 50 miles per hour. 
 
 The hurricane of 1873 that Capt. TOYNBEE tracked for 17 days in 
 the Atlantic, and which dispersed off Newfoundland, had a velocity of 
 translation of from 7 J to 32 miles per hour ; but its least rate was not, 
 as BRIDET supposes, in the lower latitudes, but near the locality of re- 
 curvature. 
 
 In order to determine the rate of progress of areas of low pressure, 
 Professor E. LOOMIS has examined the weather reports of the United 
 
 E 
 
66 THE CIECULAR THEORY OF STORMS, AND 
 
 States Signal Service, during a period of 13 years, and these show that 
 the average velocity (per hour) of the progress of storms for the entire 
 year is 2 8 '4 miles ; also that the velocity is greatest in February and least 
 in August, and that the former velocity is 50 per cent, greater than the 
 latter. The velocity also varies very much for the same month in 
 different years, the greatest mean velocity for several of the months being 
 nearly double the least mean velocity for the same months. 
 
 The rate of progress of storm-centres in Europe the Professor gives 
 as 16 '7 miles an hour, the average velocity over the United States being 
 about two-thirds greater. 
 
 Professor LOOMIS also makes a similar examination for the Atlantic 
 Ocean, from which he obtains 18 miles per hour as the average velocity 
 for the entire year. This average is, however, probably too low ; for as 
 regards those storms that have been recorded as lying in the track of 
 vessels bound from Europe to America, it appears that from 20 to 25 
 miles an hour is a safer estimate of the rate. But there is undoubtedly 
 often a very wide margin of difference for individual storms. 
 
 Within the area of the storm the wind has been estimated to attain, 
 on frequent occasions, a rotatory velocity of from 70 to 100 miles an 
 hour, and even more within the tropics. BRIDET gives a velocity of 
 upwards of 125 miles an hour for one of the Reunion cyclones. 
 
 DOVE'S rules for trimming a vessel to the hurricane winds, in the two 
 hemispheres, are embodied in Tables I. and II. on p. 67. The directions 
 of the wind- vane as set down in the first column, are tangents to the 
 whirlwind in its course ; these indicate the quarter from which the storm 
 sets in, or is encountered. The points of the compass in the second 
 column show (on the circular theory) the position of the storm's centre 
 as regards the vessel. The headings of the remaining columns, and the 
 instructions within the columns, sufficiently explain the various methods 
 of manoeuvring according to the shift of wind ; this shift is, however, 
 supposed to be from a stationary point. 
 
 On pp. 68-69 are given taken from the French various methods 
 of manoeuvring ; in these tables, N = navigable semicircle, D = dangerous 
 semicircle, P = port tack, S = starboard tack. 
 
PRACTICAL RULES BASED THEREON. 67 
 
 I. FOR THE NORTHERN HEMISPHERE (DOVE) : 
 
 Direction of 
 Wind at com- 
 mencement 
 of Storin. 
 
 And the centre 
 (or vortex) 
 bearing. 
 
 If the Wind 
 shifts from 
 
 Steer 
 
 
 But if the Wind 
 shifts from 
 
 
 N.W. 
 
 N.E. 
 
 N.W. towards W. 
 
 S.E. 
 
 
 N.W. towards N. 
 
 
 N.W. byN. 
 
 N.E. byE. 
 
 N.W. byN.,, W. 
 
 S.E. by S. 
 
 
 N.W. byN. ,, N. 
 
 
 N.N.W. 
 
 E.N.E. 
 
 N.N.W. ,, W. 
 
 S.S.E. 
 
 
 N.N.W. , N. 
 
 
 N. by W. 
 North 
 
 E. by N. 
 
 East 
 
 N. by W. W. 
 North ,, W. 
 
 S. by E. 
 South 
 
 P 
 
 
 N. by W. , N. 
 North , E. 
 
 
 N. by E. 
 
 E. by S. 
 
 N. by E. N. 
 
 S. by W. 
 
 & 
 
 N. by E. , E. 
 
 43 
 
 N.N.E. 
 
 E.S.E. 
 
 N.N.E. N. 
 
 S.S.W. 
 
 E 
 
 N.N.E. , E. 
 
 1 
 
 N.E. byN. 
 
 S.E. by E. 
 
 N.E. byN. N. 
 
 S.W. byS. 
 
 
 
 N.E. byN. , E. 
 
 EH 
 
 N.E. 
 
 S.E 
 
 N.E. N. 
 
 S.W. 
 
 OJ 
 
 N.E. , E. 
 
 1 
 
 N.E. byE. 
 E.N.E. 
 
 S.E. by S. 
 S.S.E. 
 
 N.E. byE. N. 
 E.N.E. ,, N. 
 
 S.W. by W. 
 W.S.W. 
 
 +3 
 
 g 
 
 N.E. byE. , E. 
 E.N.E. , E. 
 
 ! 
 
 E. by N. 
 
 S. by E. 
 
 E. by N. ,, N. 
 
 W. by S. 
 
 o 
 
 E. by N. , E. 
 
 1 
 
 East 
 
 South 
 
 East N. 
 
 West 
 
 
 
 East , S. 
 
 0> 
 
 E. bv S. 
 
 S. by W. 
 
 E. by S. ,, E. 
 
 W. by N. 
 
 i 
 > 
 
 E. by S. , S. 
 
 ^ 
 
 E.S.E. 
 
 S.S.W. 
 
 E.S.E. ,, E. 
 
 W.N.W. 
 
 1 
 
 E.S.E. , S. 
 
 a 
 
 S.E. byE. 
 
 S.W. byS. 
 
 S.E. by E. E. 
 
 N.W. by W. 
 
 pM 
 
 <p 
 
 S.E. byE. , S. 
 
 o 
 
 S.E. ' 
 
 S.W. 
 
 S.E. ' E. 
 
 N.W. ' 
 
 1 
 
 S.E. ' , S. 
 
 S 
 
 S.E. byS. 
 
 S.W. by W. 
 
 S.E. byS. E. 
 
 N.W. byN. 
 
 
 S.E. by S. , S. 
 
 
 
 S.S.E. 
 
 W.S.W. 
 
 S.S.E. ,, E. 
 
 N.N.W. 
 
 o> 
 
 r r* 
 
 S.S.E. , S. 
 
 
 
 S. by E. 
 
 W. by S. 
 
 S. by E. ,, E. 
 
 N. bv W. 
 
 O 
 
 S. by E. , S. 
 
 W 
 
 South 
 
 West 
 
 South ,, E. 
 
 'North 
 
 $4" 
 
 South , W. 
 
 
 S. by W. 
 S.S.W. 
 
 W. by N. 
 W.N.W. 
 
 S. by W. ,, S. 
 
 S.S.W. S. 
 
 N. by E. 
 
 N.N.E. 
 
 O 
 
 S. by W. , W. 
 S.S.W. , W. 
 
 
 S.W. by S. 
 
 N.W. by W. 
 
 S.W. by S. ,, S. 
 
 N.E. byN. 
 
 
 S.W. byS. , W. 
 
 
 S.W. ' 
 
 N.W. 
 
 S.W. ' S. 
 
 N.E. ' 
 
 
 S.W. ' , W. 
 
 
 II. FOR THE SOUTHERN HEMISPHERE (DOVE) : 
 
 Direction of 
 Wind at com- 
 mencement 
 of Storm. 
 
 And the centre 
 (or vortex) 
 bearing. 
 
 If the Wind 
 shifts from 
 
 Steer 
 
 
 But if the Wind 
 shifts from 
 
 
 South 
 
 East 
 
 South towards W. 
 
 North 
 
 ^4 
 
 South to wards E. 
 
 
 S. by E. 
 S.S.E. . 
 
 E. by N. 
 E.N.E. 
 
 S. bvE. ,, S. 
 S.S.E. S. 
 
 N. by W. 
 
 N.N.W. 
 
 1 
 
 SbyE. E. 
 S.S.E ,, E. 
 
 
 S.E. by S. 
 
 N.E. by E. 
 
 S.E. byS. S. 
 
 N.W. byN. 
 
 Is 
 
 S.E. by S. E. 
 
 
 S.E. ' 
 
 N.E. ' 
 
 S.E. ' S. 
 
 N.W. 
 
 o 
 
 S.E. " E. 
 
 
 
 S.E. bv E. 
 
 N.E. bv N. 
 
 S.E. byE. ,, S. 
 
 N.W. by W. 
 
 S 
 
 S.E. by E. ,, E. 
 
 1 
 
 E.S.E." 
 
 N.N.E!" 
 
 E.S.E. S. 
 
 W.N.W. 
 
 -2 
 
 t/2 
 
 E.S.E. ,, E. 
 
 S 
 
 E. by S. 
 
 N. by E. 
 
 E. byS. ,, S. 
 
 W. by N. 
 
 0> 
 
 E. byS. ,, E. 
 
 
 
 East 
 
 North 
 
 East ,, S. 
 
 West 
 
 ,j 
 -M 
 
 East ,, N. 
 
 & 
 
 E. by N. 
 
 E.N.E. 
 
 N. by W. 
 
 N.N.W. 
 
 E. by N. E. 
 E.N.E. , E. 
 
 W. by S. 
 W.S.W. 
 
 d 
 o 
 
 s 
 
 E. byN. ,, N. 
 E.N.E. ,, N. 
 
 O 
 
 J3 
 
 N.E. byE. 
 
 N.W. by N. 
 
 N.E. byE. , E. 
 
 S.W. by W. 
 
 S 
 
 N.E. byE. ,, N. 
 
 d 
 o 
 
 N.E. ' 
 
 N.W. 
 
 N.E. " , E. 
 
 S.W. 
 
 t 
 
 N.E. " N. 
 
 
 N.E. by N. 
 
 N.W. by W. 
 
 N.E. byN. , E. 
 
 S.W. byS. 
 
 cS 
 o> 
 
 N.E. byN. ,, N. 
 
 o 
 
 -t-> 
 
 N.N.E. 
 
 W.N.W. 
 
 N.N.E. , E. 
 
 S.S.W. 
 
 ^ 
 
 N.N.E. ,, N. 
 
 1 
 
 N. by E. 
 North 
 
 W. by N. 
 
 West 
 
 N. by E. , E. 
 North , E. 
 
 S. by W. 
 South 
 
 oT 
 
 03 
 
 '$ 
 
 N. by E. N. 
 North ,, W. 
 
 S 
 
 N. by W. 
 
 N.N.W. 
 
 W. by S. 
 W.S.W. 
 
 N. by W. , N. 
 
 N.N.W. ,, -N. 
 
 S. bv E. 
 S.S.E. 
 
 b 
 
 0) 
 
 N. by W. W. 
 N.N.W. ,, W. 
 
 
 N.W. byN. 
 
 S.W. by W. 
 
 N.W. byN.,, N. 
 
 S.E. byS. 
 
 o 
 
 N.W. byN.,, W. 
 
 
 N.W. ' 
 
 S.W. 
 
 N.W. ,, N. 
 
 S.E. 
 
 6 
 
 N.W. W. 
 
 
68 THE CIRCULAR THEORY OF STORMS, AND 
 
 III. FOR NORTH HEMISPHERE. (From the French.) 
 
 Direction 
 
 Bearing 
 
 Cyclone moving N W-ly. 
 
 Cyc 
 
 ^~ 
 
 3$ 
 
 ^ J 
 
 lone n 
 
 3 
 
 ^3 
 
 loving N-ly. 
 Course (if 
 
 Cyclone moving N E-ly. 
 
 3 J 
 
 3 
 
 Course (if 
 
 II 
 
 3 
 
 , 
 
 Course (if 
 
 of 
 Wind. 
 
 of 
 
 Centre. 
 
 S & 
 || 
 
 : on whid 
 leave-to. 
 
 possible) to 
 run out of 
 
 Ss 
 
 ^ 
 
 || 
 
 possible) to 
 run out of 
 
 & 
 
 || 
 
 | 
 
 possible) to 
 run out of 
 
 
 
 53 rC 
 
 GO GO 
 
 f* m 
 
 I 
 
 cyclone. 
 
 It 
 
 f 
 
 cyclone. 
 
 g e< 
 
 s .H 
 
 0002 
 
 f 
 
 cyclone. 
 
 N 
 
 E 
 
 N 
 
 P 
 
 S W 
 
 N 
 
 P 
 
 S W 
 
 N 
 
 P 
 
 S W 
 
 NbyE 
 
 EbyS 
 
 ... 
 
 
 S W by W 
 
 
 
 SWbyW 
 
 ... 
 
 
 S W 
 
 NNE 
 
 ESE 
 
 ... 
 
 ... 
 
 WSW 
 
 ... 
 
 ... 
 
 WSW 
 
 ... 
 
 ... 
 
 S W 
 
 NEbyN 
 
 S E by E 
 
 
 ... 
 
 WbyS 
 
 ... 
 
 ... 
 
 WbyS 
 
 ... 
 
 
 S W by W 
 
 NE 
 
 SE 
 
 D 
 
 S 
 
 W 
 
 
 ... 
 
 W 
 
 ... 
 
 ... 
 
 WSW 
 
 NEbyE 
 
 SEbyS 
 
 ... 
 
 
 WbyN 
 
 ... 
 
 ... 
 
 WbyN 
 
 ... 
 
 
 WbyS 
 
 ENE 
 
 SSE 
 
 ... 
 
 
 WNW 
 
 ... 
 
 ... 
 
 WNW 
 
 ... 
 
 ... 
 
 W 
 
 E byN 
 
 SbyE 
 
 
 ... 
 
 WNW 
 
 ... 
 
 ... 
 
 NWbyW 
 
 
 ... 
 
 WbyN 
 
 E 
 
 S 
 
 ... 
 
 
 WNW 
 
 D 
 
 S 
 
 N W 
 
 ... 
 
 ... 
 
 N W 
 
 EbyS 
 
 Sby W 
 
 ... 
 
 ... 
 
 NWbyN 
 
 ... 
 
 ... 
 
 NWbyN 
 
 ... 
 
 ... 
 
 NWbyN 
 
 ESE 
 
 SSW 
 
 ... 
 
 ... 
 
 NNW 
 
 ... 
 
 ... 
 
 NNW 
 
 ... 
 
 ... 
 
 NNW 
 
 SEbyE 
 
 SWbyS 
 
 ... 
 
 ... 
 
 Nby W 
 
 ... 
 
 ... 
 
 NNW 
 
 ... 
 
 ... 
 
 Nby W 
 
 SE 
 
 SW 
 
 ... 
 
 ... 
 
 N 
 
 ... 
 
 
 NNW 
 
 D 
 
 S 
 
 N 
 
 SEbyS 
 
 SWbyW 
 
 ... 
 
 ... 
 
 NbyE 
 
 ... 
 
 ... 
 
 NbyE 
 
 ... 
 
 ... 
 
 NbyE 
 
 SSE 
 
 WSW 
 
 ... 
 
 
 NNE 
 
 ... 
 
 ... 
 
 NNE 
 
 ... 
 
 ... 
 
 NNE 
 
 SbyE 
 
 WbyS 
 
 ... 
 
 ... 
 
 NEbyN 
 
 ... 
 
 ... 
 
 NEbyN 
 
 ... 
 
 ... 
 
 NNE 
 
 S 
 
 W 
 
 
 ... 
 
 NE 
 
 ... 
 
 ... 
 
 NE 
 
 ... 
 
 ... 
 
 NNE 
 
 Sby W 
 
 WbyN 
 
 ... 
 
 ... 
 
 NEbyE 
 
 ... 
 
 ... 
 
 NEbyE 
 
 ... 
 
 ... 
 
 NEbyE 
 
 SSW 
 
 WNW 
 
 ... 
 
 ... 
 
 ENE 
 
 ... 
 
 ... 
 
 ENE 
 
 ... 
 
 ... 
 
 ENE 
 
 S W by S 
 
 NWbyW 
 
 ... 
 
 ... 
 
 EbyN 
 
 ;.. 
 
 ... 
 
 EbyN 
 
 ... 
 
 ... 
 
 EbyN 
 
 SW 
 
 N W 
 
 ... 
 
 ... 
 
 E 
 
 ... 
 
 ... 
 
 E 
 
 ... 
 
 ... 
 
 E 
 
 SWbyW 
 
 NWbyN 
 
 N 
 
 p 
 
 EbyS 
 
 ... 
 
 ... 
 
 EbyS 
 
 ... 
 
 ... 
 
 EbyS 
 
 WSW 
 
 NNW 
 
 
 ... 
 
 ESE 
 
 ... 
 
 ... 
 
 ESE 
 
 ... 
 
 ... 
 
 ESE 
 
 WbyS 
 
 Nby W 
 
 ... 
 
 ... 
 
 SEbyE 
 
 ... 
 
 ... 
 
 SEbyE 
 
 ... 
 
 
 SEbyE 
 
 W 
 
 N 
 
 ... 
 
 ... 
 
 SE 
 
 ... 
 
 ... 
 
 SE 
 
 ... 
 
 ... 
 
 SE 
 
 WbyN 
 
 NbyE 
 
 ... 
 
 ... 
 
 SEbyS 
 
 N 
 
 P 
 
 SEbyS 
 
 ... 
 
 ... 
 
 SEbyS 
 
 WNW 
 
 NNE 
 
 ... 
 
 
 SSE 
 
 ... 
 
 ... 
 
 SSE 
 
 ... 
 
 ... 
 
 SSE 
 
 NWbyW 
 
 NEbyN 
 
 ... 
 
 ... 
 
 SbyE 
 
 ... 
 
 ... 
 
 SbyE 
 
 ... 
 
 ... 
 
 SbyE 
 
 N W 
 
 NE 
 
 ... 
 
 ... 
 
 S 
 
 ... 
 
 ... 
 
 S 
 
 ... 
 
 ... 
 
 S 
 
 NWbyN 
 
 NEbyE 
 
 ... 
 
 ... 
 
 Sby W 
 
 ... 
 
 ... 
 
 Sby W 
 
 N 
 
 p 
 
 Sby W 
 
 NNW 
 
 ENE 
 
 ... 
 
 ... 
 
 SSW 
 
 ... 
 
 
 SSW 
 
 ... 
 
 ... 
 
 SSW 
 
 Nby W 
 
 EbyN 
 
 ... 
 
 ... 
 
 SWbyS 
 
 ... 
 
 
 SWbyS 
 
 ... 
 
 
 SWbyS 
 
PRACTICAL RULES BASED THEREON. 69 
 
 IY. FOR SOUTH HEMISPHERE. (From the French.) 
 
 Direction 
 of 
 Wind. 
 
 Bearing 
 of 
 Centre. 
 
 Cyclone moving S W-ly. 
 
 Cyclone moving S-ly. 
 
 Cyclone moving S E-ly. 
 
 Semicircle in which 1 
 Ship takes cyclone. 
 
 Tack on which to 
 heave-to. 
 
 Course (if 
 possible) to 
 run out of 
 cyclone. 
 
 Semicircle in which ] 
 Ship takes cyclone. 
 
 Tack on which to 
 heave-to. 
 
 Course (if 
 possible) to 
 run out of 
 cyclone. 
 
 14 
 
 II 
 
 
 
 II 
 
 "o "** 
 
 "r* PH 
 
 Tack on which to 
 heave-to. 
 
 Course (if 
 possible) to 
 run out of 
 cyclone. 
 
 N 
 
 W 
 
 D 
 
 P 
 
 SE 
 
 D 
 
 P 
 
 SE 
 
 D 
 
 P 
 
 SSE 
 
 NbyE 
 
 WbyN 
 
 ... 
 
 ... 
 
 SEbyS 
 
 ... 
 
 ... 
 
 SEbyS 
 
 ... 
 
 ... 
 
 SSE 
 
 NNE 
 
 WNW 
 
 
 .. . 
 
 SSE 
 
 
 
 SSE 
 
 
 ... 
 
 SSE 
 
 NEbyN 
 
 NWbyW 
 
 
 ... 
 
 SbyE 
 
 ... 
 
 ... 
 
 SbyE 
 
 ... 
 
 ... 
 
 SbyE 
 
 NE 
 
 N W 
 
 ... 
 
 
 S 
 
 
 
 SSW 
 
 ... 
 
 ... 
 
 S 
 
 NEbyE 
 
 NWbyN 
 
 
 ... 
 
 SbyW 
 
 ... 
 
 ... 
 
 ssw 
 
 N 
 
 S 
 
 SbyW 
 
 ENE 
 
 NNW 
 
 ... 
 
 ... 
 
 SSW 
 
 
 . . . 
 
 SSW 
 
 ... 
 
 ... 
 
 SSW 
 
 EbyN 
 
 NbyW 
 
 ... 
 
 ... 
 
 S W byS 
 
 ... 
 
 ... 
 
 S WbyS 
 
 ... 
 
 ... 
 
 SWbyS 
 
 E 
 
 N 
 
 ... 
 
 ... 
 
 WSW 
 
 
 ... 
 
 S W 
 
 ... 
 
 ... 
 
 S W 
 
 EbyS 
 
 NbyE 
 
 ... 
 
 ... 
 
 WSW 
 
 N 
 
 S 
 
 SWbyW 
 
 ... 
 
 ... 
 
 SWbyW 
 
 ESE 
 
 NNE 
 
 ... 
 
 
 WSW 
 
 
 ... 
 
 WSW 
 
 ... 
 
 ... 
 
 WSW 
 
 SEby E 
 
 NEbyN 
 
 ... 
 
 ... 
 
 WbyS 
 
 ... 
 
 ... 
 
 WbyS 
 
 
 ... 
 
 WbyS 
 
 SE 
 
 NE 
 
 ... 
 
 ... 
 
 W 
 
 ... 
 
 
 W 
 
 .. . 
 
 .. . 
 
 W 
 
 SEbyS 
 
 NEbyE 
 
 N 
 
 S 
 
 WbyN 
 
 ... 
 
 ... 
 
 WbyN 
 
 ... 
 
 ... 
 
 WbyN 
 
 SSE 
 
 ENE 
 
 
 .. . 
 
 WNW 
 
 ... 
 
 ... 
 
 WNW 
 
 
 ... 
 
 WNW 
 
 SbyE 
 
 EbyN 
 
 ... 
 
 
 NWbyW 
 
 
 
 NWbyW 
 
 ... 
 
 ... 
 
 NWbyW 
 
 S 
 
 E 
 
 ... 
 
 .. . 
 
 NW 
 
 ... 
 
 
 N W 
 
 
 ... 
 
 NW 
 
 Sby W 
 
 EbyS 
 
 
 ... 
 
 NWbyN 
 
 ... 
 
 ... 
 
 NWbyN 
 
 ... 
 
 ... 
 
 NWbyN 
 
 ssw 
 
 ESE 
 
 ... 
 
 ... 
 
 NNW 
 
 ... 
 
 .. . 
 
 NNW 
 
 ... 
 
 
 NNW 
 
 S Wby S 
 
 SEbyE 
 
 ... 
 
 ... 
 
 NbyW 
 
 ... 
 
 ... 
 
 NbyW 
 
 ... 
 
 ... 
 
 NbyW 
 
 S W 
 
 SE 
 
 ... 
 
 ... 
 
 N 
 
 ... 
 
 ... 
 
 N 
 
 D 
 
 P 
 
 N 
 
 SWbyW 
 
 SEbyS 
 
 ... 
 
 
 NbyE 
 
 ... 
 
 ... 
 
 NbyE 
 
 
 
 NbyE 
 
 WSW 
 
 SSE 
 
 ... 
 
 ... 
 
 NNE 
 
 ... 
 
 . . . 
 
 NNE 
 
 .. . 
 
 .. . 
 
 NNE 
 
 WbyS 
 
 SbyE 
 
 ... 
 
 ... 
 
 NEbyN 
 
 ... 
 
 ... 
 
 NEbyN 
 
 ... 
 
 ... 
 
 NEbyN 
 
 W 
 
 S 
 
 ... 
 
 ... 
 
 NE 
 
 D 
 
 P 
 
 NE 
 
 ... 
 
 ... 
 
 NE 
 
 WbyN 
 
 SbyW 
 
 ... 
 
 ... 
 
 NEbyE 
 
 ... 
 
 ... 
 
 NEbyE 
 
 ... 
 
 ... 
 
 NEbyE 
 
 WNW 
 
 SSW 
 
 ... 
 
 ... 
 
 ENE 
 
 ... 
 
 ... 
 
 ENE 
 
 ... 
 
 ... 
 
 ENE 
 
 NWbyW 
 
 SWbyS 
 
 ... 
 
 
 EbyN 
 
 
 ... 
 
 EbyN 
 
 ... 
 
 ... 
 
 EbyN 
 
 N W 
 
 S W 
 
 D 
 
 P 
 
 E 
 
 ... 
 
 ... 
 
 E 
 
 ... 
 
 ... 
 
 E 
 
 NWbyN 
 
 SWbyW 
 
 
 ... 
 
 EbyS 
 
 ... 
 
 ... 
 
 EbyS 
 
 ... 
 
 ... 
 
 EbyS 
 
 NNW 
 
 WSW 
 
 ... 
 
 ... 
 
 ESE 
 
 ... 
 
 ... 
 
 ESE 
 
 ... 
 
 ... 
 
 ESE 
 
 NbyW 
 
 WbyS 
 
 ... 
 
 ... 
 
 SEbyE 
 
 
 ... 
 
 SEbyE 
 
 ... 
 
 
 SEbyE 
 
SPIRAL ROTATION OP THE WIND. 
 
 MODIFICATION OF THE CIRCULAR THEORY OF STORMS : 
 RESULTS OF RECENT INVESTIGATION. 
 
 The controversies of meteorologists in the early part of our century 
 eventuated in the triumph of the circular theory of storms, and the 
 establishment of laws based on that theory. The chiefs in the investiga- 
 tion REDFIELD, REID, and PIDDINGTON had their doubts about the 
 exactitude of a strictly circular revolution of the winds, and were 
 earnest in the expression of their doubts on this head ; they, however, 
 nevertheless thought the eight-point rule for the bearing of the 
 centre of the storm, and the tack on which to heave-to, according to 
 the semicircle in which the ship was found to be, were sufficiently, 
 although only approximately, accurate for safety, provided always that 
 the barometer was carefully watched, together with the veering or the 
 backing of the winds. It came to pass, however, that writers and 
 many of them in high authority who ostensibly followed the teachings 
 of the men who first discovered and popularized the " Law of Storms," 
 insidiously, and without sufficient justification, persisted in advocating 
 the strictly circular rotation of the winds in the tropical hurricane ; and, 
 for at least twenty-five years after the first promulgation of the " Law," 
 it was deemed a heresy to dispute the validity of the crude reasoning 
 brought forward to substantiate the purely circular theory, though to 
 some, at least, that theory appeared erroneous from the standpoint of 
 mechanics and physics. One of the earliest writers that perceived and 
 recognised the discrepancy between the theory and the facts, published 
 a work on the subject in 1853 : " A Memoir on the Equinoctial Storms 
 of March- April 1850 : an Inquiry into the Extent to which the 
 Rotatory Theory may be applied." By F. P. B. Martin, Esq., M.A. 
 (published by Thomas Harrison, Pall Mall). 
 
 The work was altogether desultory somewhat crude and in form 
 not very readable ; but the substance of the book may be summarized 
 thus : After stating it to be "a settled conviction, that, subject to the 
 varying modifications of local influences, the Law of Storms is universal, 
 semper, ubique, et db omnibus" he proceeds, in a note, to remark, " that 
 although these gales consist of a storm's eye, a central calm, .... 
 the wind blowing round that centre is not of the same force at every 
 portion of the circumference equally distant from that centre. More than 
 one instance will occur in these pages, of steamers having crossed the track 
 of very violent hurricanes without encountering any bad weather, or even, 
 but for the barometrical depression, having had reason to suppose it to 
 be near, Neither does the term mean that a storm is circular in the strict 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 71 
 
 mathematical acceptation of the word, as though the central storm's eye 
 were equally distant from every part of the circumference at which its 
 force is the same, inasmuch as the area of the storm, whilst advancing, 
 expands, but unequally, and this inequality arises from the diminished 
 pressure of the atmosphere in the rear quadrants of the storm : as the 
 vortex moves forward, it draws within its influence the atmosphere in 
 advance of it with a steady equal operation ; but as that portion in the 
 rear of it does not collapse and subside with the same regularity with 
 which it was set in motion, a current parallel to the axis of advance 
 will frequently be found to blow on the left-hand semicircle for some 
 time after the body of the storm has passed." 
 
 The work scarcely attracted the attention it merited, and for the 
 reasons already given. Many of the storms, however, of which 
 particulars are given, coincide in character with those described by later 
 observers, in India, China, the West Indies, and Mauritius. 
 
 We come now to a later period of the controversy as to the strict 
 validity of the circular theory. 
 
 For years after the establishment of the Mauritius Meteorological 
 Society, the members were constantly discussing the improbabilities of 
 many of the " well-established laws," and the errors of some of the 
 " practical rules : " the subject, though ever recurring, was always fresh, 
 for every season brought disabled ships to Port Louis to refit j all sorts 
 of forms for the cyclone, novel modes of progression, and new rules for 
 avoiding the fury of the tempest were constantly before the members, 
 and their value debated with considerable vigour : the perception of a 
 theory differing from the circular one had dawned on them years before 
 much notice was taken of their " Proceedings " in Europe, and in 1861 
 the controversy was at its height. Mr. C. MELDRUM, the secretary, had 
 (during ten or a dozen years) often called attention to the apparent in- 
 curving of the wind in cyclones, and to the losses occasioned by masters 
 of ships acting on the supposition that the bearing of the centre of the 
 storm was at right angles to the direction of the wind, and that the 
 movement of the wind was in concentric circles : he had come to the 
 conclusion that there was a converging of the wind in cyclones, when 
 the late Capt. DOUGLAS WALES, then harbour-master at Port Louis, 
 addressed the following important letter to him in 1872 : 
 
 " Some remarks of yours respecting the uncertainty of the real 
 position of the centre of a cyclone set me thinking, and I send you a 
 few ideas on the subject, which, as a sailor, I think worthy the serious 
 attention of seamen, and the correctness of which they may put to the 
 test of experience, whenever they have opportunities of doing so. 
 
 " Allow me to premise that I have no intention of dogmatizing. I 
 believe our knowledge of the causes of these fearful tempests, of their 
 
72 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 origin, their progress in this or that direction, their rate of progression, 
 their recurving, the reason of those recurvings, and their ultimate dis- 
 persion, to be still in its infancy. No doubt, the knowledge already 
 acquired has saved many a good ship from becoming entangled in these 
 storms, especially ships approaching them on their equatorial sides ; but 
 at the same time it must be admitted that more than one intelligent 
 seaman, who thought himself well up in the subject, has actually run 
 into the very centre of a cyclone, when, by all known rules, he ought to 
 have been certain of avoiding it. 
 
 " There must be some reason for such an error, and it is that reason 
 that I have been seeking for, and which, I trust, I have to some extent 
 discovered. I assume that within a diameter of 40, 50, 60, 70, or 80 
 miles a true circular storm of terrific violence must be found in every so- 
 called hurricane, and that, to a considerable distance outside and around 
 this central and circular storm, winds are to be found gradually decreasing 
 in force from 11, near the outer edge of the central storm, to 7 and 6, 
 at the outer edge of the bad weather, but which, instead of blowing in 
 ever-enlarging circles further and further out from one common centre, 
 are always converging to that centre, and on all sides gradually increasing 
 until, at a certain distance from the central calm, they acquire the force 
 of a hurricane (12), and thence inwards blow with great violence in 
 what, in all probability, is as nearly as may be a circle. 
 
 " It is these converging lines of wind that are, I think, likely to lead 
 men into error as to the position of the centre of the storm. 
 
 " I shall, to prevent confusion, confine myself to cyclones south of 
 the equator, every one acquainted with the cyclonic theory knowing that 
 the inverse of rules for the guidance of seamen in the southern hemi- 
 sphere will be the rules for their guidance in the northern hemisphere. 
 
 " Let us suppose that a ship bound to Europe arrives at a point in a 
 converging curve where the wind being N.E. with force 7, that is, double 
 reefs and jib, barometer falling, sky overcast, confused swell, and, in 
 short, every appearance of bad weather; Lat. 12 S., Long. 70 E. 
 What ought her Commander to do 1 ' Heave-to on the Port Tack,' says 
 one, ' and wait for the weather to clear.' ' Run to the S.W.,' says 
 another, * and make use of the storm.' Being a pushing fellow he makes 
 up his mind to run, and, truth to say, there are as many reasons for 
 approving that proceeding as for finding fault with it. 
 
 " If he succeeds in making use of the hurricane, he is considered a 
 smart fellow ; if he runs into it and is dismasted, or worse, ' rash,' 
 ' headstrong,' * ignorant,' etc.,* are the best terms he can look for : 
 
 * These expressions are mild as compared with some that were heaped on shipmasters who 
 (following, as they thought, "definite and unambiguous rules") had the misfortune to find their 
 way to the centre, instead of out, of the hurricane. 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 73 
 
 and yet he might as easily have been wrong in heaving-to as in 
 running. 
 
 " The wind being N.E., he infers that the centre bears N.W. (see 
 diagram 17, p. 53). He considers that the barometer and weather in- 
 dicate that he is on the S.E. edge of a cyclone the N.E. wind upon 
 which he is running forming a part of a circidar storm, and that 
 necessarily the centre is N.W. of him. Considering, further, that in 
 that Lat. and Long, the storm is probably travelling W.S.W., he thinks 
 that if he runs S.W. he will be diverging from it, and that, by making 
 use of the storm, he will get fine runs, perhaps for days to come. 
 
 " But if the N.E. wind be only converging towards the fearful storm 
 raging near the centre, that centre, in the first place, bears W. by 
 N. J N. instead of N.W., so that the vessel, by steering S.W., is not 
 diverging from the centre, as supposed, but is really drawing nearer to it. 
 In due time the weather gets worse from this very cause ; the wind veers 
 more to the eastward ; the barometer continues to fall ; and the Captain 
 begins to doubt whether the storm may not after all be progressing more 
 to the southward than he supposed whether indeed it may not, although 
 so far to the eastward, be actually recurving, and he naturally becomes 
 anxious, and uncertain what to do. If he decides on running at all 
 risks, he finds the wind still drawing at first more and more easterly, 
 and then more and more southerly, always increasing in fury, and the 
 sea becoming more and more heavy and tumultuous. But run he must 
 now, and he must run dead before it, and being on what I have supposed 
 a line of wind converging to a centre, he finishes by getting into the real 
 hurricane, and loss and disaster are imminent. He may, however, if 
 his ship be tight and staunch, and runs well, get round to the N.W. side 
 of the storm, and so get clear, probably with loss of spars and sails ; but 
 he has clearly run into what he was running to avoid, because he was 
 under the impression that the winds within the influence of a cyclone, 
 although far from its centre, blew in circles round that centre, the wind 
 everywhere clearly indicating the exact, or nearly exact, position of that 
 centre. 
 
 " These opinions I submit with very great diffidence for the considera- 
 tion of seamen and cyclonists. I am not going to attempt the setting 
 up of any dogmatic theory of mine own, but I am inclined to think that 
 this theory of converging winds will probably account for the manner in 
 which many vessels have become entangled in hurricanes when seeking 
 to avoid them according to cyclonic rules. Like all other theories on 
 this very important subject, it requires very careful consideration ; but 
 there can be no possible risk in deducing from it the rule that vessels 
 on approaching what the barometer, the state of the weather, and the 
 force of the wind, clearly indicate as the dangerous side of a cyclone, 
 
74 MODIFICATION OP THE CIRCULAR THEORY OF STORMS. 
 
 should, in seeking to avoid it, keep the wind quite four points on the 
 Port Quarter. 
 
 " With the wind thus free, a fast ship would run with great rapidity 
 through the water, and, unless the storm were advancing on her in a 
 direct line, would be always increasing her distance from its centre, and 
 getting into finer weather and, in any case, would have a very good 
 chance of running across its track, and thus avoiding it. 
 
 " Ships running into cyclones on their equatorial sides are to a very 
 great extent without excuse. There are, however, some exceptional 
 instances ; but they are very rare." 
 
 The Theory of Revolving Storms had passed into what was supposed 
 to be a well-defined Law, and the Practical Rules deduced therefrom 
 were taken as established on the very best basis that of experience : it 
 may, however, be said, without misstatement, that, from a critical point 
 of view, the circular theory had come to be accepted without much con- 
 troversy ; true, it had at first encountered opposition ; but from some 
 cause, whether from the influence of names (always very potent), or 
 from the apparent simplicity of the "Law" and its deductions, the 
 opposition had been silenced, or was in abeyance. Now no place better 
 than the Mauritius, owing to its situation, could have been selected for 
 a crucial examination at once of the theory and its practical bearing 
 on navigation ; and the letter of Capt. WALES is important, as showing 
 that the regular discussions of the Meteorological Society of the island 
 were (in 1872) beginning to develop valuable and unexpected results. 
 There were advocates of the circular theory, and others, who strongly 
 opposed it, and the discussions were at times very animated : the form 
 of cyclones whether circles or ellipses ; the calm was it in the centre 
 of a circle, or in one of the foci of an ellipse ? were the winds at a given 
 place in the storm from such direction as the theory required them to 
 be ] did this or that ship run into the vortex through disregarding or 
 obeying the practical rules? these had been the topics debated and 
 discussed at almost every meeting of the Society for ten or twelve years : 
 meanwhile the Secretary (Mr. C. MELDRUM) was quietly examining the 
 archives for new data ; the investigation resulted in his faith in the old 
 principles being gradually shaken; and in February 1873 he read to the 
 Society the following important paper : 
 
 Notes on the Form of Cyclones in the Southern Indian Ocean, and on 
 some of the Rules given for avoiding their Centres. 
 
 " According to the earlier writers on the ' Law of Storms,' cyclones 
 are of a circular form, and they are invariably represented by a number 
 of concentric circles, in which the wind in the Northern Hemisphere 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 75 
 
 blows from right to left, or against the hands of a watch, and in the 
 contrary direction in the Southern Hemisphere. 
 
 " Following in the footsteps of his predecessors, who had written 
 some thirty years previously, M. BRIDET, in his excellent treatise, Etude 
 sur les Ouragans (Paris, 1869), has adopted the same views. 
 
 " Now, if cyclones are circular, it is very easy to know the bearing 
 of the centre, and a very simple rule is laid down by cyclonologists for 
 the guidance of shipmasters in that respect. ' Turn your face to the 
 
 * wind/ they say, ' and if you are in the Northern Hemisphere, the 
 ' centre will be to your right, but if in the Southern Hemisphere, to 
 ' your left.' Or ' the bearing of the centre is always at right angles to 
 ' the direction of the wind, but in different directions in the two 
 ' hemispheres.' Thus, with the wind at East, the centre in the Northern 
 Hemisphere bears due South, and in the Southern Hemisphere due 
 North ; and so on. 
 
 " But it is evident that if cyclones are not circular, the above rule 
 is inapplicable. It is, therefore, of great practical importance to know 
 something of the form of cyclones. 
 
 " The first time attention was called to this particular point was, so 
 far as I can find, in January 1860. The winds and weather at the 
 Mauritius Observatory had just indicated the passage of a cyclone, and 
 I determined to deduce its track from my own observations alone, 
 before any vessels could arrive. Proceeding on the supposition that 
 the circular theory was correct, I was obliged to come to the conclusion 
 ' that the E.S.E. wind, which blew for forty-eight hours, did not belong 
 ' to the cyclone, unless the latter travelled at first S. by E. and S.S.E., 
 ' and then S. by W. and S.S.W., or in some other irregular way.' 
 
 " These remarks, which were published in the Commercial Gazette, 
 attracted the attention of Mr. CLARE BERNARD. 
 
 " ' At the end of last January,' remarks Mr. BERNARD in a paper 
 which he communicated to the Society, ' I saw with great interest, in 
 1 the Commercial Gazette, the remarkable paper read by Mr. MELDRUM 
 ' to the Meteorological Society on the cyclone of the 8th to 16th 
 ' January, and the perusal of it directed my ideas towards cyclonology, 
 
 * to which I had hitherto paid but little attention. 
 
 " < Soon afterwards I met with an extract from the Archives of 
 ' Mauritius, containing an account of the celebrated hurricane of 1818, 
 ' and I was anxious to trace the track which this cyclone had taken. I 
 ' did so according to the circular theory, following the method described 
 ' by Mr. MELDRUM in his paper, and was much surprised to find myself 
 ' confronted by an impossible result. For the wind in 1818 commenced 
 ' at E.S.E,, which placed the centre of the cyclone to the N.N.E. ; it 
 afterwards veered to South, thus placing the centre in the East ; it 
 
76 MODIFICATION OP THE CIRCULAR THEORY OF STORMS. 
 
 1 then veered to North and N.W., which made the centre pass to the 
 ' Westward and S.W.-ward of Mauritius. This cyclone, therefore, had 
 
 * travelled in a zig-zag direction.' 
 
 " Mr. BERNARD then proceeds to develop a theory by which, on the 
 supposition that cyclones are elliptical, that the centre is in the posterior 
 focus of the ellipse, and that the inclination of the major axis to the 
 meridian varies, he accounted for all anomalies. But it was evident 
 that, without further observation, this ingenious theory would not be 
 sufficient to overthrow the circular theory, which had held its ground for 
 nearly thirty years. (See Mercantile Marine Magazine, vol. 10, p. 321.) 
 
 " From January 1860 down to the present time, no hurricane has 
 been examined here which has not afforded fresh evidence that cyclones 
 are not so circular as they have been supposed to be. To reproduce all 
 the remarks that have been made on the subject would occupy too much 
 time and space, but it may be proper to remind the Society of some of 
 them. 
 
 " In a paper read on the 31st July 1860, giving an account of the 
 hurricane of the 6th to the 17th February 1860, it is said: 'The 
 
 * Southerly and Northerly winds extended many degrees South and 
 ' North of the centre, giving the storm an elliptical appearance ; and on 
 1 at least the llth, 12th and 13th, the trade, at some distance to the 
 ' Eastward, was directed towards the centre. Hence the bearing of the 
 
 * centre, from many of the vessels, was not at right angles to the wind's 
 1 direction.' 
 
 " The following quotation is from an account of the hurricanes of the 
 18th to the 29th February 1860 : ' There are no means of determining 
 ' the exact forms of the cyclones on each day ; but it is evident that the 
 
 * wind did not blow in true circles. From observations made on all sides 
 
 * of the Phoenix storm, on the 25th, it would appear that the meteor had 
 ' a helical form ; for, while the S.E. trade swept round on its Western 
 ' side, and the N.W. monsoon on its Eastern side, in two distinct curves, 
 ' there was an Easterly stream of air (between the N.E.-ly and S.E.-ly 
 ' winds) extending over several degrees, and directed towards the vortex. 
 ' (See Fig. 22, p. 80.) On that day, then, and on subsequent days, the 
 
 * bearing of the centre at some distance could not have been determined 
 ' by the rules laid down on the subject. From the winds and weather 
 
 * at Port Louis on the 25th, for example, the centre, according to the 
 ' Law of Storms, should have been due North of the Island, whereas it 
 ' was W.N. W. of it ; and the bearing of the centre from St. Denis, 
 ' where the wind was S.E., should have been N.E., whereas it was 
 about N.N.W.' 
 
 " In a description of another storm which occurred in March 1860, 
 we read as follows : ' Although the storm on some days appeared to be 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 77 
 
 ' more circular than on others, yet it is probable that during the whole 
 ' time the wind blew round the centre spirally.' 
 
 "An account of the hurricane of the 29th January to the 7th 
 February 1863 contains the following remarks : ' But it is evident 
 ' that at some distance from the centre the wind was not blowing in a 
 * circle. The probability is that it was blowing towards it spirally. 
 ' At all events, it was impossible for the John Barrow, William Carvill, 
 1 Louisa, etc., to determine by the Law of Storms the bearing of the 
 ' centre at noon, although the wind was increasing fast and the barometer 
 ' falling. The winds and weather experienced by these vessels would, 
 ' according to the "Law," place the centre to the N.E.-ward of them; 
 ' whereas it was nearly due North of the William Carvill, and N.N.W. 
 ' of the John Barrow. All of them, in short, had the S.E. trade; but 
 ' it is difficult to know when the trade forms a part of the " storm circle" 
 ' Capt. TILLMAN of the Louisa, who was well aware that a cyclone was 
 ' somewhere to the Northward, ran to the Westward, apparently under 
 ' the impression that, with the wind at S.S.E. and S.E., he could cross 
 ' the storm's track before the centre could reach him ; but after running 
 ' nearly 200 miles W. J S. he found the wind hauling to the East, which 
 ' made him give up the idea of passing in front of the storm. The 
 ' storm was a revolving one, but it was only towards its centre that the 
 1 wind blew in a circular direction, the curvature increasing as the centre 
 1 was approached.' 
 
 " A description of a severe hurricane, which took place from the 8th 
 to the 22nd May 1863, contains the following remarks: 'Although, 
 ' however, it is quite evident that the wind blew round the centre 
 ' uninterruptedly, it does not appear that it did so in a circle. The disc 
 ' of revolving air was a vortex, the "Western side of which was circular, 
 ' or nearly so, while on its Eastern side the wind blew more or less 
 ' towards the centre, the Westerly winds curving sharply to Northerly 
 ' and North-easterly winds, and the Easterly winds blowing towards the 
 ' centre.' 
 
 " The years 1864-68 being remarkable for a comparative absence of 
 hurricanes in the Southern Indian Ocean, we had not many oppor- 
 tunities of studying their forms, and the bearings of their centres with 
 respect to the directions of the wind. But from what we already knew, 
 it was thought proper to call attention to the subject in a paper which 
 was read before the British Association in 1867. The following is a 
 quotation from the published report of that paper : ' As the trade- wind 
 ' in front of a revolving storm often blows in strong gales with a falling 
 ' barometer over many degrees in longitude, and the direction of the 
 ' wind, especially at a distance, is far from being at right angles to the 
 ' bearing of the centre, severe losses have occurred in consequence of 
 
78 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 ' vessels, having the wind at S.E., running to the West or N.W., with 
 ' the view of crossing the storm's path, under the impression that the 
 ' centre bore N.E. In place of bearing N.E. when the wind is from 
 ' S.E., the centre may bear North or N.N.W.; and if the storm be 
 ' travelling to the S. W., as is often the case, a vessel steering Westward 
 ' or N.W. may be running to her destruction. During a hurricane in 
 1 February 1860, for example, a number of vessels left the roadsteads of 
 ' Reunion with the wind at S.E., and running to the N.W. got into the 
 * heart of the storm. Several of them were wrecked on the coast of 
 ' Madagascar, others were never heard of, and of those that returned 
 ' some had to be abandoned.' 
 
 " On approaching a cyclone on its Southern side, a vessel always 
 encounters a strong trade- wind, and the bearing of the centre can seldom 
 be inferred from the direction of the wind. This is often exemplified at 
 the commencement of a storm at Mauritius, as already referred to in 
 the cases of the hurricanes of March 1818 and January 1860. But we 
 are not now alluding to such cases, but to the direction of the wind in 
 what may be called the body of the storm. 
 
 "At 8.30 A.M. on the 25th February 1860 a number of vessels in 
 the roadstead of St. Denis received orders to put to sea. The barometer 
 was then at 29 '68 inches; the sea was very high; the wind was blow- 
 ing from S.E. in gusts ; much rain was falling in the squalls ; the nimbus 
 clouds were moving with great rapidity; the sea was increasing more 
 and more, and everything indicated the approach of a hurricane, 
 (BRIDET, Etude, etc.) 
 
 "In all, 41 vessels left the roadsteads of Reunion, either on the 
 morning of the 25th or previously. 
 
 " Now, as the wind was from S.E., the ' Law of Storms ' placed the 
 centre away to N.E., and, naturally enough, the advice given to the 
 commanders of these vessels would be to run to the N.W., so as to cross 
 the storm's path in front of the centre, and get into the N.W. quadrant. 
 But we have ample proof that, at noon on the 25th, the centre of the 
 storm was nearly 150 miles N.N.W. of St. Denis, and that as it was 
 travelling very slowly, the centre bore between N. by W. and N.N.W. , 
 instead of N.E., at the time the vessels put to sea from that roadstead, 
 with the intention, no doubt, of crossing the storm's path. 
 
 " Most of the vessels steered towards some point between N.W. and 
 West, but only four of them succeeded in crossing the storm's path in 
 front of the centre. One of these was a steamer, and two of the other 
 three had put to sea from other roadsteads before the 25th. Fortunately 
 for them, the storm was travelling at the rate of only two to three miles 
 an hour, otherwise they would probably have been less fortunate. If 
 

 MODIFICATION OF THE CIRCULAR ^o^^OF STORMS. 79 
 
 the storm had been progressing at the rate of five or six miles an hour, 
 as often happens, they could not have crossed in time. 
 
 " Four other vessels which steered West, in place of N.W., escaped 
 without serious loss. This is entirely attributed by M. BRIDET to their 
 having kept on the port tack, but their comparative safety was fully as 
 much due to their having been considerably further from the centre 
 than the vessels which ran to the N.W. -ward. One of the commanders, 
 on his return, thought fit to apologise to the Port Captain for not 
 having followed his instructions. ' According to the Law of Storms I 
 ' should have run to the N.W. to try and pass to its navigable side, but 
 ' the fear of broaching-to prevented me, and I resigned myself to receive 
 ' the full force of the hurricane, my ship, in fact, being quite prepared 
 * for lying-to.' It was well that he did not run to the N.W. 
 
 " As for the remaining 33 vessels the results were deplorable : 55 
 men perished, either swallowed up by the sea, remarks M. BRIDET, or 
 succumbing to disease contracted after the sufferings of shipwreck. 
 Three vessels disappeared and were never heard of. Three more were 
 wrecked on the coast of Madagascar. Four were afterwards condemned 
 at St. Denis, and two more at Port Louis. In fine, only 10 vessels 
 sustained slight damage; the remaining 31, either through total loss 
 or necessary repairs, costing the insurers, according to M. BRIDET, 
 3,360,000 francs, or 134,400. 
 
 " While lamenting these disasters, M. BRIDET, in very strong terms, 
 ascribes them to the ignorance or the culpability said to have been 
 shown by the commanders in not having acted in conformity with ' the 
 principles of the new science.' But it will not, we think, be difficult to 
 prove that, on this occasion at least, the ' principles ' themselves were 
 more at fault than the commanders. 
 
 "When the wind was from S.E. at St. Denis (Reunion) on the 
 morning of the 25th, with a rather low and falling barometer and eveiy 
 appearance of a hurricane, M. BRIDET, believing that the centre of the 
 cyclone bore N.E., and that the vessels were in front of it, would have 
 no doubt counselled the commanders to run to the N.W. with all 
 possible speed. 
 
 "Now the centre did not bear N.E., but between N. by W. and 
 N.N.W. 
 
 This very important point is fully proved by observations made at Mauritius (M), 
 and Reunion (R), and by the following vessels: Emily Smith (e), Swallow (s\ 
 steamer Somme (arrow above R ), Colbert (arrow below R , or Reunion), Ocean Wave 
 (w) near Foule Point, PJmnix (p), Johanne (j), Chene (c), steamer Bahiana (b), 
 and Jemmy (i).* 
 
 * The letters show the positions of the vessels and of Mauritius and Reunion on Fig. 22, between 
 at. 14 and 26 S., long. 45' to 63 E. 
 
80 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 "We have, then, the direction of the wind at 12 different points on 
 all sides of the centre ; and these show that, at noon of February 25th, 
 the centre was in latitude 18 40' S., longitude 54 30' E., near the 
 position of the Chene. These observations, however, not only enable us 
 to determine with sufficient accuracy the position of the centre at noon 
 on the 25th, but they 
 also afford a good idea 
 of the form of the 
 cyclone (see Fig. 22). 
 We see that from the 
 Emily Smith and Swal- 
 low northwards to the 
 Ocean Wave, the wind 
 curved round from 
 S.E.-ward to S. by W. 
 on the S.W. and West 
 sides of the storm, but 
 that as it pursued its 
 course to the N.E.- 
 ward it must have 
 sharply curved so as 
 to form the N.N.W. wind experienced by the P/icenix. It then gently 
 curved round on the N.E. side of the storm, as shown by the N.E.-ly 
 winds of the Chene and Jolianne. To the Southward and Eastward of 
 these vessels, again, we find, as at Mauritius and with the Bahiana, 
 an Easterly wind almost directed to the centre, while at and near 
 Reunion there was a strong S.E.-ly wind, also blowing more or less 
 towards the centre. 
 
 " About 150 miles to the N.E. by N. of St. Denis the Johanne had 
 a hard gale from N.N.E., whereas if the S.E. wind at St. Denis had 
 belonged to a circular storm, the Johanne should have had the wind 
 from N.W. by N. 
 
 " M. BRIDET does not give the position of the centre at noon on the 
 25th, but he places it at noon on the 26th in lat. 19 2' S., long. 53 20' E., 
 and it was placed by us in lat. 19 S., long. 53 30' E., which is 
 practically the same. But we have seen that at noon on the 25th the 
 centre was in 18 40' S. and 54 30' E. From noon on the 25th to 
 noon on the 26th, therefore, the storm travelled nearly 62 miles on a 
 W.S.W. course, or at the rate of 2-6 miles an hour. 
 
 (See Fig. 22 for the position of the centre on the 25th towards which the spirals 
 converge, and on 26th, indicated by a small white circle. ) 
 
 " We thus arrive at the following important results : 
 
 " Three hours after the vessels left St. Denis on the 25th, with the 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 81 
 
 wind from S.E., the centre bore about 150 miles to the N.N.W.; and as 
 the storm was travelling at the rate of only 2 *6 miles an hour, it is clear 
 that at the time the vessels put to sea (9 A.M.) the centre must have still 
 been nearly in the same direction, the storm in the three hours' interval 
 having progressed only 7 -8 miles to the W.S.W. At noon on the 26th, 
 the centre was 160 miles due N.W. of St. Denis. Hence, if the vessels 
 had all steered N.W. 011 leaving St. Denis, at the average rate of seven 
 knots an .hour, they would have reached the spot occupied by the centre 
 at noon on the 26th (as given by M. BRIDET himself) at 8 A.M. on the 
 26th, or four hours sooner than the centre, and have been at a distance 
 of only ten miles from it. 
 
 " A part of M. BRIDET'S censure is directed against those commanders 
 who, by running to the N.W., and getting ultimately involved in the 
 dangerous part of the hurricane, did not from the outset run on the port 
 tack. But what we contend for is, that as the centre did not bear N.E. 
 but N.N.W., it was, under any circumstances, most dangerous to steer 
 to the N. W., and that heavy losses were to be fully expected, seeing that 
 the vessels were going straight to the centre. Four vessels did succeed in 
 crossing the storm's path in front of the centre. But this was the result 
 of mere hazard, for it is certain that none of these vessels, when they 
 left St. Denis, knew either the bearing of the centre or the storm's rate 
 of progression. They owed their safety to the fact that the storm was 
 travelling very slowly, and probably to the circumstance that they were 
 enabled to run faster than the other vessels, and very little merit is due 
 to their commanders. 
 
 " Moreover, the vessels that were fortunate enough to cross in front 
 of the storm kept on the starboard tack, which, as the storm was 
 travelling on a W.S.W. course, is the tack usually recommended to 
 vessels about to pass into the right-hand semicircle, it being expected 
 that the S.E. and S.S.E. winds would soon veer to South and S.W. 
 That the less fortunate vessels, therefore, should be blamed for running 
 on the starboard tack at the outset, is, we think, a mistake. For aught 
 they knew, the storm might have been travelling on a S.W. by S. 
 course, in which case they would have been at the outset in the right- 
 hand semicircle. It is true that, when they found the wind hauling to 
 the East, they should, if possible, have been put on the port tack. But 
 this may have been impossible. They ran to the Northward, in 
 expectation of the wind veering to South and S.W. ; and before the S.E. 
 and S.S.E. wind began to veer, not to South, but to East, they were 
 irretrievably involved in the storm. Their great misfortune was that 
 they held to the N.W. at all, on any tack. 
 
 "We pass to another point which bears on the subject : 
 
 " It has been observed at Mauritius, that when cyclones pass on the 
 
 F 
 
82 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 North side of the island, the wind often veers only from S.E.-ward to 
 N.E.-ward, or to North at the utmost, although, in pursuing its course, 
 the centre of the storm is afterwards found to bear West, or even S.W., 
 of the Observatory. 
 
 "In the destructive hurricane of the 18th to the 29th February 
 
 1860, for example, already referred to, the wind veered only from S.E. 
 to N.N.E., and then gradually backed to E.S.E. ; and when the wind 
 was from N.E. on the 27th, the centre of the storm did not bear N.W., 
 according to the * Law,' but West. 
 
 "Again, in the great hurricane of the 9th to the 19th February 
 
 1861, the centre of which passed between Mauritius and Reunion, the 
 wind, it is true, veered from S.S.E. to East, N.W., and West, but its 
 direction was not a sure indication of the bearing of the centre. On the 
 17th, 'for instance, when the wind at Port Louis was from N.N.W., the 
 centre did not bear W.S.W., but between S.S.W. and S. by W. 
 
 "From the 4th to the 10th January 1871 a storm passed on the 
 North side of Mauritius. The wind at the Observatory veered only from 
 S.S.E. to N.E., and then backed to E.S.E. The storm was traced from 
 latitude 17 S., longitude 60 45' E. at noon on the 4th, to latitude 
 30 28' S., longitude 57 48' E. at noon on the 10th. It travelled at 
 first on a W.S.W. course, and afterwards on a S.W., South, and S.E. 
 course, passing a few miles to the Westward of Reunion on the 6th, 
 where it committed considerable damage. Now at noon on the 6th, 
 when the wind at Port Louis was strong from N.E., the centre, in place 
 of bearing N.W., was 180 miles to the S.W., the wind at Port Louis 
 blowing directly towards it. Many persons at Mauritius thought that 
 as the wind did not veer beyond N.E., Reunion would be spared, but 
 the arrival of vessels with full information showed not only that Reunion 
 had suffered, but that the storm had passed to the Westward and South- 
 ward of that Island. 
 
 "At a meeting of this Society held on the 23rd March 1871 the 
 storm's course and extent were discussed, and particular attention again 
 called to the fact that the bearing of the centre was far from being at 
 right angles to the direction of the wind, especially at a good distance from 
 it ; and it was also remarked that, as usual, some of the vessels that had 
 put to sea from Reunion had not since been heard of, they having run 
 to the N.W. when the wind was from S.E., and probably got into or 
 near the centre. 
 
 "About this time I had a conversation with the harbour-master, 
 Captain WALES, on the subject, and made a rough sketch of what I 
 conceived to be the form of cyclones in these latitudes. Captain WALES 
 at once perceived the practical bearing of the remarks submitted to him, 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 83 
 
 and in a few days he drew up a paper, ' On the Converging of the Wind 
 in Cyclones,' which was read before the Society (see pp. 71-74). 
 
 " Between the 8th and 19th February 1872 a very severe hurricane 
 came down from the E.N.E., the centre passing about 70 miles North 
 of Mauritius early on the morning of the 6th, and 40 to 50 miles N.W. 
 of Reunion in the night of the 16th to the 17th. It did little harm at 
 Mauritius, but the crops at Reunion suffered greatly, and several lives 
 were lost. At the Observatory (Mauritius) the wind veered from S.S.E. 
 to N.N.E., and then backed to E.S.E. The storm's path and extent 
 were discussed in a paper read before the Society on the 28th March 
 1872; and we know bej^ond all doubt that when, at noon on the 17th, 
 the wind was blowing at Mauritius in strong breezes from N.E. by N., 
 the centre of the storm bore about 240 miles W.S.W. of the Observatory. 
 On this occasion, also, a number of vessels put to sea from Reunion, and 
 some of them have never returned. 
 
 "From the 4th to the 10th January 1873 another severe hurricane 
 came from the N.E. -ward, passed North of Mauritius, and committed 
 great damage at Reunion, blowing down buildings, destroying the crops, 
 and causing general distress. We know that at Mauritius the wind 
 veered only from S.E. to N. by E., remaining but a short time at the 
 latter point, from which it backed to East, and that when the wind was 
 from N. by E. the centre of the storm bore about S.W. by W. 
 
 " Shortly after the storm of 1871 or 1872 (I forget which), I received 
 some copies of a Reunion newspaper containing a discussion of the storm 
 by M. BRIDET, as well as several letters, in which the writers, or some 
 of them, requested that gentleman to explain the fact that the wind at 
 St. Denis did not veer more than it did, the circular theory requiring 
 it to have passed to the West of North. I am not aware that M. 
 BRIDET has given any satisfactory answer to these inquiries, his reply 
 merely re-asserting the correctness of the ' Law of Storms ' as expounded 
 by himself and others. Another writer, however, has suggested that, 
 in approaching Mauritius and Reunion, cyclones might be flattened in 
 consequence of the resistance presented by the high lands of these 
 islands, and, if so, that this would account for the wind veering so little. 
 But an examination of storms which have taken place at a great distance 
 from high land, shows that their forms were nearly the same as in the 
 case of storms passing near Mauritius and Reunion, and therefore the 
 supposed influence of the mountains of these islands in altering the 
 form of cyclones cannot be accepted as an explanation of the observed 
 facts. This is a point of so much importance that we do not hesitate to 
 give the details, for one day (the 16th), of the hurricane of the 8th to 
 the 22nd May 1863. 
 
84 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 (The ships involved were the Golconda (g f ), Francis Banfidd (6), Formosa (/), 
 Rajasthan (r\ Lochnagar (I), Fez Eabannee (z), James Mussel (/), Glendevon (g), 
 Tjielingsie (t), Sugar (s f ), Ms (i), Alice Maud (m), Scinde (s), Earl Dalhousie (d), 
 Herald (h), Adela (a), and the Fairy, represented on Fig. 23 by their respective 
 letters : the area embraced is from latitude to 26 S., and between longitude 71i 
 and 89 E.) 
 
 "The centre of the storm at noon was in latitude 11 20' S., longitude 
 78 50' E., or about 20 miles to the Eastward of the James Russel. 
 
 " The relative positions of the vessels, and the winds they had (as 
 laid down on Fig. 23), show that the S.E. trade-wind gently curved 
 round on the West side of the 
 storm, as shown by the logs of 
 the Golconda, Francis Banfield, 
 Formosa, Rajasthan, Lochnagar, 
 and Fez Rdbannee ; that nearly 
 380 miles due North of the 
 centre the Ibis had a strong 
 breeze from West ; that between 
 the Ibis and the Alice Maud the 
 wind must have curved sharply 
 round from West to N.E.; that 
 the N.E.-ly winds of the Alice 
 Maud, Scinde, Earl Dalhousie, 
 Herald, and Adela were directed 
 towards the centre; and that 
 when the Tjielingsie had a storm 
 from E.S.E. to East, the centre 
 bore about 120 miles to the 
 N.W.-ward of her. 
 
 " The form of this cyclone, therefore, strongly resembled that of the 
 25th February 1860, which will be best seen by comparing Figs. 22 
 and 23. 
 
 "The Earl Dalhousie had been scudding since 5 P.M. on the 14th, 
 and by 8 A.M. on the 16th she had gone three times round the centre, 
 the wind veering regularly from N.E. to East, S.E., South, S.W., West, 
 etc., back to N.E. , until at last she apparently got into the central calm 
 for a few minutes at 3.30 A.M. on the 16th. But no sooner did she get 
 out of the centre, with the wind at N.N.W., than she again began to 
 scud. By this time, however, the storm, which had been almost 
 stationary, was travelling at the rate of 15 miles an hour, and she could 
 not manage to run round the centre a fourth time. Still she held to the 
 S.W. before the N.E. wind at the rate of 10 to 11 knots an hour, until, 
 finding little improvement, she was hove-to at 6 P.M. 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 85 
 
 "During the latter part of the 16th and the earlier part of the 17th 
 the vessels that had the wind from the N.E.-ward (Alice Maud, Scinde, 
 Earl Dalhousie, Herald, and Adda) were either hove-to, or proceeding 
 slowly to the Southward or S.W.-ward. But on the 17th, the weather 
 having much improved, and the barometer risen considerably, they ran 
 to the Southward and S.W.-ward; and on the 18th, the storm recurving 
 and travelling slowly, most of them got into the centre, or close to it, a 
 second time, and suffered severely. 
 
 " The Earl Dalhousie scudded as before ; but by the time she had 
 brought the wind from N.E. to S. by W., the rudder-head twisted, and 
 refusing her starboard helm, the ship broached-to on the starboard tack, 
 and went over on her beam ends, with her lower yards in the water, the 
 lee main deck full, and the sea washing up half-way on the poop deck. 
 In this dilemma it was found necessary to cut away the three masts by 
 the deck. 
 
 " The Alice Maud foundered, but fortunately her officers and crew 
 were saved by the Chanaral. 
 
 " The Herald, proceeding more slowly to the S.W.-ward than the 
 Earl Dal/iousie, got into the centre only at 10 A.M. on the 19th, and was 
 dismasted. 
 
 "At 10 P.M. on the 16th, the James Russel had the wind from 
 W. by S., from which it veered to W.N.W. and North early on the 17th. 
 The weather being now more moderate, she ran South with the wind 
 veering to N.N.E. and increasing, and by 3 P.M. on the 18th, after being 
 dismasted, she was in the central calm, with her barometer at 2 7 '30. 
 At 5 P.M. the wind recommenced from West to S.W., and blew with 
 great fury for two hours. 
 
 " The Scinde, going at the rate of only 2 to 3 knots, did not get 
 near the centre a second time, but the Adela came up close to it, and 
 was in great danger of foundering. 
 
 " The Fairy, which had also the wind moderating at N.E. on the 
 16th (position not given), bore away S.W., and on the 17th she was 
 1 running before the wind (from N.E.-ward) with all possible speed.' 
 At noon on the 18th, in latitude 14 9' S., longitude 78 10' E., with a 
 furious hurricane from the N.E.-ward, she was laid-to on the starboard 
 tack. 'At 6 P.M. the hurricane, which was then raging with tremendous 
 1 fury, ceased in less than one minute to an entire calm.' At 7 the 
 hurricane recommenced from the opposite quarter, namely, S.W. She 
 was then kept right before the wind, but in ten minutes she was thrown 
 on her beam ends, and was for some time in great danger. On the 
 20th she was abandoned, her officers and crew taking refuge on board 
 the Arundel from Bombay to Liverpool. 
 
 " It is quite certain that all these vessels, if they had not run to the 
 
86 MODIFICATION OF THE CIRCULAR THEORY OF STORMS, 
 
 Southward and S.W.-ward, but had stood back for 48 hours, would 
 have escaped. But what we have especially to remark is this, that, in 
 all probability, the commanders were entirely mistaken as to the bearing 
 of the centre. Most of them had been more or less involved in the 
 hurricane from the 12th to the 16th, but at last the wind began to 
 moderate, and the barometer to rise rapidly. Now, with the wind from 
 N.E., and the barometer low, though rising, the centre, according to 
 the circular theory, should bear N.W. ; and as the weather improved 
 and the barometer continued to rise, it would naturally be inferred 
 that, with the centre in that direction, and its distance increasing, 
 there could be no great danger in running to the S.W.-ward. But at 
 noon on the 16th, when the wind was from N.E.-ward, the centre was 
 from 200 to 300 miles to the S.W.-ward of the vessels, and therefore 
 at that distance the wind blew directly towards it ; and it cannot fail 
 to be remarked that this is analogous to what has already been pointed 
 out in the case of hurricanes passing near Mauritius and Reunion, 
 the N.E.-ly winds at a distance blowing more or less towards the 
 centre. 
 
 " On the other hand, we have evidence that, near the centre, the 
 wind blows, if not in a circle, at least nearly so. A good example of 
 this is furnished by the Earl Dalhousie in running round the centre 
 of the hurricane of May 1863. At noon on the 15th she was in lat. 
 8 55' S., long. 84 32' E. The following is an extract from her log : 
 
 " 'At 1 A.M. a heavy gale from East ; course West, 10 to 11 knots. 
 At 2, wind S.E. ; course N.W., 10 to 11 knots ; raining in torrents. At 
 3, wind South; course North, 10 knots; the roaring of the wind is 
 something fearful, and the intense darkness is truly appalling. At 4, 
 wind S.W. ; course N.E., 10 knots ; bar. 29-70. At 5, wind W.N.W. ; 
 course E.S.E., 10 to 11 knots. At 6, wind N.N.W. ; course S.S.E., 10 
 knots. At 7, wind North ; course South, 1 1 knots ; blowing with 
 tremendous fury, and raining in torrents ; bar. 29 '45. At 8, wind 
 N.N.E.; course S.S.W., 11 knots. At 9, wind N.E. by E. ; course 
 S.W. by W., 11 knots. At 10, wind East; course West, 11 knots; 
 bar. 29-35 ; same weather and sea ; still scudding.' 
 
 " It would thus appear that it took 9 hours to go completely round 
 the centre ; and as the distance run was 95 miles, the diameter of the 
 cyclone, on the supposition that the wind blew in a circle, and that the 
 storm was stationary, would, with the Earl Dalhousie, be nearly 30 
 miles. The storm, however, was not stationary, but travelling at the 
 rate of 3-3 miles an hour to the S.W.-ward. During the two hours, 
 therefore, it took to bring the wind from East to South, the centre 
 would have advanced nearly 7 miles, and have approached the vessel 
 when on the S.W. side of the storm. But in getting round on the 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 87 
 
 Northern and Eastern sides, her distance from the centre would be 
 increasing. This may partly account for the fact that in the Western 
 half of the storm the wind veered rapidly, taking only about 3J hours 
 to pass from East to West, whereas it took 5 J hours to pass from West 
 to East. There is reason to suppose, however, that even at that short 
 distance from the centre, the wind on the N.E. and Eastern sides of the 
 storm was incurving towards the centre, for it took only one hour to 
 bring the wind from S.W. to W.N.W., apparently showing the sharp 
 curving in that part of the storm already alluded to, while it took a 
 whole hour to bring the wind from W.N.W. to N.N.W.; and it will be 
 seen that it remained 2 hours between North and N.E. by E., whereas 
 on the Western side it veered four points in one hour. During all this 
 time the vessel was gradually approaching the centre, for with the 
 N.E.-ly and Easterly winds she gained more than she had previously 
 lost. 
 
 " Her next run round the centre was performed in a little more than 
 8 hours ; and as the subject is rather interesting, it may be proper to 
 give the log in full, commencing where we stopped with the wind at 
 East : 
 
 "'At 11 A.M., wind S.E. by E. ; course N.W. by W., 11 knots; 
 ship labouring very much. Noon, wind S. by E. ; course N. by W., 10 
 to 1 1 knots ; sky densely overcast ; same fearful hurricane, and rain in 
 torrents ; bar. 29-25. At 1 P.M., wind S.W. by W. ; course N.E. by E., 
 
 10 to 11 knots; dense upper clouds, and light lower scud flying in 
 various directions. At 2, wind W.N.W. ; course E.N.E., 10 to 11 
 knots. At 3, wind N.W. by W. ; course S.E. by S., 10 knots ; blowing 
 a most severe storm, with constant heavy rain. At 4, wind North ; 
 course South, 11 knots; bar. 29-20. At 5, wind N.E. ; course S.W., 
 
 11 knots. At 6, wind E. by N. ; course W. by S., 12 knots. At 7, 
 wind E. by S. ; course W. by N., 12 knots.' 
 
 From 10 A.M. to about 6J P.M. the wind veered right round the com- 
 pass, the vessel making 92 miles ; and it will be seen that, as before, 
 the wind must have curved sharply from S.W. by W. to W.N.W., and 
 very slowly from W.N.W. to N.W. by N. 
 
 " It would be easy to adduce many more instances showing that at 
 least at some distance from the centre the form, of cyclones in the 
 Southern Indian Ocean is not truly represented by concentric circles, 
 and that the rule usually given for ascertaining the bearing of the centre 
 is often inapplicable. But we must stop here for the present, and now 
 proceed to state briefly the conclusions at which we have arrived. 
 
 " To find the exact directions in which the wind blows in a hurri- 
 cane is not an easy matter. The true positions of the vessels are not 
 always given, especially when their courses are frequently changed, and 
 
88 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 it is impossible to get observations of the sun or stars. It doubtless 
 happens, also, that the directions of the wind as given in the log-books 
 are often incorrect to the extent, it may be, of two or three points. 
 Then, although it may be quite certain that a cyclone exists, yet there 
 may be only two or three vessels involved in it, and these may be all in 
 the same part of the storm. For these reasons, it takes a long time, 
 and the investigation of many cyclones, to obtain an approximate know- 
 ledge of their forms. We, in fact, get only occasional glimpses, at such 
 times as in consequence of many vessels being on all sides of a storm 
 we can obtain good averages ; and this will account for the want of 
 precision which sometimes characterizes the allusions made to the forms 
 of cyclones in the descriptions which we have given of them. 
 
 " Comparing, however, the results obtained during the last 20 years, 
 and especially the last 12 years, we have no hesitation in saying that, 
 generally (we do not say always), fully developed cyclones in the 
 Southern Indian Ocean have the same form as the cyclones of the 25th 
 February 1860 and the 16th May 1863. 
 
 " The forms of these two cyclones are given in Fig. 22 (p. 80), and 
 Fig. 23 (p. 84), as examples of what may generally be expected. 
 
 " It will be seen, as already remarked, that the S.E. trade- wind 
 curves round on the Western side, giving the storm in that part of it a 
 more or less circular appearance, but that the wind curves sharply from 
 West to N.W. and North, and that the Easterly winds (particularly 
 from E.N.E. to E.S.E.) blow nearly towards the centre, except when 
 near it. 
 
 "The line A B (Fig. 22, page 80) is supposed to pass through 
 Mauritius, and it will be seen that the storm commences with the wind 
 from S.E.-ward at A, but that as the centre advances to the S.W.-ward 
 the wind gradually veers to North at B. 
 
 " The extent to which the wind veers at a fixed point must depend 
 upon the size of the storm, the distance from the centre, and the dia- 
 meter of the more circular part near the centre. 
 
 " Having given the general form of cyclones, I should perhaps leave 
 the rest to practical men. But there can be no harm in stating, subject 
 to correction, what appears to be necessary for avoiding the centre ; and 
 in doing so I will quote here some remarks made in April 1863 : 
 
 " ' A rotatory storm, in the Indian Ocean, not being a detached body 
 of air whirling round in a circle beyond which light airs and variables 
 prevail but the result of the conflicting action of two antagonistic 
 winds which often respectively extend hundreds of miles North and 
 South of the centre, towards which they at a distance begin to curve 
 gradually, the degree of curvature increasing as the centre is approached, 
 until at length they rush round it with more or less violence, it becomes 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 89 
 
 a question of practical importance to know when can the direction of 
 the wind be relied upon for a sufficiently near estimate of the bearing 
 of the centre ; for by considering the direction of the wind, when it is 
 S.E. for example, to be tangential to a storm-circle, and then running 
 to the Westward, with the view of crossing the storm's track, the vessel 
 may be steered headlong to destruction. 
 
 " ' There does not appear to be much difficulty in avoiding a rotatory 
 storm when the vessel is in its Northern half, and has the wind at any 
 point from South to North, through the West, or even from South to 
 N.E. through the West. 
 
 " ' If the wind be between South and West, a vessel should run to 
 the N.W.-ward ; if it be between West and North, she should run to 
 the N.E.-ward ; and if it be from N.E.-ward, she should make as much 
 Easting as possible. 
 
 " ' Or, if the sea and weather allowed, she might be hove-to on the 
 proper tack till the barometer rose and the weather improved, [except, 
 perhaps, when the wind is from N.E., in which case, if she be between 
 the parallels of 18 and 30 S., the storm may be travelling to the 
 S.E. -ward, and bearing down upon her. If such should be its course, 
 she should endeavour to cross its path by running to the S.W.-ward.] 
 
 N.B. Compare this exception (which is included in brackets) with paragraph 
 (6), p. 91, where Mr. MELDRUM, as the result of ten years' additional experience, 
 advises "making as much Easting as possible," instead of running to S.W.-ward. 
 
 " ' When, however, the wind is from any point from South to North, 
 by the West, there is really no danger whatever. All that is required 
 is either to stand away from the centre, by going to the Westward, the 
 N.W.-ward, the Northward, the N.E.-ward, or the Eastward, according 
 to the direction of the wind, or to lie-to. It might easily be shown that 
 all the homeward-bound vessels that come here for repairs, in con- 
 sequence of their having sustained damage in a cyclone, which they 
 entered from the North, incurred the heavy expense and the long delay 
 thus occasioned, by running to the Southward or S.W.-ward, in place 
 of heaving-to in time, or standing off till the barometer should rise and 
 the weather improve. There is a strong temptation to such vessels to 
 hold their course with a favourable breeze ; but an increasing Northerly 
 or N.E.-ly wind, with a falling barometer and bad appearance, should 
 warn them of the extreme danger of running to the S.W.-ward at 
 the rate of perhaps 9 knots an hour. How many vessels, by taking 
 supposed advantage of these Northerly winds, have got in front of a 
 storm, been dismasted, and afterwards obliged to remain at Mauritius 
 for months, if they should escape being condemned ! A delay of 24 or 
 48 hours would, almost in every instance, have saved both vessel and 
 cargo. 
 
90 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 " ' On the other hand, to vessels with the wind between South and 
 East, or E.N.E., especially between S.S.E. and East, the case is widely 
 different ; for the storm is travelling towards some point between 
 W.S.W. and S.E., and that point cannot always be known, because the 
 wind may not, as yet, be blowing round the centre of the storm in a 
 circle, although the barometer be falling and the wind increasing. 
 
 " ' To vessels in this situation the safest course seems to be to lie-to, 
 and carefully watch the wind and barometer. 
 
 " ' If the wind should haul decidedly, either to the East or to the 
 South, the passage of the centre, with respect to the vessel's position, 
 may be approximately inferred. 
 
 " ' When the barometer has fallen four-tenths of an inch, the 
 direction of the wind may be taken to be nearly at right angles to the 
 direction of the centre. 
 
 N.B. After ten years' more experience, a fall of tf/Z-tenths is considered 
 necessary before the above conclusion can be drawn ; see paragraph (5), p. 91. 
 
 " ' If the vessel should then have the wind from S.E., she should 
 endeavour to run to the Westward, so as to cross the storm's path. By 
 running to the Westward before the barometer has fallen more than a 
 tenth or so, she may be approaching the centre of the storm, as the 
 Louisa was doing when she had the wind from the S.E.-ward on the 
 2nd February 1863. 
 
 " ' With the wind from S.E.-ward, the same difficulty presents itself 
 to vessels at anchor at the bell-buoy (Mauritius) and at Reunion, 
 namely, that of knowing the bearing of the centre, as, at the outset, the 
 wind may be far from blowing at right angles to the direction of the 
 centre. If the centre should bear North or N.N.W., in place of N.E., 
 and the vessel run to the N.W.-ward, she may, by keeping that course, 
 get into the heart of the storm, as has frequently happened to vessels 
 leaving St. Denis, and as probably happened to the Shah Allum, after 
 she left the bell-buoy in January 1863.' 
 
 " * The best course would perhaps be (if the vessels cannot be brought 
 into port) to put to sea on the first decisive indications of the coming 
 storm, and, after obtaining sufficient sea-room, to lie-to, watch the wind 
 and barometer, and be guided by them ; not hesitating to run to the 
 Westward or N.W.-ward, if the barometer should have fallen three or 
 four-tenths of an inch, with the wind between S.E. and South ; but, if 
 the wind should haul from S.E. to East and N.E., taking care that the 
 vessel be hove-to on the port tack, after getting as far to the Eastward 
 as possible.' 
 
 " After ten years' further experience, we have little to add to what 
 was said in 1863. 
 
MODIFICATION OP THE CIRCULAR THEORY OP STORMS. 91 
 
 " (1.) The really dangerous position for a vessel is with the wind from 
 N.E.-ward to S.S.E., and more especially from the S.E.-ward. 
 
 " (2.) When the wind is between N.E. and E.S.E., our advice would 
 be to lie-to at once. 
 
 " (3.) With the wind from S.E.-ward, and the barometer still high, 
 though falling, we would also heave-to, because, as yet, the bearing 
 of the centre cannot be known. If the wind hauled decidedly to the 
 Southward, and passed to the South of S.S.E., we would stand to the 
 N.W. -ward without loss of time, but if it veered to the Eastward we 
 would either lie-to on the port tack, or endeavour to make Easting. 
 
 " (4.) On no account whatever would we run to the Southward or 
 S.W.-ward with the wind anywhere between North and East, but make 
 as much Easting as possible. 
 
 " (5.) The most dangerous case of all is when the wind is steady from 
 S.E., the barometer falling, and the wind gradually increasing in force. 
 What is a vessel to do under these circumstances ? If she runs to the 
 N.W. she may be going straight to the centre. Upon the whole, we 
 think she cannot do better than lie-to and watch the wind and barometer. 
 The chances are that the wind will veer, and that she will know, before 
 it is too late, whether it is best to run to the N.W. -ward, or remain 
 where she is, or hold to the Eastward. But if the wind does not veer, 
 and the barometer has fallen from the commencement fully s-tenths of 
 an inch, we would, as a last resort, run to the N.W.-ward if possible. 
 We say ^/JT-tenths, because four-tenths, which was the amount of fall 
 mentioned in 1863, is too small. (In the centre of the severest cyclones 
 in these seas the barometer always falls below 28 inches.) 
 
 " The above remarks apply chiefly to vessels encountering cyclones 
 which have not yet commenced to curve to the Southward and S.E.- 
 ward. 
 
 " (6.) When a storm is travelling to the Southward or S.E.-ward, I 
 would still, with the wind anywhere between North and East, make as 
 much Easting as possible, and not, as recommended in 1863, run to the 
 S.W.-ward, for we now know the N.E.-ly and Easterly winds often, if 
 not always, blow towards the centre. 
 
 " (7.) Storms do not always curve to the Southward and S.E.-ward, 
 and those that do so seldom proceed far in those directions. But if it 
 was found that a storm was travelling to South or S.E., a vessel with the 
 wind from E.S.E. to South should hold to the Westward. 
 
 " It may be proper to remark that the modifications above mentioned, 
 as to the form of cyclones, do not very materially affect the general 
 theory of revolving storms, as established by REDFIELD and REID 
 in America and the West Indies, and by PIDDINGTON and THOM in 
 Calcutta and Mauritius, although, I think, they deserve the attention of 
 
92 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 all interested in navigation. They are, in some measure, a compromise 
 between Prof. ESPY'S views and those of the advocates of the circular 
 theory. 
 
 " The forms given in Fig. 22 (p. 80) and Fig. 23 (p. 84) apply to 
 storms when fully developed, and before they begin to break up. These 
 storms are usually generated between an Easterly and Westerly current 
 of air, and cease between a Northerly and Southerly current, so that at 
 their commencement they are flattened on their Northern and Southern 
 sides, and towards their close on their Western and Eastern sides." 
 
 Such are the conclusions at which Mr. MELDRUM arrived as to the 
 form of cyclones, and the gyration of the winds within the cyclonic 
 area, in the southern hemisphere. But no less important, and much to 
 the same effect, are the results of later observations made in the Bay of 
 Bengal, in the China Seas, the Western Pacific, and among the West 
 Indian Islands \ and to these it will presently be necessary to refer. 
 
 The Law of Storms now enters on a new phase, to appreciate which 
 it is necessary to briefly explain certain barometric phenomena intimately 
 connected with atmospheric circulation. 
 
 The navigator who, day by day, or at shorter intervals, records, in 
 a graphic form, atmospheric pressure, as indicated by the barometer, 
 understands that such a projection represents (according to his limited 
 observations, and hence imperfectly) a barometrical curve for the route 
 he has traversed. If, after a somewhat similar method, but from more 
 exact and continuous observations made at many different places, baro- 
 metric curves be drawn to show the mean annual pressure of the atmo- 
 sphere over certain parts of the globe, as, for example, along meridional 
 lines between lat. 80 N. and 60 S., it will be found that, however 
 much these curves vary in curvature over different meridians, they 
 are characteristically 
 alike in this that 
 there are two parts 
 indicative of a higher 
 pressure than else- 
 where, so that, on 
 viewing them trans- 
 versely to the meri- 
 dian, i.e. in the direction of the parallels of latitude, they would present 
 the atmosphere which encircles the globe as apportioned into five distinct 
 bands or areas, viz., two of high pressure and three of low pressure. 
 
 (See Fig. 24 above, in which the upper curve represents (roughly) the atmospheric pressure over 
 the meridian of 20 W., and the lower curve the atmospheric pressure over the meridian of 100* E.) 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 93 
 
 To Mr. A. BUCHAN (in 1868) is due the credit of having been the 
 first to approximately solve the physical problem of the distribution of 
 atmospheric pressure, and his views have since been corroborated by 
 WOJEIKOF and others. In respect to the mean annual pressure, Mr. 
 Buchan deduces the following important general conclusions : 
 
 "There are two regions of high pressure the one north and the 
 other south of the equator passing completely round the globe as broad 
 belts of high pressure. They enclose between them the low pressure 
 of the tropics, through the centre of which runs a narrower belt of still 
 lower pressure, towards which the trade-winds blow. Since these belts 
 of high pressure can only be maintained by air flowing in upon them in 
 the upper regions of the atmosphere, it is towards these belts of high 
 pressure that the upper currents of the air must flow. The southern 
 belt of high pressure lies nearly parallel to the equator, and is of nearly 
 uniform breadth throughout ; but the belt north of the equator has a 
 very irregular outline, and great differences in its breadth and in its 
 inclination to the equator, these irregularities being due to the unequal 
 distribution of land and water in the Northern Hemisphere. Considered 
 in a broad sense, there are only three regions of low pressure one round 
 each pole, bounded by, or contained within, the belts of high pressure 
 just described, and the equatorial belt of low pressure. The most re- 
 markable of these, in so far as known, is the region of low pressure 
 surrounding the South Pole, which appears to remain pretty constant 
 during the whole year. The depression round the North Pole is divided 
 into two distinct centres, at each of which there is a diminution of pressure 
 greatly lower than the average north polar depressions :" these two 
 centres lie one in the Atlantic in the vicinity of Iceland, and the other 
 in the northern part of the North Pacific. 
 
 These belts, following the course of the sun in its to-and-fro vibration 
 from hemisphere to hemisphere, change their form and position, month 
 by month, more or less rapidly according to the relative distribution of 
 land and water. They have more uniformity in April and October, 
 when atmospheric pressure appears to be more uniformly distributed 
 over the globe, than during the other months of the year, while in 
 January and July (the representative months of extreme seasons) occurs 
 the greatest deviation from the annual mean. 
 
 Of the latter months Mr. BUCHAN remarks : 
 
 " In January the highest pressures are over the continents of the 
 Northern Hemisphere, and the larger the continental mass the greater 
 the pressure ; the lowest pressures are distributed over the northern parts 
 of the Atlantic and Pacific, South America and South Africa, and the 
 Antarctic Ocean. In Central Asia the mean pressure of the atmosphere 
 in this month is fully 30*40 inches, whereas in the North Atlantic, round 
 
94 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 Iceland, it is only 29-34 inches, or upwards of an inch lower than in 
 Central Asia. This area of high barometer is continued westward through 
 Central and Southern Europe, the North Atlantic between lat. 5 N. and 
 45 N., North America (except the north and north-west part), and the 
 Pacific for some distance on either side of lat. 15 N. : it is thus an 
 exaggerated form of the high belt of annual mean pressure, spreading, 
 however, over a much greater breadth in North America, and a still 
 greater breadth in Asia. 
 
 " In July, on the other hand, the mean pressure of Central Asia is 
 only 29*47 inches, or nearly an inch lower than during January; or, 
 putting this striking result in other words, about a thirtieth of the pres- 
 sure of the atmosphere is removed from this region during the hottest 
 months of the year, as compared with the winter season. The lowest 
 pressures of the Northern Hemisphere are now distributed over the 
 continents, and the larger the continental mass the greater is the de- 
 pression. At the same time, the highest pressures are over the ocean, 
 between lat. 50 N. and 50 e S., particularly over the North Atlantic and 
 North Pacific, between lat. 25 and 40 N.; and in the Southern Hemi- 
 sphere, over the belt of high mean annual pressure, which in this month 
 reaches its maximum. Pressure is high in South America and in 
 Australia, just as in the winter of the Northern Hemisphere pressure is 
 high over the continents." 
 
 The distribution of atmospheric pressure at different seasons is best 
 elucidated by maps of the earth's surface, on which the various belts of pres- 
 sure are delineated in two different colours, as in BUCHAN'S " Introductory 
 Text Book of Meteorology," SCOTT'S "Elementary Meteorology," or 
 WOJEIKOF'S " Atmospharische Circulation," to which the reader is referred 
 for a fuller exposition of the subject. 
 
 Our two plates I. and II. represent the distribution of atmospheric 
 pressure for January and July respectively ; and the substance of the 
 foregoing remarks is also projected into curves, in Figs. 25 and 26. 
 
 A mean or average value is often but an inadequate expression for 
 extreme differences, 
 and this is especially 
 the case in respect to 
 meteorological pheno- 
 mena, as may be 
 seen by a comparison 
 of the curves of at- 
 mospheric pressure in 
 Figs. 24 (p. 92), 25, 
 and 26, which are 
 given (in each case) 
 for the same meridians, viz., 20 W. and 100 E. 
 
MODIFICATION OP THE CIRCULAR THEORY OF STORMS. 95 
 
 The meridian of 20 W. passes along the length of the Atlantic, and 
 wholly over water, with the exception of Iceland ; comparing the January 
 curve of pressure (Fig. 25) with that of July (Fig. 26), there is a 
 difference between them, which is most marked in the vicinity of Iceland, 
 lat. 62 N., but which is also sufficiently characteristic in the equatorial 
 regions, where the belt of calms (or area of low pressure), towards which 
 the trade-winds blow, shifts its position through several degrees of 
 latitude, and the area of lowest pressure from lat. 2 or 3 N. in January 
 to lat. 1 1 N. in July; 
 and where also the 
 parts of the curve 
 which represent the 
 belts of high pressure 
 (in lat. 25 to 30) 
 on the polar sides of 
 the trades show the 
 greatest change in 
 the Southern Hemi- 
 sphere. The excessive 
 low pressure of January, with its concomitant storms, in the vicinity of 
 Iceland, disappears in July, while in the region of the great Southern 
 Ocean the pressure remains comparatively unaltered, and always 
 abnormally low. 
 
 The meridian of 100 E. mainly passes over land (Central Asia) in 
 the Northern Hemisphere, but over the ocean south of the equator ; 
 hence the two seasonal (January and July) curves of pressure are 
 markedly contrasted, nor do they bear the slightest resemblance to the 
 curves which pass over the oceanic meridian of 20 W. ; they, in fact, 
 illustrate the effect of the irregular and unequal distribution of land and 
 water, and show the difference which characterizes areas of pressure as 
 principally influenced through one or the other element j thus, it is the 
 northern portion of the curve which undergoes the greatest change : in 
 January, from the high pressure region of Central Asia the N.E. 
 monsoon blows towards the low pressure area, in lat. 5 S. ; in July a 
 reversion occurs, and from the vicinity of the equator the S. W. monsoon 
 blows towards the same region of Central Asia, which is now one of 
 remarkably low, as it had previously been one of remarkably high, 
 pressure. (See Figs. 25 and 26.) 
 
 Exactly as the inter-tropical winds the perennial trades and the 
 periodical monsoons are connected with and influenced by belts of high 
 and low pressure, so are the prevailing winds, and hence the climate of 
 every part of the globe. With the march of the sun from one side of 
 the equator to the other, these belts (and with them the whole atmo- 
 
96 MODIFICATION. OF THE CIRCULAR THEORY OF STORMS. 
 
 sphere) make annually one general oscillation northwards in the 
 northern spring, southwards in the northern autumn not in one unin- 
 terrupted movement, but by a series of minor vibrations, until they have 
 reached their most northern or southern limit, when the oscillations 
 recur in the opposite direction. 
 
 The purpose has here been to indicate in a general way the geogra- 
 phical distribution of the belts or areas of high and low pressure, not to 
 enquire into their cause whether they originate from temperature and 
 humidity, or from these and other motive forces, the relation of which is 
 at present but little understood, for, as Prof. H. MOHN says, " the chief 
 problem in meteorology is the law of the variation of atmospheric 
 pressure, and we may consider ourselves only on the threshold of a 
 knowledge of it."* As a matter of fact, the general atmospheric circu- 
 lation is intimately connected with, if not entirely dependent on, these 
 areas ; and as, on a large scale, the prevalent winds and general climate 
 of tropical, extra-tropical, and polar regions are more or less due to them, 
 so, on a smaller scale, all fluctuations of wind and weather are coincident 
 with local and temporary variations of pressure. 
 
 And how does the general circulation take place 1 An area of high 
 pressure indicates a superfluity of air ; an area of low pressure a 
 deficiency of air. Wind, which is simply air in motion, in flowing away 
 from the area of high pressure, circles round it with the sun, and towards 
 and around the area of low pressure against the sun, in each hemisphere. 
 The Trades, the Monsoons, the S.W. winds of Northern regions, and the 
 " brave " N. W. winds of the Great Southern Ocean, all speak to the same 
 effect ; and so, in a similar manner, wherever there is a local and 
 temporary disturbance of the atmosphere, whether resulting in a high 
 barometer, or in a low barometer as in a cyclone, the system of circula- 
 tion is universal ; and, being so, it is on this that BUYS BALLOT has based 
 the " Law of the Winds " which goes by his name, viz., the wind always 
 flows from the region of the higher to that of the lower pressure, veering, 
 in consequence of the earth's rotation, to the right in the northern 
 hemisphere, and to the left in the southern hemisphere. 
 
 Dismissing this law for the present, and turning from areas of pres- 
 sure which influence the general atmospheric circulation to those which, 
 though they appear to have but a local origin and temporary existence, 
 yet not unf requently take a wide geographical range, as is the case with 
 
 * As to the question of the origin of these areas of high and low pressure, respecting which there 
 is some controversy, it may be well to quote an observation of DOVE'S : " Whenever two phenomena 
 constantly occur together, we may, with some probability, presume that one depends on the other 
 in the relation of cause and effect. But it will remain to be decided which of the two is the cause 
 and which the effect, because it is not impossible that both phenomena may be the result of a third 
 phenomenon ; nor, in fact, can it be at once decided whether, when one of the phenomena is really 
 the immediate effect of the other, the same result uiay not be brought about in a different way." 
 
MODIFICATION OF THE CIRCULAR 
 
 97 
 
 storms, and especially those of the tropics, it may be premised that in 
 this connection, as an area of low pressure has been called a cyclone, so 
 an area of high pressure has been termed an anti-cyclone, owing to the 
 reversal of the order of the circling winds. 
 
 Areas of Low Pressure, or Cyclones. In the northern hemisphere (see 
 Fig. 27) the wind circles round the central area of lowest pressure in a 
 direction against the sun, 
 or contrary to that of the 
 motion of the hands of a 
 watch. In the southern 
 hemisphere (see Fig. 28) 
 the wind circles round the 
 centre of lowest depression 
 in a direction also against 
 the sun (astronomically 
 considered),but this motion 
 is there similar to that of 
 the hands of a watch. 
 But so far as observation 
 goes shown by an ex- 
 amination of the weather 
 maps of the northern hemisphere, and indicated by the reports of able 
 observers in all parts of the world this circling takes the form of a 
 spiral, the tendency of the wind being inwards, and towards the place 
 of lowest barometer, at an angle dependent upon the latitude in which 
 the cyclone is moving, and the lay of the land along or towards which 
 it is travelling : thus the direction of the wind is intermediate between 
 that of the commonly accepted theory of rotatory gales winds blowing 
 in circles returning on themselves and the directly in-blowing theory 
 of ESPY, but more nearly approximating to the former than the latter 
 the inclination being rather centrifugal than centripetal. The amount 
 of incurvature varies considerably in different latitudes : between the 
 tropics it is probably greatest in the rear of the storm ; in the higher 
 latitudes, in the front. In the central area of the depression it is 
 generally dead calm, but not always so in extra-tropical regions. 
 
 The force of the wind in a cyclone in fact, in all storms, and when- 
 ever wind blows around an area of low pressure varies in different 
 parts according to the amount of difference in the pressure over a given 
 distance, which is called the barometric gradient. Now, an isobar is the 
 line (curved, circular or irregular) which passes through all places that 
 at any given time have the barometer at the same height, thus, in Fig* 
 28 the outer irregular curve indicates in every part of it a barometric 
 pressure of 29 '7 inches, the next (inwards) a pressure of 29 5 inches) 
 and so on. Regions of both high and low pressure are surrounded by 
 
 G 
 
98 
 
 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 such isobars, i.e. by lines of equal barometric height. An area of low 
 
 pressure has them crowded together, generally more crowded on one 
 
 side than on the other; and 
 
 the wind will be strong in 
 
 proportion as the isobars appear 
 
 to close on each other, or, in 
 
 other words, to the steepness of 
 
 the barometric gradient.* Thus 
 
 in Figs. 27 and 28, at the parts 
 
 of the diagrams marked a, where 
 
 the isobars close on each other, 
 
 the wind will be stronger than 
 
 at 6. 
 
 Connecting these diagrams 
 with the hurricanes or cyclones 
 of the tropics, it will be seen 
 that the direction of the wind 
 at any part will much depend on the shape or form of the area of low 
 pressure, and the strength of the wind upon the distance separating 
 successive isobars, together with the position of the surrounding areas 
 of high pressure. 
 
 Areas of High Pressure, or Anti-Cyclones. Wherever there is an 
 area of low pressure there must be in the vicinity (so to speak) one or 
 more areas of high pressure, 
 over which the isobars are 
 more widely separated than 
 is the case with those apper- 
 taining to cyclonic areas, and 
 generally with greater irre- 
 gularity of form ; the winds 
 are light, the weather fine, 
 and the sky clear, especially 
 at the centre. From the 
 centre of highest pressure 
 the winds stream outwards 
 (see Fig. 29), often nearly at 
 right angles to the isobars, 
 but, as they pursue their 
 course, they turn in an 
 opposite direction to that 
 
 * The velocity or strength of the wind greatly depends on the difference of the barometrical 
 readings over a given distance ; if, with three places x y z, each successively 50 miles apart, the 
 difference of readings is greater between y and z than between x and y, it will blow stronger 
 between the former (y z) than the latter (x y) : if there is no difference in the barometrical readings 
 there will no wind. 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 99 
 
 assumed by the winds blowing inwards upon an area of low pressure 
 hence, with watch hands in the northern hemisphere, and against watch 
 hands in the southern hemisphere. These are anti-cyclones in contra- 
 distinction to cyclones, and their meteorological difference is exceedingly 
 well marked : not only do they differ as regards the direction and circling 
 of the winds, but the temperature and humidity of the winds connected 
 with each have perceptibly distinct characteristics. 
 
 Together, the cyclone and anti-cyclone are concomitant agents in all 
 atmospheric disturbances ; but it does not follow that every cyclone, 
 considered from a meteorological stand-point, is a hurricane; the zephyr 
 of summer may be cyclonic no less than the storm of winter, differing 
 only in intensity. In great storms, the rotation of the wind and the 
 various weather aspects are much complicated by cyclone following in 
 the wake of cyclone. 
 
 As the areas of high and low pressure are connected together 
 at the surface of the earth by horizontally-flowing currents of air, 
 and since the outflow from the anti- cyclone must be replenished and 
 the inflow to the cyclone must be carried away, it is the most accepted 
 view that the air ascends over areas of barometrical depression (cyclone), 
 and descends over areas of barometrical elevation (anti-cyclone) ; from 
 which it is further inferred that in the upper regions of the atmosphere 
 the conditions of pressure must be reversed as compared with those 
 below, in other words, over an area of low pressure there must be an 
 excess of pressure at a certain level, causing an outflow of the air which 
 is rising ; and conversely over an anti-cyclonic area there must be a 
 deficiency of pressure at a certain height, the tendency of which is to 
 attract the air towards the region of its existence.* 
 
 It must be borne in mind that what has* here been said of areas of 
 high and low pressure, and their geographical distribution, has inci- 
 dentally been introduced to explain to the seaman and general reader, 
 in as brief a manner as possible, the connection of such areas with 
 storms not to discuss a subject that is now engaging the attention of 
 the best meteorologists of Europe and America ; the writer is conscious 
 that the remarks, with the accompanying plates and diagrams, can give 
 but the rudest outline of an all-important branch of meteorology. 
 
 At the annual meeting (1876) of the Meteorological Society, Dr. 
 Mann, the President, in his address compared the meteorological doctrine 
 
 * Besides BUCHAN, WOJEIKOF, MOHN, and others, who agree on this point, Professor HILDE- 
 BRANDSSON, from observations of the cirrus cloud, shows that air flows away from centres of 
 minima, and towards centres of maxima, in the upper regions of the atmosphere. See also, for very 
 valuable and full information on the subjects here incidentally mentioned, " Aids to the Study and 
 Forecast of Weather," by W. CLEMENT LEY, M.A. ; " The Barometer Manual for the use of Seamen," 
 and " Principles of Forecasting by Plans of Weather Charts," by the Hon. RALPH ABERCROMBY 
 pamphlets recently published by the Meteorological Council. 
 
100 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 of high and low pressure areas of the atmosphere, and of the movement 
 of currents of the air under the influence of the barometric gradient, to 
 the Newtonian doctrine of gravitation in astronomical physics, to the 
 Daltonian hypothesis of atomic proportions in chymistry, to the dynamic 
 theory of the tides, and to Avogadro's law of the uniformity of the 
 atomic constitution of gases under like conditions of pressure and 
 temperature. 
 
 Within the last few years meteorology has been studied on a sounder 
 and more philosophical basis than formerly ; and, in that department of 
 the science bearing especially on temperature, moisture, and pressure, 
 together with the winds, research (or, in other words, systematized 
 investigation founded on carefully recorded observations) has taken the 
 place of pure conjecture based on imperfect data ; much has been learnt, 
 but much more remains unknown ; discussion and controversy are 
 equally rife at home and abroad, where the constant topics among 
 meteorologists are : hygrometric questions causes of barometrical de- 
 pressions the barometer and real pressure of the air relation between 
 pressure and velocity of wind relation of the velocity of the wind to 
 the gradients the effect of rainfall on the barometric pressure the 
 distribution of temperature in an ascending current of air the origin 
 and theory of storms in fact, whatever tends to improve our knowledge 
 of the physics of the atmosphere, with its vapour, together with the 
 mechanical effects of ascending, descending, and horizontally-flowing 
 currents of air, on which depends the theory of all atmospheric circulation. 
 The more systematic enquiry into the character of tropical storms, which 
 was some time ago the great desideratum, and which was ably initiated 
 and conducted by MELDRUM, BLANFORD, and WILLSON, has been very 
 efficiently pursued and carried out by the PADRE BENITO VINES, the 
 PERE DECHEVRENS, and E. KNIPPING (in Japan), so that more is now 
 known than formerly of these intermittent visitors in the Southern 
 Indian Ocean, the Bay of Bengal, the China Sea, and the West Indies ; 
 but it still remains certain that whatever is known of the cyclone as a 
 hurricane has been derived from extra- tropical sources ; that knowledge 
 however, so far as it goes, does not lead to any other inference than 
 that the cyclonic storms of all regions are very much alike though 
 those of the tropics may within a limited area be more persistently 
 severe and destructive. 
 
 The latest and most strenuous advocate of the purely circular theory 
 of storms is M. FAYE, whose essay appeared in the " Annuaire du 
 Bureau des Longitudes" for 1875 ; it was written to uphold the older 
 views in opposition to those of MELDRUM, as well as to controvert the 
 theory of some Continental authors who had lately maintained the 
 efficiency of ascending currents of air, and an upward motion, as the 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 101 
 
 first cause of a whirlwind. The essay endeavours to establish the 
 reasonableness of most of the theories of the elder cyclonologists, and 
 advances others of a later date showing that cyclones, and all similar 
 phenomena, are descending currents of air of a conical form, like the 
 whirlpools and eddies that form in water, and that their origin must be 
 sought for in the upper regions of the atmosphere; the progressive 
 motion is taken to be due to the upper part of the cyclone being dragged 
 along by the upper current in which it takes its rise ; and waterspouts 
 and whirlwinds (trombes) are considered to come into the same category 
 as cyclones and hurricanes. M. FAYE'S proofs being disputed by PESLIN 
 and COUSTE, and the invalidity of his arguments being pointed out, on 
 returning to a re-discussion of MELDRUM'S notice of the Mauritius hurri- 
 cane of February 1860, he endeavours to show that the storm was really 
 circular, but modified by the action of the S.E. trade- wind the circular 
 form being well developed in front, while the winds in the rear were 
 centripetal ; he at the same time affirms that observations made at sea 
 are not so accurate as desirable for the purpose, which is a rather good 
 joke, considering that the circular theory was mainly based on data 
 derived from ships' log-books. By various papers that have recently 
 appeared in the " Revue Maritime et Colonial e," and the " Annales 
 Hydrographiques" the views of French meteorologists and navigators do 
 not coincide with those of M. FAYE. 
 
 Prof. W. BLASIUS, in a work on the theory of storms, seems to 
 indicate that the storm area is a region of conflict between two opposing 
 currents of air where the equatorial and polar currents come into 
 collision. Discarding on the one 
 hand the rotary or cyclonic, and on 
 the other the centripetal, theory as 
 proposed respectively by REDFIELD 
 and ESPY, he asserts that storms 
 assume the form of an ellipse, in 
 which the winds blow in straight 
 lines from the circumference to- 
 wards the region of the uprising 
 equatorial current, and that the 
 barometrical depression is only a 
 secondary effect of the collision of 
 the two winds, and not the original 
 exciting cause of the storm. He 
 also divides storms into two classes North-East, or those of winter, 
 and South-East, or those of summer ; and shows that each is connected 
 with its own peculiar cloud. (Fig. 30 is the form of the S.E. storm 
 contrasted with a circular one.) After pointing out that the old rules 
 
102 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 of the circular theory are worse than useless, he propounds new rules 
 adapted to his two forms of storms. The work is that of an observer, 
 and the statements are given with much positiveness ; but the inferences 
 are not always correct, though several of the weak points in meteoro- 
 logical reasoning are well exposed ; and the author's remarks are chiefly 
 valuable as showing that the circular theory of storms requires re- 
 consideration. But the most valuable works on Meteorology and the 
 Law of Storms issued of late in the United States are Professor FERREL'S 
 paper on " Cyclones" in the U. S. Survey Report 1878, and Professor 
 LOOMIS' " Contributions to Meteorology" 
 
 Before passing on, it may be well to quote Mr. BUCHAN (the great 
 Scotch meteorologist) on the theory of cyclones : 
 
 " The spiral rotation, instead of the purely circular rotation, of the 
 winds in storms, completely alters the whole complexion of the question 
 of the theory of storms. For since it follows from it that enormous 
 quantities of air are constantly being poured all around into the area 
 of the storm, and since, notwithstanding these accessions tending to 
 increase the pressure, observation shows that the pressure is not thereby 
 increased, but on the contrary sometimes diminished, we are forced to 
 the conclusion, that, from a large area within and about the centre of the 
 storm a vast ascending current must arise into the upper regions of the 
 atmosphere ; and arriving there, must flow away over into neighbouring 
 regions. The physical cause of the ascending currents is to be found in 
 the moist and warm, and therefore light, air which all observation shows 
 to prevail in the front and in the central part of storms. And since 
 most of the rain which accompanies storms falls in those parts of the 
 storm, the barometer will be still further reduced by the removal of 
 the elastic aqueous vapour which is condensed into rain-drops, and by 
 the latent heat set free in the condensation of the vapour." 
 
 As a knowledge of the "law of storms" now forms part of the 
 examination of candidates for the master's and mate's certificate, it is of 
 importance that the parties most interested in this department of meteoro- 
 logy should be well posted in the latest views respecting cyclones, and 
 to this end an abstract is given of such parts of Professor FERREL'S 
 paper as may be most useful to the navigator. 
 
 Professor FERREL has investigated the subject from the twofold stand- 
 point of (1) theory, and (2) the practical applications of the theory and 
 comparisons with observations. In the first of these, embracing as it 
 does much abstruse mathematical investigation, the general reader and 
 the navigator will, unless well read in mathematics, take but little 
 
MODIFICATION OP THE CIRCULAR THEORY OP STORMS. 103 
 
 interest. It is sufficient, therefore, to say that in dealing with the 
 mechanical theory of cyclones the elements considered are the earth's 
 rotation and its effect on the movements of the atmosphere, the gyratory 
 velocity of the wind around the centre of depression, the friction, the 
 inertia, and the temperature and humidity of the air. The Professor 
 aptly concludes the theoretical investigation into the mechanics of 
 cyclones, by an illustration of the causes of their motion over the earth's 
 surface : 
 
 " So far we have regarded the cyclones as existing in an atmosphere 
 having no motions excepting those belonging to the cyclones. But on 
 account of the general motions of the atmosphere (treated of in the first 
 part of these researches), cyclones usually occur in an atmosphere having 
 a general motion independent of those of the cyclone itself. In order, 
 therefore, to have the true motions relative to the earth's surface, we 
 must take the resultant of the two motions. In the middle and higher 
 latitudes of the northern hemisphere the general or normal motion of 
 the air is from the west or south-west. In a cyclone in these latitudes 
 this motion of the air coincides with that of the southern side of the 
 cyclone, and increases the motion relative to the earth's surface, but on 
 the northern side it is in the contrary direction, and counteracts in some 
 measure that of the cyclone, or it may even entirely destroy it. On the 
 east or north-east side of the cyclone and its opposite, this general motion 
 from the west or south-west would have little effect upon the velocity 
 relative to the earth's surface, but would affect mostly the inclination, 
 that is, the angle between the tangent to the isobar and the direction of 
 the wind, decreasing it on the east or north-east side of the cyclone, and 
 increasing it on the west or south-west side. 
 
 " In the latitude of the trade-winds, where the general or normal 
 direction is from the east or north-east, the reverse takes place, so that 
 the velocities of the air belonging to the cyclone are diminished or, it 
 may be, entirely destroyed by the general or normal motion on the south 
 side of the cyclone, but increased on the north side. The effect upon 
 the inclination is in this case likewise reversed, being increased on the 
 east or rear side and decreased on the opposite side. 
 
 " But since the general motions of the atmosphere form two great 
 systems of cyclone and its corresponding anti-cyclone, all ordinary cyclones 
 are simply cyclones within a cyclone, so that all that has been stated 
 with regard to smaller cyclones within ordinary cyclones is applicable 
 here to ordinary cyclones with reference to the general motions of the 
 atmosphere. Hence the isobars of cyclones as laid down upon the 
 charts are the resultants generally of two or more systems of cyclones 
 contained within or encroaching upon one another ; and to have the isobars 
 belonging to any one system, it would be necessary to eliminate the effects 
 
104 MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 
 
 of the others. The isobars of all ordinary cyclones are therefore affected 
 generally by the isobars belonging to the general motions of the atmo- 
 sphere, as well as by the smaller class of cyclones which may be contained 
 within the ordinary ones. 
 
 "The progressive motion of a cyclone depends mostly upon the 
 general motions of the atmosphere, but also upon the tendency of the 
 cyclone to press towards the pole. In the trade-wind latitudes the wind 
 at the earth's surface is westward, or at least has a large westerly com- 
 ponent, and hence the cyclones in these latitudes are carried westward 
 by this westward motion of the air, especially at certain seasons, and, 
 having likewise a tendency toward the pole, the resultant of the two is 
 a westward motion, inclined a little toward the poles, or in the northern 
 hemisphere a motion about W.N.W. After having arrived at the 
 parallel of 30 or 35 in the tropical calm-belt, where there is no west- 
 ward motion, the progressive motion is a polar one mostly, but, after 
 progressing still nearer the pole, into the middle and higher latitudes, 
 the general eastward motion of the atmosphere here, which is great in 
 the upper regions, carries now the cyclone towards the east, and the 
 direction of the progressive motion, which is usually about E.N.E., is 
 the resultant of this eastward motion and the motion towards the pole. 
 All well-developed cyclones, therefore, having their origin near the 
 equator, have mostly a progressive motion represented by a curve some- 
 what in the form of a parabola, having its vortex in the tropical calm- 
 belt at the parallel of 30 or 35. 
 
 " Both the progressive velocities and directions, however, are very 
 different in individual cases, so that there must be other strong modifying 
 influences ; and perhaps among these the principal are the distribution 
 of the aqueous vapour and the positions of the general isothermal lines. 
 The equatorial side of a cyclone is generally warmest, and contains the 
 most aqueous vapour. As the air containing this vapour is carried 
 around to the east and ascends, it becomes colder both from its ascent 
 to higher altitudes, and from its being carried into higher latitudes, so 
 that the vapour which it contains is mostly condensed on the east or 
 north-east side of the cyclone. As the power of the cyclone is much 
 increased by the latent heat of the vapour given out by condensation, 
 this gives rise to a tendency to form a new centre of a cyclone continually 
 in advance of the old one, so that the progressive motion of the cyclone 
 is rather a continual forming of new cyclones, at least so far as the lower 
 part of the atmosphere is concerned, in the direction in which the vapour 
 of the cyclone is mostly condensed. Hence the velocity of the progres- 
 sive motion is generally much greater than that of the general motion 
 of the atmosphere below. The further around the vapour is carried 
 before it is condensed, the more it will incline the direction of the pro- 
 
MODIFICATION OF THE CIRCULAR THEORY OF STORMS. 105 
 
 gressive motion from the east toward the pole. If it should not be 
 carried around as far as the east before it is mostly condensed, it inclines 
 the progressive motion from the east toward the equator. If the iso- 
 therms should vary much from the parallels of latitude, this also might 
 have considerable effect, so that, if they extended from S.W. to N.E., 
 the progressive motion would be more inclined toward the poles than 
 they would be if they extended from N.W. to S.E. 
 
 " These views with regard to the effect of the unequal distribution of 
 vapour upon the direction of the progressive motions of cyclones were 
 first given by Rev. W. CLEMENT LEY." 
 
 The second part of the subject is easily understood, and of the highest 
 importance, corroborated as it is by the observations of MOHN, HILDE- 
 
 BRANDSSON, LOOMIS, CLEMENT LEY, MELDRUM, WlLLSON, BLANFORD, 
 
 TOYNBEE, the PADRE VINES of Havana, and others, whose investiga- 
 tions have been specially directed to the storms and cyclones of the 
 Atlantic, the Southern Indian Ocean, and the Bay of Bengal. 
 
 The particular feature to be noted is the undoubted incurvature of 
 the wind more or less according to the quadrant of the storm in which 
 the ship is found, and also according to the latitude. MELDRUM, as far 
 back as 1867, first drew attention to the disasters resulting to vessels, 
 through the direction of the centres of cyclones being estimated on the 
 circular theory; and in March 1878, with the wind N.E. at Mauritius, 
 it was found that he had correctly indicated the centre of the cyclone to 
 be to the W.S.W.-ward, whereas on the circular theory it should have 
 been to N.W.-ward. Captain TOYNBEE, from the discussion of the 
 North Atlantic storm of August 1873, inferred that the wind in a 
 hurricane draws in towards its centre : the average inclination of all the 
 observations was 29 about the parallel of 50 ; the indraft is probably 
 greater in one quarter than in another ; the indraft is probably greater 
 near the centre than further from it. CLEMENT LEY'S observations for 
 Europe make the mean inclination about 25 ; those of LOOMIS, for 
 America (nearer the equator), about 47 ; those of the PADRE VINES, 
 for the West Indies, about 45 ; and those of BLANFORD, for the Bay of 
 Bengal, in some cases 42. 
 
 Professor FERREL, on the same basis, remarks as follows : 
 " From the deductions of theory, confirmed now and supplemented 
 by numerous observations, it is evident that many of the usual rules and 
 sailing directions must be very much modified, especially in low lati- 
 tudes. Although the horn cards of PIDDINGTON, and all the rules based 
 upon the strictly circular theory of the winds, may still be used at sea 
 in high latitudes without great error, yet nearer the equator they must 
 become more erroneous, and entirely fail at the equator if cyclones could 
 
106 MODIFICATION OP THE CIRCULAR THEORY OF STORMS. 
 
 exist there. It is beginning to be pretty generally acknowledged that, 
 in sailing directions in a storm, some allowance should be made for a 
 certain small amount of inclining of the winds, but it seems to be 
 thought that this should be the same in all latitudes and at all distances 
 from the centre of the storm. If in low latitudes the inclination of the 
 winds may be 60 or more, the determination of the direction of the 
 dangerous centre of a cyclone from the usual rules based upon the 
 circular theory, in which the inclination is supposed to be naught, would 
 lead to an error of five or six points of the compass. For this reason 
 the amount of inclination at sea, under average circumstances, should be 
 determined from observation for different latitudes, so that the navigator 
 could modify his rules to suit the latitudes ; but unfortunately we have 
 but few observations yet for this purpose at sea, and very few on land 
 for the lower latitudes. Even on the same latitudes considerable allow- 
 ance should be made for distance from the centre. While the centre of 
 the cyclone is yet at a considerable distance and the winds have not a 
 great velocity, it should be considered that they are probably, even in 
 high latitudes, considerably inclined toward the centre, and that nearer 
 the equator they may be nearly radial ; but even in these latitudes at 
 places near the centre, where the velocities are very great, the gyrations 
 may become nearly circular. 
 
 " Where the cyclone has a progressive as well as cyclonic motion, 
 the former makes the inclinations differ on different sides. The pro- 
 gressive motion of the atmosphere at the earth's surface in the United 
 States is very small, only in general two or three miles, and even less in 
 places, and from the S.W. This, however, with Professor LOOMIS' small 
 average velocities of the wind, should increase the inclination very 
 sensibly in the west quadrant and decrease it in the east quadrant, 
 while the inclinations of the north and south quadrants should be little 
 affected, and have about the average value of all. This is the case with 
 his results, and hence they are in accordance with theory. But the pro- 
 gressive motion of the air in the northern part of the Atlantic Ocean 
 between Europe and America is comparatively great, as shown by the 
 closeness of the isobars ; and hence this motion must affect the inclina- 
 tions in cyclones there very much, especially in such parts of them as 
 have only moderate velocities of the wind, and from this cause the winds 
 should have smaller inclinations in the front than the rear part of the 
 cyclone. 
 
 " Mr. LEY obtained greater inclinations in Europe for the front than 
 the rear parts of the storms, while Professor LOOMIS for the United 
 States, and the PADRE VINES for the Antilles, obtained the reverse. This, 
 however, may be explained from the consideration that the cyclones in 
 Europe, especially those of the British Isles and Norway, are cyclones 
 
MODIFICATION OP THE CIRCULAR THEORY OP STORMS. 107 
 
 within a great cyclone with its centre near Iceland, and this is especially 
 the case during the winter season. They are consequently cyclones upon 
 an inclined plane, as it were ; and hence the isobars as usually charted 
 are very much distorted, and the centre of the cyclone as given by the 
 charted isobars is thrown to the left or N.W. of the real centre of the 
 cyclone. This decreases the angle with the radius on the front or N.E. 
 side of the storms and increases it on the rear or S.W. side, and con- 
 sequently the inclination is increased in the former and decreased in 
 the latter. 
 
 " The effect of the progressive motion of the air on the inclinations of 
 the winds in the cyclones of the Antilles is quite large, since the pro- 
 gressive motion of the trade-winds is comparatively great. Here the 
 trade- winds become nearly east winds. Hence the winds whose cyclonic 
 component of motion is from the N. or S. suffer the greatest deviation 
 from this direction, the inclination being decreased in the former case 
 and increased in the latter. 
 
 " By a reference to the citations (of the PADRE VINES of Havana), it 
 is seen that in every instance the winds in the anterior part of the 
 cyclone are said to be nearly circular, or to deviate but little from the 
 tangent, while in the posterior part, where the winds are mostly from 
 S.E. to S.W., the inclination is said to be very great, and the winds to 
 blow almost towards the centre. 
 
 " It is further stated with regard to these cyclones generally, ' that 
 the winds which have suffered least deviation with respect to the 
 tangent of the circle having for its centre the centre of the vortex of 
 the cyclone, have been those which have blown from the first and fourth 
 quadrants, and those deviating most with an inclination towards the 
 centre, those (which have blown) from the third, and especially the 
 second quadrant. In the winds from the E. and S. the inclination is 
 such sometimes at all distances from the vortex that the wind blows 
 directly towards it.' Here it is stated again that the winds which blew 
 from the first and fourth quadrants, which are the winds of the W. or 
 N.W. quadrant of the cyclone, and consequently the anterior part of it, 
 had the least deviation from the tangent, and that those winds which 
 blew from the third, and especially those from the second quadrant, and 
 which are necessarily winds of the posterior part of the cyclones, had the 
 greatest inclination, as they should. This inclination seems to have been 
 so great in winds from the S.E. that they blew directly towards the centre. 
 
 " The navigator, therefore, on determining the direction of the vortex 
 of a cyclone from the direction of the wind, should, in addition to 
 considering latitude, distance from centre, and velocity, likewise con- 
 sider in what quadrant of the cyclone he is situated, since the direction 
 of the vortex with reference to that of the wind is so different in 
 
108 MODIFICATION OP THE CIRCULAR THEORY OF STORMS. 
 
 different quadrants, especially where there is a large progressive motion, 
 as in the trade-wind regions. In the front part of a cyclone, where the 
 mariner is in the greatest danger, the direction of the vortex is generally 
 more nearly at right angles to the direction in which the wind blows to 
 the left in the northern hemisphere, and the contrary in the southern* 
 and consequently the old rules of the circular theory are more nearly 
 correct here than on any other side of the storm. We have seen that 
 this is the case in the front part of the cyclones of the Antilles, where 
 the direction of the vortex is nearly at right angles to the direction of 
 the wind, while in the rear it is nearly in the direction in which the 
 wind blows. On either side of the direction of progress, the direction 
 of the vortex in cyclones here may be estimated to be about 45 to the 
 left of the direction in which the wind blows." 
 
 In reference to the prognostics of an approaching cyclone, the Professor 
 dwells upon two which should particularly receive attention. 
 
 " The almost universal precursor of a distant cyclone is the appear- 
 ance of more cirrus clouds than usual, not only differing from those of 
 the general currents in form, but also in the direction of the currents 
 indicated by these clouds. The approach of cyclones from the E. or S.E. 
 is always indicated at Havana by the appearance of these clouds when 
 the vortex is yet 500 miles or more distant, and while fair weather is 
 yet prevailing. The same is also observed as they pass off at a distance. 
 Here the directions of the currents as indicated by the cirrus clouds 
 seem to be very nearly radial, as they should be by theory in these low 
 latitudes, and at great distances from the centre ; and hence the direction 
 from which they come indicates very nearly the direction of the vortex 
 of the cyclone. In these low latitudes the progressive motion of the 
 atmosphere above is small in the cyclone season, and does not interfere 
 much with the directions given by regular cyclonic motions." * * * 
 
 " With regard to the existence of an anti-cyclone in connection with 
 every cyclone, and the broad annulus of high barometer with its maxi- 
 mum at the dividing limit between the cyclone and anti-cyclone, there 
 are many confirmatory observations, but this maximum may be interfered 
 with and masked by numerous irregularities. On the ocean, where the 
 cyclonic systems are more regular, if we had a sufficient number of 
 accurate observations, the annulus of high barometer and the winds of 
 the anti-cyclone could be more readily distinguished from the other 
 irregularities, especially within the tropics, where these irregularities 
 are smaller. In the Antilles, where the islands being small in com- 
 parison with the expanse of ocean around them, the annulus of high 
 
 * This is according to BUYS BALLOT'S law, and with reference to the back to the wind; the seaman 
 is accustomed to face the wind, in which case the statement here made would read : "to the right in 
 the northern hemisphere, and the contrary (or left) in the southern," 
 
PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 109 
 
 barometer is readily observed, and the approach of the cyclonic centre 
 indicated several days in advance of its arrival by the rising of the 
 barometer above its normal level." The hurricane of September 1875 
 was thus indicated at Havana while the cyclone was yet 1200 miles 
 distant, and that of September 1876 was similarly shown by a rise at 
 Havana, while it was causing great destruction at Porto Rico. 
 
 The observations of Mr. C. MELDRUM (of the Mauritius Observatory) 
 on the " Form of Cyclones in the Southern Indian Ocean " have already 
 (pp. 71-92) been given at some length. It remains to briefly collect 
 the researches of other meteorologists who have discussed the subject on 
 the same basis. 
 
 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 (By RHO. P. BENITO VINES, S. J. OF HAVANA.) 
 
 FIRST INDICATIONS OF THE APPROACH OF A HURRICANE. 
 
 The existence of a hurricane at a distance is manifested ordinarily 
 by the passage of an anti-cyclone, the presence of which is revealed by 
 the following phenomena : 
 
 1. Extraordinary rise of the barometer. 
 
 2. Anti-cyclonic winds of some duration. 
 
 3. Dry, bracing, and beautiful weather, clear sky, and exceedingly 
 transparent atmosphere. 
 
 The greater or less rise of the barometer in a given locality during 
 the passage of an anti-cyclone depends not only on its force, but also on 
 the distance of its centre. 
 
 The anti-cyclonic winds are easily recognised by their coolness and 
 dryness, and by their force and persistency with a high barometer. As 
 a rule they dominate completely the land and sea breezes, and prevail 
 during the hours of calm incident to the change from one to the other. 
 Not unfrequently are they in complete opposition to these breezes, and 
 also to the indications of the barometer under ordinary conditions. At 
 Havana, for example, the barometer may rise during the day with a 
 southerly wind, and fall during the night with a northerly wind. 
 
 If the winds of the anti-cyclone rotate and diverge, the bearing of its 
 centre can be determined by the following simple rule : 
 
 If the observer stands with his back to the wind the centre of maximum 
 pressure lies to his right and rear, so that with easterly winds the centre 
 of the anti-cyclone bears between N. and E. ; with southerly winds, 
 between E. and S. ; with westerly winds, between S. and W* ; and with 
 northerly winds, between W. and N* 
 
110 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 
 It is to be observed that in general the anti-cyclonic winds change 
 direction very slowly, and sometimes blow for a long time from the 
 same point of the compass. This arises from the slow motion of the 
 body of the anti-cyclone, and its tendency to become fixed in certain 
 regions. 
 
 In respect to the weather experienced at Havana under the influ- 
 ence of anti-cyclones, the following exception has been observed : 
 When the centre of an anti-cyclone bears N.E. or N.W. from Havana, 
 and a cyclone belonging to the same system at the same time crosses 
 North America and the Atlantic between 45* and 50 north latitude, 
 its presence is sometimes indicated by rainy weather ; this frequently 
 happens with Northers and other strong winds. Considering the 
 position of the anti-cyclone in this case, we may infer that it is supplied 
 from above from the upper currents of the cyclone and from the warm 
 and humid anti-trade winds. Anti-cyclones of this character do not 
 usually precede a hurricane. 
 
 As the centre of maximum pressure draws away from the observer, 
 and he finds himself within the influence of the hurricane, the barometer, 
 which has already dropped from its extraordinary height, goes on falling 
 slowly. The upper regions get less clear, and a veil of cirrus clouds 
 commences to form. This thickens gradually, and is the cause of the 
 solar and lunar halos. At the time of the rising and setting of the sun 
 the clouds are remarkable for their dark-red and violet tints. The 
 peculiar cirro-stratus clouds,* the invariable precursors of the hurri- 
 cane, soon appear ; the weather becomes heavy, the heat oppressive and 
 sticky, causing profuse perspiration, and the humidity of the air 
 increases. In a short time the cloud-bank of the hurricane appears on 
 the horizon, the wind freshens every moment, and the first nimbus and 
 broken cumulus clouds commence to disintegrate and fly about with 
 gusts, light rains, and passing squalls. These increase in number and 
 intensity, with the more rapid fall of the barometer, as the storm-centre 
 approaches. Electrical discharges are rarely seen at this period, and 
 their absence is a very bad sign. 
 
 BEARING OF THE STORM-CENTRE. 
 
 For an observer who has to rely solely on his own personal 
 observations, as is the case with seamen, it is of the greatest importance 
 to be able to determine with the closest possible approximation the 
 bearing of the centre at different times and during different phases of 
 the hurricane. 
 
 * The cirro-stratus clouds, called by the author cirro-stratus plumtformes, appear like white and 
 delicate feathers or like great and showy plumes crossing the firmament. These beautiful clouds 
 remain fixed at times, and do not change their forms for hours. 
 
PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. Ill 
 
 The following indications deduced from the appearance and dis- 
 position of the clouds and from the direction of the air-currents will be 
 of great utility : 
 
 1. As soon as the upper region commences to cloud over, the cirrus 
 veil appears most dense in a particular part of the horizon where a whitish 
 arc is formed, which, when the sun rises and sets, changes to an intense 
 dark-red. This cirrus arc forms part of a misty aureole which surrounds 
 the hurricane, and is consequently the first indication which enables us 
 to fix its bearing. 
 
 2. It has been observed that the cirro-stratus clouds (before men- 
 tioned) are so situated that their focus of radiation or divergence nearly 
 corresponds to the bearing of the centre. This is considered a very good 
 indication, and is fortified by numerous observations. 
 
 To readily determine the position of this focus, it is only necessary 
 to observe that when several of these cirro-stratus clouds diverge 
 materially, and consequently are at some distance from the zenith, we 
 must suppose them prolonged to the point of meeting. This should also 
 be done when the cirro-stratus clouds show a slight curvature due to 
 rotation, which has been observed in a few instances. (It is very re- 
 markable, and entirely in agreement with the theory adopted by the 
 author, that the few times this phenomenon has been observed it has 
 been seen that the cirro-stratus clouds participate in the movement of 
 rotation, and the whole body of clouds moves slowly round the focus of 
 divergence, like the motion of the spokes of a wheel round its fixed axis.) 
 
 3. The cloud-bank of the hurricane presents itself above the horizon 
 so that its centre nearly corresponds to the bearing of the storm-centre. 
 
 In order to distinguish it easily, and not confound it with the squall 
 clouds, particularly when the horizon is indistinguishable, it is necessary 
 to note the following indications : 
 
 The cloud-bank of the hurricane is distinguished from any other cloud 
 by its appearance, by its relative fixity of position, by the movements of 
 the squalls, and by the direction which the scud takes in relation to it. 
 
 Its appearance is that of a cumulo-stratus cloud, which is formed in 
 its upper part by a combination of rounded and cone-shaped clouds, while 
 its lower part consists of a very dark nimbus cloud, whose base cannot 
 be seen. The nimb\is of the squall has no particular form. Its base, 
 however, is sometimes well defined, forming a dark band above the 
 horizon. 
 
 As to its relative fixity of position, it is evident that the cloud-bank 
 of the hurricane does not move readily from one point of the horizon to 
 another, but remains for a long time in one position, or moves very 
 slowly, particularly during its first phases. The clouds of a squall, on the 
 
112 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 
 contrary, appear in different parts of the horizon, and change position 
 frequently. 
 
 With respect to the movements of the squalls, it is seen that these at 
 first disengage themselves from the cloud-bank, and then diverge. The 
 cumulus clouds, which are first seen near the middle of the cloud-bank, 
 assume gigantic proportions as they rise towards the zenith, gradually 
 spreading out and covering the sky. Very soon appears the base of the 
 cloud, forming above the horizon the dark band characteristic of the 
 squall. With this cloud comes the rain, when the wind freshens and 
 veers immediately to the right, blowing sometimes almost directly from 
 the cloud-bank, which, when the squall has passed, is seen in the same 
 point of the horizon as before. Later, the squalls arise from one ex- 
 treme of the cloud-bank, and follow more or less closely the general move- 
 ment of rotation. 
 
 Finally, if the direction of the scud above the cloud-bank is noticed, 
 it will be seen that it flies parallel to it as it rotates, so that an observer 
 who is looking at the cloud-bank sees the clouds fly always from left to 
 right in horizontal lines ; thus, if the cloud-bank bears south, the clouds 
 above it fly horizontally from east to west. This does not happen when 
 the observer looks at any other point of the horizon where there may be 
 a heavy cloud, as he will soon see that the scud does not fly in horizon- 
 tal lines, nor does not move from left to right relatively to himself. 
 
 4. As a rule, the lower air-currents converge, forming with the bear- 
 ing of the storm-centre a variable angle, which is almost always greater 
 than a right angle. For this reason the law of storms cannot be applied 
 to them in all its simplicity without grave error. It is necessary to 
 record here that the convergence of the winds is much greater ordinarily 
 at the rear tJmn at the front of the storm, where it is sometimes entirely 
 inappreciable. 
 
 5. The low clouds in the interior of the hurricane fly ordinarily in 
 directions nearly perpendicular to the bearing of the centre, and conse- 
 quently, applying the law of storms to the direction of these clouds, a 
 much closer approximation will be obtained in the determination of the 
 bearing of the centre than if the same law be applied to the direction of 
 the winds on the outer edge of the hurricane. The same may be said of 
 the violent gusts of the squalls if the wind veers always to the right, and 
 its convergence is lessened or done away with altogether. 
 
 6. The cumulus, cirro-cumulus, and cirrus clouds that precede the 
 hurricane generally diverge, that is to say, their direction forms with 
 the bearing of the centre an angle less than eight points, with the very 
 noticeable peculiarity, that if different strata are observed it will be seen 
 that the divergence increases with the elevation. 
 
PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 113 
 
 7. The gusts of the first squalls that break away from the cloud-bank 
 diverge. In order to determine, therefore, the bearing of the centre by 
 taking advantage of the direction of the air-currents, provided the ob- 
 server is not in the outer edge of the hurricane, the following simple rule 
 may be applied : 
 
 If the observer places himself with his back to the wind, the centre 
 lies to his left and, in general, a little to the front ; if the same supposi- 
 tion is made with respect to the direction of the nimbus clouds and the 
 squall gusts, the centre lies approximately to his left. 
 
 In the table that follows will be found expressed the bearings of the 
 centre, corresponding to the eight principal points as they are deduced 
 from the application of the law of storms to the direction of the nimbus 
 clouds and squalls. By means of these general cases, the bearing corre- 
 sponding to intermediate points can be easily obtained. 
 
 
 
 If the nimbus 
 clouds and 
 squalls move 
 
 The centre will 
 bear 
 
 NORTH. 
 
 EAST. 
 
 N.E. 
 
 S.E. 
 
 EAST. 
 
 SOUTH. 
 
 S.E. 
 
 S.W. 
 
 SOUTH. 
 
 WEST. 
 
 S.W. 
 
 N.W. 
 
 WEST. 
 
 NORTH. 
 
 N.W. 
 
 N.E. 
 
 MEAN DIRECTION OF THE TRACK OF THE HURRICANE. 
 
 If at one point where the hurricane has touched complete observa- 
 tions are obtained, the mean direction of the track can, in general, be 
 determined with a sufficient degree of approximation by the following 
 method : Choose two equal heights of the barometer before and after 
 the barometric minimum, and, after having determined the two bearings 
 of the centre corresponding to these heights, measure on a chart from the 
 point of observation two equal distances in the direction of the given 
 bearings ; the line that unites the extremes of these two distances gives 
 the mean direction of the track, but not its true position, nor its nearest 
 point to the observer. This last can be deduced from its angular velocity, 
 or rather from the rapidity with which the wind shifts, if the speed of 
 the body of the storm is known beforehand by observation at other 
 points. In general it can be known if the centre has passed very close, 
 if while the barometer is at its minimum the wind shifts rapidly so that 
 in a few hours it changes from eight to twelve points ; and this may also 
 
114 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 
 be the case when the barometer has not fallen much and the gusts have 
 not been very violent. 
 
 POSITION OF THE OBSERVER RELATIVE TO THE STORM-TRACK. 
 
 As long as the bearing of the centre remains the same for some time, 
 although the barometer falls rapidly and the wind increases in strength, 
 the observer is directly in the track of the centre of the hurricane, or 
 very close to it. If the bearing of the centre moves to his right, in 
 which case the wind will also shift to the right, through K, E., S., and 
 W., the observer is to the right of the track. When the wind shifts to 
 the left and the bearing of the centre is changed in the same direction, 
 the observer is to the left of the track. 
 
 DANGEROUS AND MANAGEABLE SEMICIRCLES. 
 
 The track divides the area of the storm into two parts, of which that 
 to the right is called by seamen the dangerous semicircle, and that to the 
 left the manageable semicircle. This relates to the hurricanes of the 
 northern hemisphere. This classification depends on the greater or less 
 danger usually encountered, and on the greater or less effort required to 
 struggle against the storm and to withdraw from its centre, according as 
 the vessel is on one side or the other of the track. In the dangerous semi- 
 circle the gyratory winds tend to force the ship towards the front of the 
 storm, and consequently to throw her in the path of the centre. It is 
 necessary, therefore, to employ the maximum of effort to resist the 
 violent force of the wind, and to prevent being overwhelmed by the 
 hurricane. 
 
 The manoeuvre laid down in this case is to lie-to on the starboard 
 tack, and, as the wind shifts, which will always be towards the right, to 
 head more and more towards the same side, whenever the force of the 
 wind will permit, until the hurricane disappears in a direction between 
 West and North, or between North and East, according as the observer 
 is in the first or second branch of the track.* In the manageable semi- 
 circle the winds of the hurricane can be utilized in order to run away 
 from the centre. The manoeuvre that is usually practised in this case 
 is to run with the wind on the starboard quarter. 
 
 POSITION OF THE STORM IN ITS TRACK. 
 
 It is very necessary to know the position of the storm in its track, 
 because on this depends, in general, its direction, its velocity of trans- 
 
 * As the tropical cyclones of the northern hemisphere first travel in a general north-westerly 
 direction, then curve abruptly and advance to the north-eastward, it will be seen that their tracks 
 consist of two parts or branches : the first branch is that part of the track before the curve, the 
 second branch that portion of the track after the curve. 
 
PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 115 
 
 lation, or the movement of the storm body, its intensity, and the magni- 
 tude of its diameter. East of the island of Santo Domingo, and between 
 10 and 20 N. latitude, it is certain that the hurricane is in the first branch 
 of its track, and sweeps generally towards the fourth quadrant between 
 N.W. and West with increasing velocity, which sometimes reaches 15 to 
 20 miles an hour, is of small diameter, and generally of great violence. 
 In latitudes above 35 the hurricane is generally in the second branch of 
 the track, and sweeps to the N.E. with increasing velocity, at times at- 
 taining a speed of from 30 to 35 miles an hour, is of great diameter, and, 
 in general, has lost the violent character of tropical hurricanes. 
 
 In intermediate latitudes, between longitude 65 and 80 W., and 
 very particularly in the island of Cuba, every possible case may occur. 
 At Havana, for example, a hurricane may come from East, from E.S.E., 
 or from S.E. by E., in the first branch of its track with variable velocity, 
 according as it is near or far from the curve of its track ; from the S.E. 
 or from the South, moving very slowly at the curve in its track ; finally 
 from S.S.W. and from S.W., with increasing velocity in the second 
 branch. When the hurricane presents itself from the East, E.S.E., or 
 S.E. by E., with winds from North to N.E., as is most frequently the 
 case, it is still in the first branch of its track. It will be known if it 
 is about to curve or not if its phases succeed each other slowly or rapidly, 
 or if its movements are retarded or accelerated. If it first appears from 
 S.E., S.S.E., or South, it is much to be feared that it is already curving 
 and about to cross the island (Cuba). This probability becomes a cer- 
 tainty from the moment that it is seen that the bearing of the centre 
 varies very little and the phenomena succeed each other slowly. Finally, 
 when the hurricane comes from S.S.W. or from S.W. it is almost certain 
 that it is in the second branch of its track or just entering it. In this 
 case there is no time to lose, since the phases of these hurricanes succeed 
 each other with great rapidity. 
 
 GENERAL TRACKS OP THE CYCLONES IN THE DIFFERENT MONTHS. 
 
 At the beginning of the season some of the hurricanes of June and 
 July keep well to the south and cross the Caribbean Sea, and at times 
 also the island of Cuba, in a W. by N. direction approximately. From 
 the beginning of August to the end of September the tropical cyclones 
 of the northern hemisphere in the first part of their track take generally 
 a W.N.W. direction or a course between W.N.W. and N.W., and all of 
 them curve outside of the tropics, generally between 27 and 33 latitude. 
 
 As the season advances the initial direction of the track inclines 
 more to the westward, the hurricanes keep further south and curve in 
 lower latitudes, so that those of October, and even some at the end of 
 
116 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 
 September, come to us from the southern part of the Caribbean Sea and 
 curve in the vicinity of the Tropic of Cancer, or before reaching it, with 
 this peculiarity, that where several have come in succession each one has 
 curved not only in a lower latitude, but also further west than the pre- 
 ceding. The interval in these cases from one to another is not less than 
 twelve or fourteen days. These hurricanes are without doubt the most 
 to be feared for the western portion of the island. 
 
 The gales of November, December and January, which cross the 
 United States in a N.E. to an E.N.E. direction, and which are manifested 
 at Havana by a Norther due to the anti-cyclone which follows them, 
 may be considered as being in the second branch of the track. 
 
 PRACTICAL HINTS FOR THE BENEFIT OF VESSELS LEAVING THE PORT 
 OF HAVANA OR NAVIGATING IN THE WEST INDIES DURING THE 
 HURRICANE SEASON. 
 
 When there are indications of a hurricane to windward 
 
 1. Navigation from the port of Havana to the eastward is very 
 dangerous in the months of August and September. It is not so much 
 so in October, however, as the hurricanes then come in lower latitudes. 
 Nevertheless, in this last case a sailing vessel should wait until the 
 hurricane has passed into the third quadrant or bears somewhere 
 between South and West, in order to take advantage of the winds from 
 the S.E. shifting to S. and S.W. 
 
 2. Navigation to the northward is also very dangerous in this case, 
 as a sad experience has demonstrated. No sailing ship should try it ; a 
 steamer can undertake this voyage if she is sure to have time to get out 
 of the Strait of Florida and away from the coast before the storm can 
 reach her. If the captain resolves to do this he should go at full speed, 
 to get clear of the Strait and as far away as possible from the coast of 
 Florida and the Gulf Stream in order to have plenty of sea-room. He 
 should do this from the moment he observes the first indications of the 
 hurricane, which may possibly be close upon him. This is without doubt 
 one of the regions of greatest danger, not only on account of the many 
 storms which cross it with various directions and velocities, but also 
 because in the bight between Cape Canaveral and Cape Fear there is 
 scarcely a manoeuvre possible which is not dangerous. 
 
 In this part of the ocean it is preferable for a vessel to be in the 
 dangerous semicircle of the storm, because at least there is always left 
 the expedient of getting out by lying-to with the ship's head to the N.E., 
 East, or S.E., as the occasion demands, while in the manageable semicircle 
 the ship is squeezed between the track and the coast without space to 
 
PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 117 
 
 run. Innumerable are the ships which in the last few years have been 
 wrecked on this coast. If the steamer is going to Spain, by making an 
 easterly course south of the Bermudas after leaving the Strait she soon 
 leaves the zone where the storms occur. She will probably have head 
 winds at first, which will somewhat retard her voyage, and which would 
 prevent a sailing vessel from following the course indicated. 
 
 3. Navigation to the westward in the case supposed is scarcely 
 attended with any danger whatever, on account of the few hurricanes 
 which cross the Gulf of Mexico on the first branch of their track, and 
 because in case of curving well to the southward they do not enter the 
 Gulf very far. They give plenty of time, so that captains, being fore- 
 warned and having enough sea-room, can easily keep clear of the hurri- 
 cane in case of necessity. 
 
 If the hurricane presents itself from the S.S.E. to South or S.W., or 
 with Easterly winds shifting to the Southward 
 
 1. Navigation to the eastward presents no danger. 
 
 2. The voyage to Yera Cruz and New Orleans may be very danger- 
 ous, and the same may be said of navigation to the northward, for these 
 hurricanes cross the eastern portion of the Gulf and the Southern States 
 with great velocity after curving, and frequently surprise ships which 
 are sailing westward and reach those which a short time before left for 
 the north. 
 
 When the hurricane is to the north of Havana or bears between West 
 and North 
 
 1. Navigation to the eastward is not only without danger, but also 
 has the advantage of favourable winds. 
 
 2. Navigation to the north or to the west is also without danger if 
 the hurricane has already curved, because then the storm travels faster 
 than a ship, and outstrips her rapidly. But if the hurricane is curving, 
 although navigation to the westward is without danger in some cases, 
 it will always be disadvantageous, because it is necessary to struggle 
 against head winds and heavy seas, wasting time, consuming coal use- 
 lessly, and straining the vessel without advancing much. To start north 
 this case presents no danger, but there may be danger in continuing in 
 that direction. At all events the case is very simple ; and if the captain 
 has not intentionally put himself into the storm or does not get in front 
 of it, it is very certain that it will not seek him. It is evident, then, 
 that he should not venture to pass the Strait in the storm and leave 
 the centre to the west or bearing between West and South, but rather 
 wait until it has finished curving and disappears to the northward or 
 between North and East, 
 
118 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 
 Captains of vessels that make the voyage from Spain to Cuba by 
 way of Puerto Rico from July to the end of October, on entering the 
 zone of storms should keep a good look-out for the first indications of the 
 hurricane, in order not to be surprised and perhaps overwhelmed. This 
 refers to storms in their first branch, distance from their origin and from 
 the curve of their track, and which are violent, of short radius, and ad- 
 vance with great rapidity. If the first indications show that the bearing 
 of the centre is South or S.W., by laying the ship to or slowing down, 
 the storm will withdraw to the westward or between West and North. 
 If it appears between S.E. and S. it will not be prudent to go on, or at least 
 it will be necessary to use much caution, moderating the speed as long 
 as the storm does not move off to the S. W. If the first indications show 
 the centre to bear S.E. or S.E. by E., the ship is in a much more critical 
 situation perhaps in the track itself, and not very distant from the 
 centre in the dangerous semicircle. She should be laid-to immediately 
 and allowed to reach N.E. as much as possible, utilizing for this purpose 
 the diverging gusts of the first squalls even before the cyclonic winds of 
 the body of the storm are well established. If during this manoeuvre it 
 is seen that the centre moves towards the south, keep on lying-to always 
 on the starboard tack, luffing as the wind shifts to the right until it 
 blows from the third or fourth quadrant. If, having laid his ship to at 
 the first indication of the storm, and it veers towards the first quadrant, 
 the captain should see that the centre maintains the same bearing, that 
 the barometer falls rapidly, and the wind increases in strength every 
 moment, then the ship is in the track of the centre, or very close to it. 
 Only in this extreme case, and when the captain is very certain of his 
 observations, is it deemed prudent to bear up and risk running before 
 the storm in order not to be overwhelmed. In performing this 
 manoeuvre, taking into consideration the convergence of the superficial 
 or outer currents of the cyclone, it is considered better judgment not to 
 run directly before the wind, but to keep it always on the starboard 
 quarter. 
 
 If the centre is first seen bearing E.S.E., or between East and E.S.E., 
 it is probable that the ship is either in the track itself or in the manage- 
 able semicircle. To lie-to in this case would be a loss of precious time. 
 It would be best, therefore, to run without losing a single moment, on a 
 course between South and West, thus giving a chance for the hurricane 
 to withdraw in the direction of the first or fourth quadrant, or between 
 East and West through North. 
 
 Finally, if the hurricane bears East or E.N.E., run on a course be- 
 tween South and West, providing the wind freshens and the barometer 
 falls. This case is almost free from danger, and presents no difficulties 
 in manoeuvring, as long as the ship is well away from the coast. 
 
PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 119 
 
 Ships that navigate from South America to the port of Havana in 
 August and September should go south of the island (Cuba). They 
 should enter the Caribbean Sea in the vicinity of the island of Trinidad 
 and make their longitude without going much to the northward before 
 reaching the meridian of Cape San Antonio. If while crossing the 
 Caribbean Sea they should have indications of a hurricane to the east- 
 ward, they should run immediately to the southward as far as possible. 
 From the end of September to the first of November, on the contrary, 
 the navigation of the Caribbean Sea is very dangerous, particularly in 
 the vicinity of Cape San Antonio, while the voyage to the north of the 
 island is less dangerous than during the preceding months. At this 
 time (during October) ships should go well to the eastward of the Wind- 
 ward Islands and to the north of Puerto Rico and endeavour to make 
 their latitude as soon as possible as far as the parallel of 20 between 
 the meridians of 40 to 50. Considering the track that hurricanes 
 generally take in October, steamers that start for Spain during this 
 month will go safer by the way of Puerto Rico than to the northward. 
 Steamers that navigate from Cape San Antonio to Havana in October, 
 as soon as indications of a hurricane are observed, strong gales, low baro- 
 meter, ugly appearances, etc., if they are near Cape San Antonio, will 
 do well to run for the Bank of Yucatan and wait there, or else run out of 
 their course to the westward until the storm has passed, and take advan- 
 tage of the favourable winds which follow it to continue their voyage. 
 
 If on the first indications of the hurricane the ship finds herself be- 
 tween Cape San Antonio and Havana, the first thing to do is to get clear 
 of the coast and the currents ; when this is done, if the centre bears be- 
 tween E. and S., which is the most probable, continue running with the 
 gale to W.S.W., and when the storm has passed to the N.E. the winds 
 which follow it can be utilized to continue the voyage. If, when well away 
 from the coast, the centre bears in the third quadrant, or between South 
 and "West, lie-to for some time in order to observe if the centre remains 
 on the same bearing or veers towards the second quadrant. If either is 
 the case, run immediately with the wind on the starboard quarter j but 
 if it is seen that it veers to the westward, the wind shifting to the right, 
 lie-to with the head to the eastward. This last case is exceedingly 
 dangerous, and navigation in these waters should be avoided as much as 
 possible during the month of October. 
 
 Of the many cases explained in these pages, there have been, during 
 the last few years, examples with varying results. 
 
 Knowing that occasions may again arise when these observations will 
 be of use, they are recorded here in the hope that they may prove of 
 benefit in some emergency. 
 
120 PRACTICAL HINTS IN REGARD TO WEST INDIAN HURRICANES. 
 
 The danger of navigating these waters is great, the difficulties many 
 and often insuperable, the losses and shipwrecks all too frequent. 
 
 If by means of these incomplete notes a single disaster should be 
 averted, they will not have been written in vain. 
 
 The " Practical Hints in regard to the West Indian Hurricanes " 
 here introduced (pp. 109-120) have been translated from the Spanish 
 by Lieut. G. L. DYER, U.S.N., and are issued, "to lessen the dangers 
 of navigation," by the United States Hydrographic Office. The small 
 but valuable work of the PADRE BENITO VINES (Director of the 
 Magnetical and Meteorological Observatory of the Royal College of 
 Belen, in Havana), from which the translation has been compiled, is 
 entitled " Apuntes relatives a los Huracanes de las Antillas en Setiembre 
 y Octubre de 1875 y 1876." To this work Professor FERREL, in his 
 paper on Cyclones, frequently refers, and the following extracts are 
 especially worthy of note : 
 
 From observations of the PADRE VINES on the hurricanes of the 
 Antilles during September and October of 1875 and 1876, we may 
 deduce some valuable results with regard to the inclination of the winds 
 in cyclones in those low latitudes. Although we do not have in them 
 the averages of many observations, yet we can rely on them with con- 
 siderable certainty, since the cyclones there, being nearly the same as 
 oceanic cyclones, are more regular and not so much affected by irregular 
 abnormal disturbing influences, which require large numbers of obser- 
 vations for their elimination. The following extracts from the Padre's 
 book, some in substance merely, but mostly somewhat literal transla- 
 tions, have a bearing on the general subject. In all the hurricanes of 
 the Antilles it was observed "that the gyrating winds cease to be 
 circular at a long distance from the vortex, and are found to deviate 
 from the tangent to the circle with an inclination toward the centre, 
 forming a kind of large converging spiral." This converging is likewise 
 said to " vary not only in different hurricanes, but likewise in the same 
 hurricane, with different directions and intensities of the wind, and 
 with different distances from the vortex." In the same connection it is 
 stated that the inclination is " especially small at no great distance from 
 the vortex." 
 
 In the hurricane of September 1875, "the winds of the anterior 
 part of the storm were approximately circular, or with a slight inclina- 
 tion toward the centre in some cases." But from observations made 
 at numerous places, "the winds of the second (S.E.) quadrant, which 
 remained at all these places when the vortex was at a considerable 
 
INCURVATURE OP THE WIND IN WEST INDIAN HURRICANES. 121 
 
 distance, suffered a great deviation toward the centre, and in some cases 
 likewise the winds not far from the vortex." 
 
 Similar observations were made of the hurricanes of 1876. In the 
 anterior part of the storm the winds deviated but little from the tangent 
 toward the centre, but in the posterior part they blew almost directly 
 toward the centre. On the island of Porto Rico "little deviation in 
 the winds of the third and fourth (S.W. and NYW.) quadrants was 
 noted, some greater convergency in those of the first (N.E.) quadrant, 
 and a great inclination toward the centre in those from the E. and S., 
 especially as they became at a great distance from the vortex." 
 
 The hurricane of the 19th of October 1876 is presented as a notable 
 example of the great convergency of the winds. " After its passage by 
 Havana, the winds, which in the posterior part blew from West to South, 
 suffered a great deviation towards the centre, and that not only at a 
 distance from the vortex, but even in its vicinity." And " the winds 
 which prevailed from S.S.E. to S., and which, during the passage of the 
 vortex by Havana, blew with force in the different towns situated to 
 the E.S.E. of Havana, suffered likewise a very notable inclination 
 towards the centre. With respect to the winds which prevailed in the 
 first quadrant in the different localities of the anterior part of the 
 cyclone, there was observed in them likewise a notable convergency, 
 though in general in a less degree." 
 
 From the preceding extracts, and many others of the same kind 
 which might be cited, it is evident that there is not only an inclination 
 of the winds in the cyclones of the Antilles, but that this inclination is 
 much greater than in high latitudes. This latter deduction from theory, 
 if true, is very important from practical considerations ; and although 
 there cannot be any doubt with regard to it where the theory is under- 
 stood, yet any observations which tend to confirm it must be regarded 
 as being very important, and many more observations of the same kind 
 are very desirable. In the extreme cases namely, in front and rear of 
 the storm the angle of inclination in the former is stated to be very 
 small, and the direction of the wind to be nearly at right angles to the 
 radius of the cyclone, and in the latter to be so large that the wind blew 
 almost directly toward the centre. On other sides of the cyclone the 
 observations seem to give about a mean between these extremes. The 
 angle of inclination, therefore, in the cyclones of the Antilles, and at 
 all places at sea in the same latitudes, is probably about 45 on the 
 average of all cases of great velocities of the wind, and still much 
 greater where the velocities are not very great. But we do not have 
 observations enough yet for determining quantitative results with much 
 accuracy. 
 
122 REMARKS ON THE CYCLONES OF THE BAY OF BENGAL. 
 
 There are many evidences in the observations of the hurricanes of 
 the Antilles to show that the inclination near the vortex or centre of 
 the cyclone is less than at considerable distances. It is not only so 
 stated, as in the cases cited, but it may also be inferred from many 
 other places where, in referring to the great inclinations in special cases, 
 the phrase is often added, " even near the vortex," as though it was con- 
 sidered that the inclination there is usually less than at greater distances. 
 
 These observations are valuable as confirming those of MELDRUM in 
 the Southern Indian Ocean, with regard to the great incurvature of the 
 winds in the rear of the cyclone. 
 
 REMARKS ON THE CYCLONES OF THE BAY OF BENGAL. 
 
 The observations of Mr. W. G. WILLSON on the Balasore (June 1872) 
 cyclone elucidate the more recent views respecting cyclones, and his 
 general remarks are well deserving the attention of those navigating the 
 Bay of Bengal during the cyclone season. 
 
 Of the cyclone itself Mr. WILLSON says : 
 
 "It was not one of great violence, and was unusually small in 
 dimensions its radius, probably, never exceeding forty or fifty miles ; 
 the vortex, it is calculated, was formed on the night of June the 29th 
 or the morning of June the 30th, at or about lat. 19 50' N., and long. 
 89 E. For some time after its formation, it seems to have been almost 
 stationary. It moved afterwards with a velocity of about ten miles 
 an hour, as it passed the Pilot Ridge, between 8 P.M. and midnight of 
 the 30th. It advanced in a northerly direction very slowly from 
 midnight, retarded by friction with the land, turning round to the 
 westward along a path with less resistance, with probably a smaller 
 radius. It increased in extent and velocity of propagation as it moved 
 over a freer course towards Balasore. 
 
 The approximate track of the centre, as derived from land observa- 
 tions at Akyab, Chittagong, Calcutta, Saugor Island, and Madras, and 
 from the logs and reports of no less than 26 vessels (including the light- 
 ships), appears to have been as follows : 
 
 June 30, 10 P.M Lat. 20 W N....Long. 87 55' E. 
 
 Midnight..., 20 55',, ,, 87 50' 
 
 July 1, 3 A.M 
 
 ,, 8'30 A.M. ... 
 
 11 A.M. . 
 
 21 8' 87 45' 
 21 9' 87 10' 
 21 7' 86 45' 
 
 "It is probable that, previous to the formation of this vortex, 
 another distinct vortex had been formed on the 28th, in about the same 
 place viz. lat. 19 50' N., long. 89 E. ; but it appears to have been 
 almost stationary, and to have broken up on the morning of the 29th," 
 
REMARKS ON THE CYCLONES OF THE BAY OF BENGAL. 123 
 
 This cyclone cannot have travelled over much more than 300 miles ; 
 neither does it appear to have had any very remarkable speciality, except 
 the very high cross sea which accompanied it a sea, reported " never 
 to have been equalled in the Bay of Bengal." "With all the vessels, the 
 gusts of wind are described as something fearful. 
 
 Experience of former cyclones, in May and June, was verified by 
 that of 1872, and therefore the following indications of the probable 
 formation of a cyclone, in the north of the Bay of Bengal, during those 
 months, hold good : 
 
 " 1. The barometer falls steadily for some days round the north of 
 the Bay. The fall occurs first to the south-east, at Chittagong and 
 Akyab, and afterwards at Saugor Island, Cuttack, and Calcutta. The 
 barometer at Saugor Island ranges lower than the barometer at Calcutta 
 (normally at this season it ranges higher). 
 
 "2. At Calcutta the winds become light and variable from S.E. 
 round to N.E. As the disturbance increases, masses of cloud drift from 
 E.N.E. or N.E., and, as the storm approaches, showers are frequent, and 
 the wind blows in gusts. 
 
 " 3. At Saugor Island the winds are at first light and variable, 
 working round from S.E. to N.E. As the disturbance increases, and 
 as the storm approaches, the barometer continues to fall steadily ; the 
 wind increases in force, with squalls from the N.E., and masses of 
 clouds float rapidly from the same direction. Outside there is a heavy 
 southerly swell. 
 
 "4. At Cuttack, or False Point, the winds are first light and 
 variable from E.N.E., working generally round, through North to West, 
 and W.S.W. As the storm approaches, the wind increases in force, 
 with squalls from West and W.N.W. 
 
 " 5. At Chittagong the winds are light and very variable, mostly 
 from S.E. and N.E. 
 
 " 6. At Akyab the winds are light at first, and saturated with 
 moisture, mostly from South and S.E. As the disturbance increases, 
 the wind increases in force, and works gradually round from S.E. to 
 South and S.S.W. 
 
 "7. In the Bay, along and south of a line drawn from about False 
 Point to Cheduba Island, south-westerly gales, accompanied with torrents 
 of rain, prevail. Standing northward, the barometer falls, and rises 
 when standing southward." 
 
 The inference drawn from the above data is that 
 
 " Under the circumstances above specified (3), it appears unwise for 
 ships, in a good anchorage at Saugor, to put to sea. The great danger 
 is that they may be caught in a cyclone, before they have got sea-room 
 to avoid it. Moreover, the pilot vessels will, under the circumstances, 
 
124 REMARKS ON THE CYCLONES OF THE BAY OF BENGAL. 
 
 be generally off their stations, and ships will be unable to land their 
 pilots. It should be remembered, also, that generally the worst part of 
 a cyclone is the heavy cross sea which accompanies it. In the late 
 cyclone it was this tremendous sea, which is reported never to have 
 been equalled in the Bay, which did such damage. It is accounted for 
 by the long-continued gales from the S.W. This danger would be 
 avoided by not putting to sea until the weather improved. From (7) 
 the following would appear to be a safe rule : Ships coming up the Bay 
 in the months of May and June, with fresh south-westerly gales, ac- 
 companied by torrents of rain and a falling barometer, should not press 
 to the northward until the weather improves, and the barometer rises 
 steadily/" 
 
 One vessel (the Sophia Joakin) appears to have been in a difficult 
 and dangerous position on the 29th, with a lee shore to the north and 
 the cyclone to the south ; afraid to heave-to, she had to make a fair wind 
 of it. On this circumstance Mr. WILLSON observes, "that running 
 round a cyclone is always dangerous, more especially in those of small 
 radius, and it is liable to be forgotten that the winds do not revolve 
 round the vortex in circles, but along in-moving spirals, so that a course 
 before the wind must inevitably bring a ship sooner or later to the centre." 
 And he proceeds 
 
 " I wish to draw particular attention to the following modern develop- 
 ment of the laws of cyclonic storms, for the disregard of this law may 
 lead, and has led, to many losses at sea. The direction of the wind, 
 especially at a distance, is far from being at right angles to the bearing 
 of the centre. The wind does not revolve round the vortex in circles, 
 but along incurving spirals, differing two, three, or more points from 
 the tangential circular directions. For example, when a ship is running 
 up the Bay with a S.W. gale, squalls, torrents of rain, a barometer falling 
 rapidly, and every appearance of a cyclone, she has no right to conclude 
 that the centre bears N.W., and that therefore a North or a N.N.E. 
 course is a safe one. The centre, if at a distance, may bear North or 
 N.N.E., and a northerly course may run the ship into it, particularly as 
 it may be then almost stationary. ' The safest course seems to be to 
 lie-to, and watch the barometer and wind till the bearing of the centre 
 be known with some certainty.' In this instance the barometer gene- 
 rally rose when standing to the S.E., and fell when standing N.W. 
 This was the experience of all the fleet to the S.E. 
 
 " It must be borne in mind that, before any vortex is formed, gales 
 blow towards and round a considerable belt of low pressure ; in deter- 
 mining the changes of position of the centre of a cyclone from the 
 veering of the wind, it must be borne in mind that the direction of the 
 currents is not round a circle, but in a spiral, differing about two points 
 
REMARKS ON THE CYCLONES OF THE BAY OF BENGAL. 125 
 
 or more from the circular direction, although close to the centre, the 
 direction is probably more nearly circular. I use the following rule for 
 the northern hemisphere : 'To find the bearing of the centre, stand 
 with your face to the wind, and measure round to your right-hand side, 
 about ten points (not eight). When the centre is at a distance, it may 
 bear eleven, twelve, or thirteen points from the direction in which the 
 wind is blowing.' " 
 
 The following information respecting the revolving storms or cyclones 
 of the Bay of Bengal is also derived from a report on the Midnapore 
 and Burdwan cyclone of the 15th and 16th of October 1874, by 
 Mr. W. G. WILLSON : 
 
 Important distinctions have been observed between the cyclones 
 which visit the shores of Bengal in October and November and those 
 which occur in April and May. The storms of October and November 
 are usually generated in the eastern part of the Bay of Bengal, near or 
 a little north of the Andaman Islands ; and although in that locality 
 the formation of the storm is preceded by many days of bad weather, 
 and probably by a low barometer, there are usually no weather indica- 
 tions along the Bengal coast-line of the existence of the coming storm 
 until a day or two before its arrival ; and it is not until the place is 
 near the limit of the gale of wind which surrounds the body of the 
 advancing cyclone that any decided fall in the barometer takes place. 
 In the north-west part of the Bay of Bengal during this season North- 
 easterly winds predominate. 
 
 The cyclones which are generated in October, somewhere between 
 14 N. and 18 N., are those which most frequently visit the Bengal 
 shores, while the storms of November, which are generated further 
 south, are (usually) only felt along the Madras coast. 
 
 The cyclones which visit the shores of Bengal in April and May are 
 usually generated in the northern part of the Bay of Bengal; the 
 barometer falls steadily and considerably four or five days previously ; 
 South-west winds prevail during this season, and winds from the North- 
 east are very unusual. The storms in April and May are not usually 
 so violent as those of October, and seldom travel far inland. 
 
 Before a cyclone in the Bay of Bengal, and during its approach, 
 North-easterly winds prevail over many degrees of longitude to the 
 north of the storm, both on the eastern and western side of the path 
 subsequently pursued, and there seems to be no marked tendency of the 
 wind to veer until the cyclone is close at hand.* This fact, affecting 
 as it does the generally received rule of allowing eight points to the 
 light of the direction of the wind to determine the bearing of the 
 
 * As far as the data show, the winds were north-easterly, all over the Bay of Bengal north of 
 latitude 17 N., as early as noon of the llth October 1874, and south-westerly south of latitude 
 15 N. 
 
126 *. REMARKS ON THE CYCLONES OF THE BAY OF BENGAL. 
 
 centre Q* a 'storm, is important for the seaman to remember; and a 
 decided fall of the barometer should take place before the centre is 
 assumed to bear eight points to the right of the direction of the wind. 
 At a considerable distance from the centre, and before the barometer 
 shall have fallen much, the centre may bear as much as twelve points 
 to the right of the direction of the wind. 
 
 From the foregoing remarks it will (in vessels situated to the north- 
 ward of the storm's path) be apparent, that whereas in April and May 
 the fall of the barometer and the shift of wind from South or S.W. to 
 N.E. give timely warning of the approach of a cyclone, yet in October 
 and November the fall of the barometer is usually small, and there is 
 no warning shift of wind if the vessel be to the north of the storm's path. 
 
 To make use of the slight fall of the barometer mentioned, the 
 seaman should know that the barometer in these latitudes has four 
 diurnal oscillations, which are perfectly regular in settled weather. 
 The mercury is highest at 10 A.M. and 10 P.M., and lowest at 4 A.M. 
 and 4 P.M. ; the difference between the heights at 10 A.M. and 4 P.M. 
 is usually one-tenth of an inch ; and if the barometer did not rise from 
 4 A.M. to 10 A.M., or from 4 P.M. to 10 P.M., the seaman should be 
 on his guard, as in these . latitudes the instrument would thus indicate 
 considerable atmospheric disturbance. 
 
 At the entrance of the River Hoogly the earliest indications of a 
 coming cyclone are probably afforded by the motions of the lower clouds, 
 which drift rapidly by in dark elongated masses from N.E. to S.W., 
 and this is one of the surest tokens of a cyclonic disturbance ; these 
 signs are never absent at the Hoogly whenever there is a cyclonic dis- 
 turbance anywhere in the middle or northern parts of the Bay of 
 Bengal. The mean direction of the wind at the Sandheads before a 
 cyclone is always N.E. or E.N.E. more usually E.N.E. when the 
 place is on the track of the storm, but it varies considerably in the 
 gusts, which increase in strength as the storm approaches. The squalls 
 become more frequent, the driving rain which accompanies them is 
 gradually heavier as the cyclone approaches, and a heavy swell comes 
 up from the south-eastward. With such indications, even if the baro- 
 meter up to the time shall have shown no considerable fall, it may be 
 concluded that the cyclone is not very far distant, and it may be 
 expected that the barometer will give less warning of the approach of 
 a storm in October' than in May or June. 
 
 The storm-wave accompanying the Midnapore and Burdwan cyclone 
 raised the level of the water in Diamond Harbour, River Hoogly, 
 16 feet. 
 
( 
 
 PRACTICAL HINTS IN REGARD TO TYPHOONS OF 
 
 . 
 
 PKACTICAL HINTS IN REGARD TO THE TYPHOONS OF THE 
 CHINA SEA. 
 
 The following observations by Mr. W. DOBERCK, the Government 
 Astronomer, were issued at Hong Kong in May 1885 : 
 
 In the China Sea the earliest signs of a typhoon are clouds of the 
 cirrus type looking like fine hair, feathers or small white tufts of wool 
 travelling from east or north, a slight rise in the barometer, clear and 
 dry but hot weather, and light winds. 
 
 These signs are followed by a falling barometer, while the tempera- 
 ture rises still further. The air becomes oppressive from increasing 
 dampness, and the sky presents a threatening and vaporous appearance. 
 A swell in the sea, and also phosphorescence of the water as well as 
 glorious sunsets, are other signs useful to the mariner who is acquainted 
 with the usual conditions in the locality. 
 
 If the typhoon is approaching, the sky becomes overcast, the tempe- 
 rature in consequence decreases, the dampness increases, and the baro- 
 meter falls more rapidly, while the wind increases in force. Nearer 
 the centre the wind blows so that no canvas can withstand it, and the 
 rain pours down in torrents, but there is no thunder and lightning. 
 Still nearer the centre there is less wind and rain, and the sky is partly 
 clear, but the sea is tremendous. This is therefore the most dangerous 
 portion. 
 
 The whereabouts of the centre of a typhoon may, in the China Sea, 
 be ascertained by the rule : Stand with your back to the wind, and you 
 will have the centre on your left side, but between two and four 
 points in front of your left hand. There are, however, certain excep- 
 tions to this rule. Thus there often blows a steady easterly gale along 
 the southern coast of China when a typhoon is crossing the China Sea, 
 and the gale blows often steady from north-east about the northern 
 entrance to the Formosa Straits when there is a typhoon in a more 
 southern latitude. 
 
 When you have ascertained from the changes in the barometer and 
 in the wind in which semicircle your vessel is situated, you should, if 
 in the right-hand semicircle, keep the wind on the starboard tack, and 
 if in the left-hand semicircle you should run on the starboard tack or 
 heave-to on the port tack ; but it is dangerous to lie-to in a typhoon, 
 particularly before you are sure that the centre is past. Vessels near 
 the coast of China or in the Formosa Straits generally seek refuge in 
 the nearest typhoon harbour indicated in the China Sea Directory, 
 
 Vessels leaving Hong Kong are warned from the Observatory. 
 Vessels leaving Singapore are enabled, by observing the rules given 
 
128 CHINA AND JAPAN. 
 
 above, to sail round the typhoon till they find themselves on the eastern 
 border, when they may regain their lost distance. The force of the 
 wind is usually greatest in the semicircle north of the centre. Typhoons 
 are not met south of nine degrees northern latitude. 
 
 Typhoons may be encountered at any season of the year, but are 
 most frequent in August and September. They appear to originate 
 S.E. of the Philippine Islands. In August and September they 
 frequently pass East of Formosa or travel towards N.W. up through 
 the Straits or strike the coast of China. Afterwards they usually 
 recurve towards N.E. and pass over Japan or across the sea North of 
 Japan, but not with the violence that is characteristic of tropical storms. 
 During the remainder of the year they most frequently cross the China 
 Sea from East to West. 
 
 CHINA AND JAPAN. 
 
 Turning to China and Japan, where the PERE DECHEVRENS (of the 
 Jesuit College at Zikawei, seven miles S.W. of Shanghai) and Mr. E. 
 KNIPPING (of Japan) have been engaged for some years on meteorological 
 observations, and devoted considerable attention to the typhoons of the 
 neighbouring seas, they give, as the results of their investigations, con- 
 clusive evidence that in every instance they found "the winds practically 
 circular in front of the typhoon and near the central calm, while, far 
 from being so in the posterior quadrants, they were either convergent 
 or, in some cases, centripetal; so that, although the eight-point rule 
 would sufficiently indicate the position of the centre when in front of 
 the storm, its direction would be uncertain to the extent of three or more 
 points when at a distance from the centre, and within the area of the 
 converging winds." 
 
 KNIPPING also observes of the typhoons that pass near or across 
 Japan : " In general, it seems that outdraf t (the anti-cyclone, to which 
 the PADRE VINES refers, see p. 109) is noticed only in front of the 
 centre, with light winds, while the greatest indraft has been found in 
 the wake of the centre." And again (speaking of the storm of 1880) : 
 "In the dangerous octant the bearing of the centre amounted on an 
 average to eight points, sometimes less ; in the rear octant to thirteen 
 points or more; all cases of indraft in the dangerous octant were 
 observed to the left of the track. As similar results have been obtained 
 in other typhoons, there is good reason to believe now that in all typhoons 
 which travel at an appreciable rate, the bearing of the centre, in front, 
 will always be less than in the rear, this difference depending principally 
 on the rate of progress." The following remark also occurs: "The 
 
CLOUDS OF THE CIRRUS TYPE. 129 
 
 shape of the barometric curves in low latitudes is more like an inverted 
 cone, in higher latitudes more like a basin, agreeing with the actually 
 observed increase in the diameter of the central space with the latitude." 
 
 The most disastrous typhoons, especially to shipping, are those which 
 consist of several subsidiary low-pressure areas following fast in the 
 wake of their predecessors, and which, by the varying changes of the 
 wind, utterly perplex the navigator. 
 
 NOTE ON CLOUDS OF THE CIRRUS TYPE. 
 
 As frequent mention has been made in the previous pages (109-127) 
 of the value of watching clouds of the "cirrus" type in connection 
 with the approach of a cyclone, the following extract from "Instructions" 
 drawn up by the Rev. C. W. LEY will be appreciated by the navigator. 
 
 The attention of the observer is especially directed to clouds of the 
 cirrus type, inclusive of what are here defined as true cirrus, sheet cirrus, 
 high cirro-cumulus, and cirrus-haze ; and it is to these types of clouds 
 only that the following instructions refer. 
 
 Cirrus is the highest kind of cloud. When in moderate quantities 
 it is commonly white, though when seen through haze it is usually 
 somewhat cream-coloured, and, when the sun's rays have reached it 
 through a long stratum of hazy atmosphere, is often of either an orange 
 or rosy tint. In all cases it has a very delicate appearance. It is 
 sometimes arranged like bunches of fine hair ; and such tufts of cirrus 
 are often called "mares' tails." At other times it resembles small 
 curled feathers. Quite as commonly, however, it lies in thin light 
 strands, like pale gossamer threads. 
 
 When cirrus overspreads a large portion of the sky, it becomes what 
 is here termed "sheet-cirrus." In this state, when not too thick, it 
 produces " halos " or large rings round the sun and moon. The sheet- 
 cirrus sometimes appears fibrous, sometimes reticulated. 
 
 When the veil of this cloud becomes thick, it assumes some neutral 
 or muddy tint, except in those cases when it is so disposed that the 
 rays of the rising or setting sun are reflected to us from its under- 
 surface, which then appears of either an orange or rosy colour. 
 
 "High cirro-cumulus" differs from simple cirrus in consisting of 
 small detached masses somewhat rounded in form, a great flotilla of 
 which is often seen in the sky, especially in fine summer weather. 
 Clouds at lower levels are frequently disposed in this manner, but the 
 observer must be very careful to distinguish from these the high cirro- 
 
 i 
 
130 TROUGH OF A CYCLONE. 
 
 cumulus, which latter is either white, or changes its colour under the 
 same circumstances as simple cirrus, no part of each cloudlet ever seem- 
 ing to be decidedly thrown into shadow by another part. The high 
 cirro-cumulus also possesses the same faintness and delicacy of outline 
 which we observe in cirrus. 
 
 Sometimes, however, cirrus is only visible as either a milky or an 
 oily-looking haze, which is here termed " cirrus haze" 
 
 THE TROUGH OF A CYCLONE. 
 
 The Hon. RALPH ABERCROMBY, in a recently published work on 
 the Principles of Forecasting Weather, being No. 60 issued by the 
 Meteorological Council, has judiciously added a new term in cyclonology. 
 He defines the trough of a cyclone as a line drawn through the centre 
 of the cyclone, generally at right angles to its path : thus the trough 
 marks out the position of all the places where the barometer has 
 attained its lowest point. Everywhere the mercury is falling in front 
 and rising in rear of this line, in consequence of the forward motion 
 of the cyclone ; this line also defines the front and rear of a cyclone ; 
 and it may so happen that, although not near to the centre of the storm, 
 the passage of the trough will occasionally be marked by a sudden 
 shifting of the wind. 
 
 The front and rear of a cyclone are also well defined by characteristic 
 meteorological conditions. A cyclone is not merely an area of low 
 pressure within which the winds are blowing more or less inwards upon 
 the centre ; but it is further distinguished in this, that in front of the 
 trough we have a watery sun, or pale moon, with the attendant halo, 
 an atmosphere warm, close, and muggy, damp weather with a dirty 
 sky, drizzling or driving rain ; at the passage of the trough, a squall 
 or shower; then the sky begins to break. The whole of the rear is 
 cool, brisk, exhilarating, and dry, with a hard, firm sky. 
 
 NOTE ON THE INCURVATURE OP THE WINDS 
 IN THE TROPICS. 
 
 The particular feature to be noted with regard to a hurricane within 
 the tropics is the undoubted incurvature of the wind more or less 
 according to the quadrant of the storm in which the ship is found, and 
 also according to the latitude. MELDRUM, as far back as 1867, first 
 drew attention to the disasters resulting to vessels through the direction 
 of the centres of cyclones being estimated on the circular theory ; and in 
 March 1878, with the wind N.E. at Mauritius, it was found that he had 
 
INCURVATURE OF THE WINDS IN THE TROPICS. 131 
 
 correctly indicated the centre of the cyclone to be to the W.S.W.-ward, 
 
 whereas, on the circular theory, it should have been to N.W.-ward. 
 
 In the Tropics the inclination as indicated alike by MELDRUM for the 
 
 Southern Indian Ocean, BLANFORD and WILLSON for the Bay of Bengal, 
 
 the PERE DECHEVRONS and DOBERCK for China, KNIPPING for Japan, 
 
 and the PADRE VINES for the West Indies may be four points or more, 
 
 according to the amount of incurvature or 
 
 convergence towards the centre ; and this 
 
 large deviation from the circle, in both 
 
 hemispheres, as the storm is progressing 
 
 to Westward, is in the rear or posterior 
 
 quadrants; while in the front or anterior 
 
 quadrants the winds more nearly conform 
 
 to the circular. Hence, for the tropical 
 
 regions of the Northern Hemisphere, 
 
 Fig. 31 will approximately represent the 
 
 general trend of the winds ; and for the 
 
 Southern Hemisphere, the trend of the winds will be approximately 
 
 represented by Fig. 32. It is thus safe 
 
 to infer that when in front of the storm, 
 
 as indicated by the winds due to that 
 
 part of it, and by giving attention to the 
 
 sky, sea, and state of the barometer, the 
 
 circular theory will approximately indicate 
 
 the position of the centre, and enable the 
 
 navigator to manoeuvre so as to avoid 
 
 disaster; it appears to be also equally 
 
 certain that the rear of the storm-with Fig< 32 (S> Hemisphere). 
 
 its incurvingvriiids especially when these 
 
 are fair for the prosecution of the voyage may be equally as dangerous 
 
 as the front unless when, using those winds, the barometer be attentively 
 
 watched. 
 
 Under ordinary circumstances, the incurvature of the winds in the 
 cyclones of the middle and higher latitudes does not appear to be so 
 great as it is within the tropics. Captain TOYNBEE found it to be 29 
 in the storm of August 1873, and chiefly in front; Mr. CLEMENT LEY'S 
 observations for Europe make the mean incurvature about 25, also 
 greater for the front than the rear ; while Professor LOOM is for America 
 (and nearer the equator) made it 47, but greater for the rear than the 
 front. It seems to be not unlikely that this difference may be due, as 
 pointed out by Lieut. SPINDLER of the Russian Navy, to the friction 
 encountered by the wind on the continental side of cyclones, increasing 
 the inclination on the west side, and therefore the rear of the American 
 
132 HURRICANE SEASONS AND STORM-PATHS. 
 
 storms, and on the east side, or front of the European storms. If this 
 be the right solution, it will account for some of the anomalies in the 
 China and Japan typhoons. 
 
 HURRICANE SEASONS AND STORM-PATHS. 
 
 North Atlantic. M. ANDRE POEY, of Havana, gave, a few 
 years ago, a chronological list of every hurricane (chiefly of tropical 
 origin) in the North Atlantic of which any record could be traced 
 viz. from 1493 to 1855 a period extending over 363 years; they 
 amounted in the aggregate to 355,* and their distribution, according to 
 the months in which they occurred, appears to be as follows I- 
 
 January ... 5 
 
 February . . 7 
 
 March ... 11 
 
 April. ... 6 
 
 May .... 5 
 
 June .... 10 
 
 July .... 42 
 
 August ... 96 
 
 September . . 80 
 
 October ... 69 
 
 November . . 17 
 
 December . . 7 
 
 Hence the four months, July, August, September, and October, may be 
 said to constitute the hurricane season of the North Atlantic, since the 
 record assigns 287 (or 81 per cent.) to those months against 68 (or 19 
 per cent.) for the remaining months of the year. The extent of surface 
 embraced by POEY'S list may be divided in five areas, as follows: 
 (1) The West African; (2) the West Indian, including the Gulf of 
 Mexico ; (3) the Eastern seaboard of North America ; (4) the Central 
 Atlantic ; and (5) the Northern and Eastern Atlantic, respecting each 
 of which some remarks are necessary. 
 
 (1.) The West African area includes the vicinity of the Cape Verde 
 Islands, the Canaries, and Madeira. Of the six hurricanes mentioned 
 by POET, three are stated to have occurred in the month of September 
 off the Cape Verdes, to which may be added another, in 1872, also 
 in September, which took a general N.W. by W. direction, from lat. 
 12 N., long. 22| W. towards lat. 17 N., long. 30 W. (Annales 
 Hydrographiques). Numberless vessels are constantly traversing that 
 part of the Atlantic, yet the records of hurricanes have hitherto been 
 few, though in that locality, and during that month, all the elements 
 which give birth to them are present the area of lowest pressure lies 
 just south of the islands, near lat. 12 N., long. 20 to 25 W., with two 
 opposing winds (the N.E. Trade and S.W. Monsoon) in close proximity 
 to each other the Trade increasing in force and pushing to the south- 
 
 * A few of the instances recorded are undoubtedly duplicates ; others, that occurred within 
 two or three days of each other, are examples of a hurricane following nearly in the track of a 
 predecessor. 
 
HURRICANE SEASONS AND STORM-PATHS. 133 
 
 ward. Of the remaining three hurricanes, one occurred in October, one 
 in November, and the other in May, near the Canaries and Madeira ; 
 and of these scanty particulars are known, except that the October 
 storm of 1842 was (according to Sir W. REID) very destructive, and 
 took a route towards Seville and Lisbon, and that all three approxi- 
 mated to the curvature of the African coast. At the Canaries and 
 Madeira, on the southern edge of the region of sub-tropical rains, gales 
 are expected from the middle of October to December, with the moist 
 S.W. winds. 
 
 (2.) The West Indian area, including the Gulf of Mexico and southern 
 part of the Caribbean Sea, is the region of the tropical hurricane in the 
 North Atlantic. The examination of the records of a hundred years 
 gives 215 hurricanes (including Northers), which, as regards the month, 
 appear to be distributed as follows : 
 
 January ... 5 
 
 February . . 2 
 
 March ... 6 
 
 April .... 1 
 
 May .... 
 
 June .... 4 
 
 July .... 29 
 
 August ... 59 
 
 September . . 61 
 
 October ... 40 
 
 November . . 12 
 
 December , 6 
 
 The true hurricane season of the West Indies is July, August, Sep- 
 tember, and October, while that of the NortJiers in the Gulf of Mexico 
 extends from September to March. The southern part of the Caribbean 
 Sea is not wholly exempted from hurricanes, though it cannot be said 
 that there is a hurricane season, since at wide intervals four of these 
 storms have taken place in September, one in December, and one in 
 January; Trinidad more frequently experiences the storm-wave than 
 the storm itself, but of the three hurricanes recorded as having visited 
 the island, one occurred in August and two in September, within an 
 interval of four years. (See also the PADRE VINES' remarks, p. 115.) 
 
 (3.) The Eastern seaboard of North America is very frequently 
 visited by the West Indian hurricane after recurvature, hence many of 
 those storms already mentioned on p. 132 are common to the two areas. 
 There are others, however, the records of which exclusively relate to 
 the seaboard of the United States, Nova Scotia, and Newfoundland, 
 together with the Bermudas. 
 
 As areas (1) and (2) are approximately connected, the paths of the 
 hurricanes prevalent in the two regions may be taken together ; but any 
 attempt to classify them so as to obtain a characteristic or " habitual 
 path " month by month will utterly fail ; yet from a seasonal point, 
 and taking them in the aggregate, the paths of the earlier hurricanes 
 undoubtedly differ in form, and must be determined by a very different 
 system of atmospheric circulation, from those that occur at a later date, 
 
134 HURRICANE SEASONS AND STORM-PATHS. 
 
 Thus Fig. 33 represents that portion of the Atlantic basin included 
 from lat. 10 to 50 N., and from long. 40 to 100 W., with the varied 
 forms of the principal storm-tracks from June to December, both 
 months inclusive. The typical progressive movement is first W.N.W.-ly, 
 recurving through North, and then advancing in a N.E.-ly direction 
 the place of recurvature being somewhere between lat. 27 and 34 
 giving an approximately parabolic curve ; but this is not persistent and 
 invariable. At the beginning of the season a storm may take the form 
 (with a short course) represented by curve a; occasionally, between 
 June and October (both inclusive), it may assume a more rectilinear 
 direction, towards the West or W.N.W., across the Caribbean Sea, or 
 the Gulf of Mexico, as at b and d ; the typical and most predominant 
 form of the West Indian hurricane, with its course along the seaboard 
 
 of the United States, appears in August and September, occasionally, 
 however, in October, as shown in curves c, e, f, and g ; but a very 
 common track for the October storm is represented by curves h, i, andj 
 especially the latter, the course being N.N.E. and N.E. ; while the 
 November and December, as well as some of the October, storms 
 originate on the American continent, and take an E.N.E. to N.E. course 
 (as k and I). 
 
 The direction (so far as known) of all the West Indian storm-paths 
 recorded in POEY'S catalogue was from the equator towards the poles, 
 with the exception of one, that of 10th October 1847, which, com- 
 mencing in the Atlantic, eastward of the Windward Islands, crossed 
 Tobago to the mainland of South America thus, in its westwardly 
 progression, moving obliquely towards the equator. 
 
 REDFIELD had shown that some of the West Indian hurricanes com- 
 menced in the eastern part of the North Atlantic, near the African 
 
HURRICANE SEASONS AND STORM-PATHS. 135 
 
 coast ; it is interesting to find that our Meteorological Office may throw 
 more light on this point, and trace the life of a cyclone, with all its 
 vicissitudes, from its birth in the tropical part of the Atlantic to its 
 death in the extra-tropical part of that ocean, or it may be, on rare 
 occasions, in some part of Western Europe. In a late publication* of 
 that department, Capt. TOYNBEE observes that " the logs of two ships 
 show that a hurricane originated in about 11 N. and 32 W., in 
 August 1871 ; at midnight of the 17th its centre seems to have been in 
 about 13J 1ST. and 38 W. ; a hurricane going to the W.N.W. also 
 passed over the island of St. Thomas, at the rate of 17 miles an hour, 
 on the 21st; there can be little doubt that this was the same hurricane, 
 and if so, it must have travelled about W. by N. at a rate of nearly 18 
 miles an hour between midnight of the 17th and its passage over 
 St. Thomas on the 21st;" and, again "in the part of the sea where 
 some of the West India hurricanes originate, fresh N.E.-ly and S.W.-ly 
 winds are in close contact, and the route which they take is to the 
 southward of a permanent area of high pressure." " A southerly wind 
 is retreating before the advancing N.E. Trade during the months in 
 which West India hurricanes prevail," and "this origin between 
 opposing winds agrees with the researches of MELDRUM in Mauritius, 
 and WILLSON in the Bay of Bengal." 
 
 In August 1873 a most destructive hurricane traversed a great part 
 of the North Atlantic; killed nearly 500 people; damaged, stranded, 
 or wrecked more than 1000 vessels, and damaged or destroyed nearly 
 1000 buildings in the neighbourhood of Nova Scotia, Cape Breton, 
 Labrador, etc. The Meteorological Office instituted an inquiry into the 
 formation and progress of this hurricane, and collected a large amount 
 of meteorological data for the whole month. 
 
 Captain TOYNBEE'S observations on this storm are contained in a 
 paper read at the Royal United Service Institution on 1st June, and 
 published in the Nautical Magazine of December 1877 the subject : 
 " On the Great Hurricane, the Tracks of American Storms, and the 
 Ordinary Winds of the North Atlantic experienced in August 1873" 
 
 The storm was tracked day by day from the 14th of August, when 
 a ship, in lat. 15 N., long. 38 W., was experiencing a W.N.W. gale of 
 nearly hurricane force, thence it passed Sombrero, then westward of 
 Bermuda, and onward to the south coast of Newfoundland, where it 
 died down, and there was no gale after the 27th. It is not unworthy 
 of note that three cyclonic systems of wind were traversing the Atlantic 
 in different directions at the same time, and extended to Northern 
 
 * " Physical Geography of the part of the Atlantic between 20* N. and 10* S., and between JO" to 
 40* W.," by Capt. TOVNBKB, F.R.A.S., F.R.G.S., etc, 
 
136 HURRICANE SEASONS AND STORM-PATHS. 
 
 Europe ; and it is remarked that one of these systems was, at its origin, 
 considerably S.E.-ward of Nova Scotia and Newfoundland, "almost 
 rubbing against the great hurricane, and having its N.W.-ly winds and 
 fine weather in very close contact with the S.E.-ly winds of the hurricane, 
 while the hurricane itself never got beyond Newfoundland." 
 
 Although the storm was tracked through its whole route, and was 
 never lost sight of from 18th August until it had died out on the 27th, 
 there was a scarcity of observations in its neighbourhood until it 
 approached Bermuda. On the 23rd most of the observers were on its 
 northern side, and it was not until the 24th, when it had moved further 
 north, that it had a fair number of observers on its southern side ; on 
 the latter day it had rather an oval shape. Fig. 34 shows the circulation 
 of air in the hurricane when south of Newfoundland on the 25th; "it 
 
 pictures the N.E. winds extending further north than the S.E., and the 
 S.E. turning rather sharply into East and N.E. It seems worthy of 
 notice that on the 24th there was a similarly-shaped whirl of air near 
 the British Islands, in which N.E.-ly winds prevailed to the northward 
 and north-eastward of S.E.-ly winds. These facts, it will be seen, agree 
 in a remarkable way with the researches of Mr. MELDRUM with regard 
 to the shape of Mauritius hurricanes." (The reader can compare 
 Fig. 34 with Fig. 22 on p. 80, which is MELDRUM'S diagram of the 
 circulation of air in the Mauritius hurricane of 25th February 1860). 
 
 (4.) Of the storms of the Central Atlantic much more is known than 
 formerly ; it was the opinion of Mr. HUNT (Consul) that the heavy gales 
 off the A zores were in some instances, and in some way, related to the 
 West Indian hurricane, which is not unlikely. 
 
HURRICANE SEASONS AND STORM-PATHS. 
 
 137 
 
 (5.) The meteorology of the Northern and Eastern Atlantic has now 
 the close attention of the Meteorological Departments of the various 
 maritime nations. The exigencies of commerce and navigation demand 
 greater precision than formerly in the forecasting of weather ; and this 
 would be impossible without systematized organizations for the collection 
 of accurate observations. It was once a very prevalent opinion that a 
 cyclone was rarely encountered out of the tropics ; it is now well known 
 that the autumnal, winter, and early spring gales of the extra-tropical 
 regions in fact bad weather of all kinds, at any season appertain to 
 the cyclonic type, and have a close connection with the relative distri- 
 bution of areas of high and low pressure. 
 
 Southern Indian OceanMauritius and Reunion. Mr. 
 
 MELDRUM and M. BRIDET are the great authorities on the hurricanes of 
 the Southern Indian Ocean, and MELDRUM has given very elaborate 
 tables illustrative of the periodicity of Cyclones and Rainfall in connection 
 with Sun-spot periodicity. From 1731 to 1844 only 22 Mauritius 
 hurricanes are on record, but this imperfect list might give merely 
 those that were remarkable for their destructive effects on the islands. 
 The number of cyclones for each year, from 1847 to 1873, is as follows, 
 force of wind from 9 to 12, and the connection to which MELDRUM 
 refers is sufficiently evident. 
 
 Years. 
 
 Hurricanes. 
 
 Storms. 
 
 Whole 
 Gales. 
 
 Strong 
 Gales. 
 
 Total 
 Cyclones. 
 
 1847 
 
 5 
 
 
 
 
 
 
 
 5 
 
 8 
 
 6 
 
 2 
 
 
 
 
 
 8 
 
 9 
 
 3 
 
 2 
 
 3 
 
 2 
 
 10 
 
 1850 
 
 4 
 
 3 
 
 1 
 
 
 
 8 
 
 1 
 
 4 
 
 2 
 
 1 
 
 
 
 7 
 
 2 
 
 5 
 
 
 
 3 
 
 
 
 8 
 
 3 
 
 1 
 
 1 
 
 5 
 
 1 
 
 8 
 
 4 
 
 8 
 
 1 
 
 
 
 
 
 4 
 
 6 
 
 8 
 
 2 
 
 
 
 
 
 6 
 
 6 
 
 1 
 
 
 
 2 
 
 1 
 
 4 
 
 7 
 
 2 
 
 1 
 
 1 
 
 
 
 4 
 
 8 
 
 3 
 
 1 
 
 3 
 
 2 
 
 9 
 
 9 
 
 3 
 
 2 
 
 6 
 
 4 
 
 15 
 
 1860 
 
 7 
 
 4 
 
 2 
 
 
 
 13 
 
 1 
 
 6 
 
 2 
 
 2 
 
 2 
 
 11 
 
 2 
 
 4 
 
 2 
 
 2 
 
 2 
 
 10 
 
 3 
 
 & 
 
 2 
 
 1 
 
 1 
 
 9 
 
 4 
 
 2 
 
 2 
 
 1 
 
 
 
 5 
 
 6 
 
 2 
 
 2 
 
 3 
 
 
 
 7 
 
 6 
 
 1 
 
 4 
 
 2 
 
 1 
 
 8 
 
 7 
 
 
 
 4 
 
 2 
 
 
 
 6 
 
 8 
 
 3 
 
 2 
 
 2 
 
 
 
 7 
 
 9 
 
 3 
 
 1 
 
 3 
 
 2 
 
 9 
 
 1870 
 
 2 
 
 1 
 
 5 
 
 3 
 
 11 
 
 1 
 
 3 
 
 2 
 
 3 
 
 3 
 
 11 
 
 2 
 
 6 
 
 5 
 
 1 
 
 1 
 
 13 
 
 3 
 
 4 
 
 5 
 
 3 
 
 
 
 12 
 
 TOTAL 
 
 90 
 
 55 
 
 57 
 
 25 
 
 227 
 
138 HURRICANE SEASONS AND STORM-PATHS. 
 
 The following table gives the distribution of the Mauritius cyclones 
 according to the months of the year : 
 
 June .... 2 
 
 July .... 2 
 
 August ... 
 
 September . . 
 
 October ... 2 
 
 November . . 24 
 
 December . . 20 
 
 January . . 46 
 
 February . . 45 
 
 March ... 38 
 
 April ... 34 
 
 May .... 22 
 
 Thus, January, February, March, and April are essentially the 
 hurricane season, but cyclones are by no means unfrequent in November, 
 December, and May. It also seems highly probable that (as a rule) the 
 more easterly, and the lower the latitude of, their origin the lower the 
 parallel on which they will recurve, the more westerly their place of 
 commencement the greater their sweep in latitude, and that, though 
 they may bend or curve to the southward, they do not always take a 
 S.E.-ly progression, but should they do so their path in that direction 
 is very limited compared with the N.E. track of the West Indian 
 hurricanes. 
 
 The cyclones of the Indian Ocean are invariably attended with 
 torrential rains in front of the trough. 
 
 The cyclones of the Mozambique Channel in January take a S. W.-ly 
 direction towards the African coast. Those off the N.W. coast of 
 Australia or, more properly, between it and the south shores of Java, 
 Sumbawa, and Timor as well as the few off the West coast of Australia, 
 between lat. 23 and 30 S. (season from December to May) seem to be 
 more or less influenced by the contour of the Australian land, but 
 nevertheless belong to the Southern Indian Ocean type. 
 
 Bay of Bengal. Lieut.-Col. GASTRELL and Mr. H. F. BLANFORD, 
 when discussing the cyclone of 5th October 1864, which passed over 
 Calcutta, appended some observations on the probable law of Cyclones in 
 the Bay of Bengal, of which the following is a general summary : 
 
 Of 52 cyclones, their distribution, according to the months, is 
 
 January ... 1 
 
 February . . 
 
 March ... 1 
 
 April ... 3 
 
 May .... 13 
 
 June .... 3 
 
 July .... 2 
 
 August ... 2 
 
 September . . 1 
 
 October ... 14 
 
 November . . 11 
 
 December . . 1 
 
 showing that, though they are prevalent at the two periods the change 
 of the monsoons they are also much more numerous at the end of the 
 S.W. monsoon than at its commencement. 
 
 When the storm-paths in the Bay are carefully plotted on a chart, 
 it appears that the progression of those that traverse the north part of 
 the Bay is towards some point between N. W. by W. and N. by W. ; 
 while the ordinary path of such as occur in the south part of the Bay is 
 
HURRICANE SEASONS AND STORM-PATHS. 
 
 139 
 
 towards some point between N.W. and West. Disregarding the vicinity 
 of the Andaman Sea (between the islands and the Tenasserim coast), 
 where the tempests are generally entirely local, it appears that the 
 cyclone-paths diverge from three principal foci. The first, which may 
 be called the Northern focus, is a little to the East of the middle of that 
 part of the Bay northward of Cape Negrais and the Godavery River ; 
 the second, or Andaman focus, is in the immediate vicinity of those 
 islands, ordinarily a little to the West or N.W. of the group ; the third, 
 or Southern focus, is at about mid-distance between Ceylon and the 
 Nicobar Islands; and, discussed in reference to these foci, it would 
 appear that, at the beginning of the S.W. monsoon, cyclones originate 
 in the South part of the Bay rather than in the North, while at the 
 end of the S.W. monsoon they appear first in the North, then at the 
 Andamans, and lastly in the South. A similar result is obtained if the 
 Bay be considered as divided into only two parts by a line traced from 
 Cape Negrais to the mouth of the Godavery : for example 
 
 CYCLONES 
 
 IN THE 
 
 1-5 
 
 1 
 
 1 
 
 1 
 
 1 
 
 c 
 
 ^ 
 
 >> 
 
 >-3 
 
 
 
 <3 
 
 | 
 
 C/J 
 
 -M 
 
 
 
 1 
 
 1 
 
 TOTAL. 
 
 N half 
 
 
 
 
 
 6 
 
 s 
 
 2 
 
 2 
 
 1 
 
 9 
 
 9 
 
 
 25 
 
 S. half 
 
 1 
 
 
 1 
 
 3 
 
 fi 
 
 
 
 
 
 6 
 
 8 
 
 1 
 
 26 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 And all the works consulted give but one case of a cyclone that had 
 originated in the North part of the Bay between November and May, 
 and no instance of one occurring in the Southern half in February, or 
 during the months when the S.W. monsoon is at its full strength 
 between May and October. 
 
 Probable Causes which determine the Storm-paths. The cyclones of 
 the extreme South part of the Bay tend towards Westward rather than 
 Northward, because at the period when they principally occur the N.E. 
 monsoon is at its height (December, January; see Fig. 35); those 
 originating in the North part of the Bay progress towards the N.W. 
 and N. by W., because at the period when they occur the S.W. wind is 
 dominant, and it is but slightly affected by the current from Northward 
 (see Fig. 36). The length of the arrows shows the comparative strength 
 of winds. 
 
 Put into a practical and concise form, the periodicity and place of 
 origin of cyclones in the Bay of Bengal may be stated thus : 
 
 1. Cyclones most frequently occur at the period of the change of the 
 monsoons ; 
 
140 
 
 HURRICANE SEASONS AND STORM-PATHS. 
 
 2. They are more frequent at the end of the S.W. monsoon (in 
 October and November) than at its commencement (in May and June) ; 
 
 3. They sometimes, though rarely, occur in the north part of the 
 Bay and in Bengal during the S.W. monsoon (in July, August, and 
 September) never in the south part ; 
 
 4. They sometimes, though rarely, occur at the south extremity of 
 the Bay, and between Ceylon and lat. 5 N. during the N.E. monsoon 
 (in December, January, and March) never in the north part ; 
 
 5. At the change from the N.E. into the S.W. monsoon they occur 
 more frequently in April in the south part of the Bay ; in May equally 
 over all parts ; and in June chiefly (or only 1) in the north part of the 
 Bay; 
 
 6. At the change from the S.W. into the N.E. monsoon, cyclones 
 are most frequent in October in the north part of the Bay ; in November 
 they are rare in the north and frequent in the south part of the Bay. 
 
 Again, as to the path of cyclones : 
 
 7. Those which originate in the north part of the Bay usually pro- 
 gress towards N. by W. and N. W. by W. ; 
 
 8. Those which originate near the Andaman Islands (not in the 
 Andaman Sea) progress towards N.N.W. and West ; 
 
 9. Those that occur in the south part of the Bay have usually a pro- 
 gressive motion towards the West, or a little north of West. 
 
 It is also worthy of note, as probable, that 
 
 1. Cyclones in the Bay are preceded by a very moist and stormy 
 wind from S.W., W.S.W., or West, to the S.W.-ward of the place 
 where they take their rise. 
 
HURRICANE SEASONS AND STORM-PATHS. 141 
 
 2. Over that part of the Bay where they originate, and before their 
 formation, the barometer stands lower than ordinarily, and lower than 
 on the coasts around the Bay. 
 
 3. This depression is generated at the time when a small general fall 
 of the barometer occurs over the Bay and Gulf, such as frequently, and 
 at all seasons of the year, interrupts its normal state, or, in other 
 words, during the passage of an ordinary barometric wave. 
 
 4. A general incurvature of the Southerly current around the place 
 of low barometer precedes the formation of the centre of a cyclone. 
 
 5. At the place where the centre is forming, the barometer falls still 
 more than around the centre, and continues to fall while the cyclone 
 acquires strength. 
 
 6. The incurving S.W. current being the principal feeder of the 
 cyclone, and the centre being formed more or less towards its Western 
 border, the weather to the East and N.E. of the cyclone is more tem- 
 pestuous than to the West and N.W. of it. 
 
 7. Within and around the cyclonic centre there is a strong indraught 
 of the aerial currents, so that their direction is a spiral, not a circle, but 
 more approximating to a circle at less than 10 to 15 miles from the 
 centre than at a greater distance. 
 
 8. In all cases (perhaps only in some cases) cyclones follow no 
 regular path, but a centre may break up or disappear while another is 
 forming at some distance in advance. 
 
 The duration of the calm at the centre of the cyclone cannot, how- 
 ever (even when the diameter is known), be taken as the measure of the 
 velocity of the progressive motion of the storm. 
 
 The following observations principally apply to Calcutta and Lower 
 Bengal : 
 
 1 . If during the ordinary (tropical) periods of rise the barometer is 
 stationary or low, or if it falls more than two-tenths of an inch during 
 the usual period of fall, bad weather is approaching. The probability is 
 greater with a small gradual decrease of the daily mean through several 
 days. 
 
 2. If the wind blowing from S.E. turns to N.E. with a falling 
 barometer, bad weather is probable. The barometer and indications of 
 the weather must then be carefully watched. 
 
 3. These signs should especially be noted at the coast stations, and 
 warning given. The sea affords another very useful indication, when a 
 strong swell comes in with gentle East or N.E. breezes. 
 
142 HURRICANE SEASONS AND STORM-PATHS. 
 
 The following indications are also very important : 
 
 4. At Ceylon a stormy moist wind from S.W. or W.S.W., with 
 clouds passing rapidly in the same direction, and the barometer falling, 
 indicates a current which will probably produce bad weather in the Bay. 
 
 5. When at Ceylon there are indications of a cyclone in the vicinity, 
 there is little chance of its reaching the north part of the Bay, but it 
 will probably be felt along the coast of Madras. 
 
 6. If at the Andaman Islands, or Alguada reef, or elsewhere on the 
 Arracan coast, after some days of gentle or variable winds, with a 
 falling barometer, a moist and stormy wind from S.W. arises, the great 
 probability is that there will be bad weather in the north part of 
 the Bay. 
 
 7. If after a steady and considerable fall of the barometer, conditions 
 4 and 6 being fulfilled, a North wind springs up at stations along the 
 west side of the lower part of the Bay, a cyclonic centre has probably 
 formed ; and if condition 1 is fulfilled, it is approaching and at no great 
 distance. 
 
 These indications ought to be especially noted at the periods of the 
 change of monsoon. 
 
 Arabian Sea and West Coast of India. Observations on 
 
 the cyclones of Bombay are sufficiently numerous to show their connec- 
 tion with the change of the monsoons, as is the case in the Bay of 
 Bengal. Records for the Arabian Sea are scanty, as cyclones rarely 
 occur in that sea ; such as we have give 
 
 January ... 
 February . . 
 March . 1 
 
 May .... 1 
 June .... 2 
 July .... 
 
 April .... 2 I August ... 
 
 September . . 1 
 
 October ... 1 
 
 November . . 5 
 
 December , 2 
 
 The cyclones of the West Coast of India originate in the vicinity of 
 the Maldives or Laccadives, and the earlier ones take a N.W. to N.N.W. 
 course along the land ; those at the end of the year trend more to the 
 W.N.W. and West. According to PIDDINGTON, one in Oct.-Nov. 1845 
 came from the Bay of Bengal, and traversed India from Pondicherry to 
 the Malabar coast, pursuing a general westerly course. A very remarkable 
 and heavy cyclone passed through the Gulf of Aden in June 1885. 
 
 China Sea, and Western Part of North Pacific. These 
 
 two regions may be classed together, because the monsoons are prevalent 
 over both, but the seasonal efiect on the typhoons of the area, though 
 sufficiently distinct, is not as well marked as in the case of the cyclones 
 in the Bay of Bengal. The following record, although it shows that 
 
HURRICANE SEASONS AND STORM-PATHS. 143 
 
 typhoons may occur in any month, sufficiently indicates that they are 
 most prevalent between May and November, and especially in August, 
 September, and October : 
 
 January ... 6 
 
 February . . 1 
 
 March ... 5 
 
 April .... ^ 
 
 May .... 11 
 
 June .... 10 
 
 July .... 22 
 
 August ... 40 
 
 September . . 58 
 
 October ... 35 
 
 November . . 16 
 
 December . . 6 
 
 As regards the China Sea and China coast there appears to be, 
 according to late observers, particular seasonal storm-paths ; and these 
 paths may be either rectilinear, or curved towards the North, South, or 
 West, according to the time and place of occurrence. The effect of the 
 high land of Formosa in diverting the storm northward or southward 
 has long been noticed in those that commence eastward of that island ; 
 none cross it, as is often the case with typhoons off the Philippines. 
 Recurvature is most clearly marked in the Western Pacific and off the 
 Japan coast, especially in the instances of such storms as originate near 
 the Mariana Islands in October ; these recurve off the Bonin or Lu-chu 
 Islands. Recur vat ure has been seen in the hurricanes commencing be- 
 tween the Carolines and Marianas. Typhoons have been experienced 
 off the coast of Japan in July, August, September, and October. 
 
 ON THE PROGRESSIVE MOTION OF TYPHOONS IN THE CHINA SEAS 
 IN 1884. The Government Astronomer (Mr. W. DOBERCK) at Hong 
 Kong published, in July 1885, the following Report on the Progressive 
 Motion of Typhoons in 1884 : 
 
 Hong Kong is situated in the region of the Trades, but the winds 
 are greatly affected by the neighbouring continents principally by the 
 immense Asiatic mainland, but to some extent also by Australia the 
 influence of which is the cause of the monsoons. Thus we find, that 
 although the average direction of the wind here is East a direction to 
 some extent caused by the trend of the coast it still exhibits a regular 
 annual variation. 
 
 The air is impelled from a region where the barometric pressure is 
 higher towards one where it is lower, its motion being, however, 
 deflected toward the right in the northern hemisphere, owing to the 
 rotation of the earth. In winter, when the pressure is high over China 
 and low over Northern Australia, East or N.E. winds blow almost 
 without interruption over the China Sea. In spring, when the baro- 
 meter is falling over Southern Asia, the direction of the wind veers 
 towards South, and reaches S.W., when about midsummer the lowest 
 pressure lies over Central Asia and a high pressure over Australia. 
 In autumn the direction of the wind backs by degrees to N.E. 
 
 The S,W. monsoon does not, however, blow so steadily as the N.E. 
 monsoon, possibly because the summer area of low pressure is not so 
 
144 ' .? '^- t HURRICANE SEASONS AND STORM-PATHS. 
 
 regular as the winter area of high pressure in Asia ; and the Easterly 
 trade-wind, supported by the (at that season) comparatively high 
 pressure over the North Pacific, intrudes even in midsummer. 
 
 It is explained in the Annual Weather Report for 1884 how the 
 changes in the height of the barometer increase with the latitude. 
 North of Hong Kong the barometric pressure is subject to much 
 greater changes than south of it. In consequence, the East wind in 
 winter increases in force with a rising barometer ; and the S. W. wind 
 in summer increases in force with a falling barometer, except in the 
 presence of a typhoon. 
 
 During the winter season depressions originate within the area of 
 Asia, which is covered by the high pressure, and pass Eastward. These 
 are analogous to the depressions which originate in Nebraska and cross 
 the Atlantic, and which also have their maximum frequency and in- 
 tensity in winter. Those depressions lie outside the field of our 
 investigations, and will no doubt continue to have the attention of 
 the Observatories in the North of China. 
 
 The typhoons appear to have their origin East or S.E. of the 
 Philippines, in the trough of low pressure between the two high- 
 pressure areas in the North Pacific and in Australia. Their paths 
 are determined according to the law, first enunciated by the Kev. 
 CLEMENT LEY, according to which an atmospheric depression moves 
 so as to keep the high-pressure area on its right. This law was ori- 
 ginally proved only in the cases of depressions in the neighbourhood 
 of the United Kingdom, but it applies equally to the typhoons. Now 
 the application of this law would greatly facilitate forecasts concerning 
 the progress of a typhoon, if the telegraphic information were sufficient 
 to give a correct idea of the position and shape of the area or areas of 
 high pressure, as it is known that these are subject to comparatively 
 little change. 
 
 Typhoons may be divided into three classes, according to the paths 
 which they generally follow. No doubt abnormal instances will occa- 
 sionally present themselves, in China as well as in other countries, but 
 probably they will be of rare occurrence. 
 
 The first class of typhoons is common at the beginning and at the 
 end of the typhoon season. Typhoons belonging to this class cross the 
 China Sea, and pass either in a W.N.W. direction from the neighbour- 
 hood of Luzon towards Hainan and Tonquin ; or, if pressure is high 
 over Siam and Annam, they pass first Westward and subsequently 
 S.W.-ward. Their life is generally between five and six days. 
 
 The second class of typhoons is perhaps the most frequently en- 
 countered, and their paths can be traced the furthest. They generally 
 
HURRICANE SEASONS AND 
 
 move N. W.-ward in the neighbourhood of Luzon, and recurve towards 
 N.E. in about 26 or rather between 22 and 32 northern latitude. 
 They either strike the coast before recurving, in which case they 
 generally at once lose the character of tropical hurricanes, or travel 
 along the coast up through the Straits of Formosa. After recurving 
 they generally cross Japan or the Sea of Japan, or strike the coast of 
 Corea. These paths are explained simply according to CLEMENT LEY'S 
 law : while S. W. of the high pressure in the North Pacific they pass 
 N. W.-ward ; when arrived West of it, they pass Northward ; and when 
 arrived N.W. of it, they pass N.E. -ward, and no doubt subsequently 
 Eastward if their energy is not expended before they are North of the 
 area. They are common about the middle of the typhoon season, and 
 their life is on an average seven days. 
 
 Typhoons of the third class are possibly the most common, but it 
 appears that they are less frequently encountered, and therefore perhaps 
 often escape observation. They pass East of Formosa moving North- 
 ward. After recurving, they frequently skirt the Southern coast of 
 Japan, or enter and traverse the Sea of Japan. A typhoon of this 
 class frequently follows after one of the second class. When the latter 
 has recurved, the former proceeds Northward. This is explained by 
 the circumstance that the effective low-pressure area in Asia the pre- 
 ceding typhoon is then, in fact, considerably East of its normal 
 position. It is also well known that depressions are attracted towards 
 places which have just been traversed by a depression. Their life is of 
 uncertain duration. 
 
 Typhoons of a fourth class pass South of Luzon, moving apparently 
 Westward, or first Westward and then S. W.-ward. They are perhaps 
 not uncommon, but are really situated outside the field of our investiga- 
 tions. They appear in some respects to differ from typhoons of the 
 other classes, and I believe that they are accompanied by thunderstorms. 
 They are situated in so low a latitude, that the effect of the earth's rota- 
 tion is much less than in the case of other typhoons, and, as they appear 
 to quickly approach the equatorial belt, they form perhaps a link between 
 typhoons proper and the atmospheric disturbances in that neighbour- 
 hood, which are unconnected with rotating winds. Their life appears 
 to be between one and two days. 
 
 The average velocity of the typhoons of 1884 was as follows: 
 East of Luzon : 7 nautical miles an hour. In the China Sea between 
 12 and 18 N. : 6 miles an hour. In the China Sea between Hong 
 Kong, Luzon, and Southern Formosa: 11 miles an hour. About 
 Hainan: 13 miles an hour. East of Formosa: 10 miles an hour. In 
 the Formosa Strait: 12 miles an hour. In Kwangtung, Fokien, and 
 
146 HURRICANE SEASONS AND STORM-PATHS. 
 
 Kiangsi : 10 miles an hour. About Shanghai: 12 miles an hour. In 
 Northern China : 23 miles an hour. About Japan : 19 miles an hour. 
 In the Sea of Japan : 30 miles an hour. 
 
 Additional Note on the NORTH PACIFIC Storms. The typhoons which 
 reach Japan affect very much the form of the hurricanes of the tropical 
 part of the North Atlantic, and after recurvature pass into and across 
 the North Pacific, producing weather there similar to that which per- 
 vades the North Atlantic during the passage of a cyclone between 
 the United States and Western Europe ; occurring between July and 
 November, they are most frequent in October. In a paper recently 
 read before the Royal Meteorological Society by Mr. H. HARRIES, the 
 progress of a typhoon, which had its origin, in September 1882, near 
 the Philippines, is tracked across the Pacific and North America to 
 Labrador and Greenland, and thence well into the North Atlantic, 
 where, by joining another area of low pressure which had come up 
 from the southward, it produced the heavy gales prevalent in the 
 English Channel and Bay of Biscay at the end of October 1882. 
 
 West Coast of Mexico and Lower California. Hurri- 
 canes are not uncommon in this locality, and thence towards the 
 Sandwich Islands, generally moving in a N.W. to W.N.W. direction 
 in July, August, September, and October. There is a record of a storm 
 which crossed Central America, passed over the Gulf of Mexico, and 
 thence into the North Atlantic. 
 
 West Coast Of North America. Hurricanes have been 
 recorded moving in a N.E.-ly direction off San Francisco in December 
 and January. 
 
 South Pacific. The tropical regions of the South Pacific are by 
 no means free from hurricanes, and the missionaries have from time to 
 time given accounts of the low islands of some of the archipelagos 
 being either nearly or completely submerged. The region between the 
 Pomotu group (Low archipelago) and New Caledonia is frequently 
 visited by hurricanes of a very severe type ; they affect the form of the 
 Mauritius cyclones, and occur in January, February, or March. At the 
 Fijis they occur in January and February, the path being first S.W., 
 then recurving and following a short S.E.-ly course. At the Friendly 
 Islands they do not seem to recurve, and the season is February and 
 March. The seas between Australia and New Zealand feel the in- 
 fluence of such storms in February and March, the course being towards 
 S.S.W. or S.S.E. to S.E. At New Caledonia cyclones are expected in 
 
HURRICANE SEASONS AND STORM-PATHS. 147 
 
 January and February, and their path may be from N.W. to S.E., or 
 from N.E. to S.W. 
 
 Mr. S. J. WHITMEE, a missionary of long experience among the 
 islands of the South Pacific, says : " There is rarely a year without at 
 least one cyclone passing through, or in the neighbourhood of, one of 
 the following groups of islands, viz. Fiji, Samoa, or Hervey. 
 
 The cyclone season extends over a greater part of the period during 
 which the sun is south of the equator ; consequently, when the Trade- 
 winds from the north reach furthest south. Cyclones are most pre- 
 valent about the middle of the season, or rather later than the middle. 
 They rarely visit us earlier than December or January. 
 
 They are usually preceded for a few days by strong, northerly winds. 
 During such winds we watch the barometer very carefully, as a sudden 
 fall is a sure indication of a cyclone near at hand." 
 
 Additional Note on the S.W. Monsoon Storms generated in the 
 BAY OF BENGAL during the years 1877 to 1881 (see also pp. 138-39). 
 While going to press, JOHN ELIOT, Esq., M.A., Meteorological Reporter 
 to the Government of Bengal, has issued a Memoir on the Storms 
 generated in the Bay of Bengal during the season of the S.W. monsoon. 
 It appears to be an established fact that cyclonic storms are a regular 
 and frequent feature of the whole S.W. monsoon period, and rainfall is 
 a prominent factor in all the large atmospheric disturbances in India. 
 Generally a much smaller number of cyclonic disturbances form in the 
 Arabian Sea, and pass across the west coast of India, than advance from 
 the Bay of Bengal to its coasts. Certain conditions which frequently, 
 if not invariably, precede and accompany the formation of the cyclonic 
 disturbances of the Bay are (1) A tendency to an approximate uni- 
 formity of pressure over the Bay and the adjacent coast districts ; and 
 (2) The winds decrease in strength, and the general monsoon circulation 
 becomes feeble, more especially at its northern limits. 
 
 In respect to the frequency of cyclones in the Bay, it is stated that 
 cyclonic storms are most frequent, but least intense and severe, during 
 the period of the S.W. monsoon; and that cyclonic disturbances of 
 considerable intensity occasionally occur during the transition periods : 
 on an average, one such storm is likely to occur every year during the 
 May transition period, and two every year during the October transition 
 period. In the five years under discussion the distribution of cyclones 
 was as follows 
 
 May. June. July. Aug. Sept. Oct. Nov. Dec. Total. 
 545 11 8661 46 
 
 From which it appears that cyclones were most frequent in August : 
 all these were generated near the head of the Bay, and their tracks 
 
148 
 
 HURRICANE SEASONS AND STORM-PATHS. 
 
 indicate that the probabilities of their passing northwards or westwards 
 are about equal. Of the 46 cyclones recorded during the five years, 
 21 passed over Orissa, and 20 of these were generated during the 
 intermediate period, or S.W. monsoon proper. 
 
 Mr. ELIOT divides the cyclones in the Bay into two classes : the 
 storms of the transition periods are cyclones of low elevation, and those 
 of the intermediate periods are generally cyclones of high elevation. 
 Of the 28 storms of the monsoon period proper, not one, so far as 
 can be judged from the various observations, had a depression of the 
 barometer exceeding a third of an inch. 
 
 Extra-Tropical Regions. The storms of the extra-tropical 
 regions of the middle and higher latitudes (northern and southern) 
 are not unlike those of the tropics, except in dimension and intensity 
 and that they originate in the collision of a different class of winds ; 
 they are essentially cyclones, but their form more nearly approaches 
 to that of an ellipse than does the typical hurricane of lower latitudes. 
 Such are the usual gales which pervade the North Atlantic between 
 the United States and Western Europe. In a succession of such storms, 
 judiciously used, ships have run down their easting when bound to 
 Australia. They are also common off Cape Horn and the Cape of 
 Good Hope. Storms in extra-tropical regions may occur at any period, 
 but they are most prevalent shortly before, during, and shortly after, 
 the winter of the hemisphere, and are the well-known winter gales of 
 the locality. 
 
 TABLE, showing the different Months of the Tear in which Hurricanes 
 or Cyclones have been recorded in various Regions. 
 
 Locality. 
 
 i 
 -a 
 
 1 
 
 i 
 
 t 
 
 ^ 
 
 6 
 
 .a 
 
 i 
 
 Irs 
 
 i 
 
 1 $ 
 
 < co 
 
 1 
 
 4J 
 
 8 
 
 1 
 
 o 
 Q 
 
 I 
 
 North Atlantic (chiefly) 
 West Indian) ) 
 
 5 
 
 9 
 
 7 
 
 11 
 
 9 
 
 6 
 q 
 
 5 
 ?! 
 
 10 
 10 
 
 42 
 8 
 
 96 80 
 4 6 
 
 69 
 81 
 
 17 
 
 18 
 
 7 
 
 q 
 
 355 
 115 
 
 Bombay 
 
 1 
 
 1 
 
 1 
 
 5 
 
 9 
 
 2 
 
 
 5 8 
 
 19 
 
 9 
 
 5 
 
 62 
 
 Arabian Sea 
 
 
 
 1 
 
 9 
 
 1 
 
 9 
 
 
 | 1 
 
 1 
 
 5 
 
 9 
 
 15 
 
 China Sea and North Pacific 
 Java Sea 
 
 5 
 
 a 
 
 1 
 5 
 
 5 
 1 
 
 5 
 1 
 
 11 
 
 10 
 
 22 
 
 40 58 
 
 35 
 
 16 
 
 6 
 2 
 
 214 
 1? 
 
 South. Indian Ocean 
 
 46 
 
 45 
 
 88 
 
 84 
 
 9^ 
 
 9, 
 
 9 
 
 
 fl 
 
 94 
 
 90 
 
 985 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 In broad terms, the hurricane season, in tropical regions, may be 
 said to begin near the time of the summer solstice of each hemisphere ; 
 
HURRICANE SEASONS AND STORM-PATHS. 149 
 
 but, for the West Indies, hurricanes do not, as a rule, occur until after 
 the sun has reached its greatest northern declination, in the Southern 
 Indian Ocean, on the contrary, they may take place a month or more 
 before the sun has attained its greatest southern declination, and then 
 with greater or less frequency throughout the season. The absolute 
 verticality of the sun at the place of origin is not borne out by obser- 
 vation ; had it any effect, there would be a marked progression in the 
 direction towards the equator of the places where such storms might be 
 expected to begin, nearer the tropic early in the season, nearer the 
 equator as the season advanced, but nothing of this kind is perceptible. 
 These remarks apply more especially to the region of the typical 
 hurricane, not to that of monsoons, where, owing to the entangled 
 relations of land and water, there may be two distinct hurricane 
 seasons. 
 
 There is no instance on record of a hurricane having been encountered 
 on, nor of one having crossed, the equator, although the simultaneous 
 occurrence of two, but on opposite sides of the equator, and 10 to 12 
 apart, nearly under the same meridian, is well known. The phenomenon 
 of two similar storms raging at the same time on different sides of the 
 equator is probably incidental to the contiguity of the Monsoon and 
 Trade-wind regions : as cyclones may be expected in the Bay of Bengal 
 as well as in the Southern Indian Ocean during April and May, and 
 again during November and December, similar meteorological con- 
 ditions occurring about the same time, within the two regions, may 
 eventuate in the simultaneous production of the same result a cyclone 
 in each region but neither in any way connected. 
 
 As it is necessary not to confound the strong monsoon winds with 
 a true cyclone, so the same caution is necessary with regard to the winter 
 storms of regions bordering on, or just within, ^the tropics. Mr. MELDRUM 
 has well described those reaching from the Cape of Good Hope to the 
 Mauritius, and thence towards the equator in June, July, and August, 
 blowing hard from the southward, accompanied by a high barometer 
 (30 '2 to 30-4 inches), and raising a high sea. 
 
 The Diameter of a Hurricane or Cyclone varies, but it is un- 
 doubtedly small at its origin, increasing in dimensions in its onward 
 progression. The West Indian hurricane has been estimated at 500 to 
 1000 miles across, and expanding to larger dimensions beyond the 
 tropics after recurvature. The Southern Indian Ocean cyclones average 
 from 400 to 600 miles in diameter ; while those of the Arabian Sea 
 
150 GENERAL SUMMARY AND PRACTICAL RULES. 
 
 and Bay of Bengal do not exceed 250 to 350 miles, and the typhoons 
 of China about 50 to 200 miles or more. Widespread areas of low 
 barometric pressure, having several centres of cyclonic action, may, in 
 extra-tropical regions, have a diameter of 6000 miles, or may even form 
 a belt extending nearly round the globe. 
 
 Rate of Progressive Motion of Hurricanes. The velocity of the 
 progressive motion of the entire area of a hurricane varies not only in 
 different localities, but in the same locality and at the same season. 
 The West Indian storms have been estimated to travel at the rate of 
 15 to 18 miles an hour (360 to 430 miles a day) previous to reaching 
 the point of recurvature, and to have afterwards passed into the 
 northern part of the North Atlantic at the accelerated speed of 25 to 
 40 miles per hour, though it is often much less. In the Bay of Bengal 
 and China Sea the progression, except on rare occasions, does not exceed 
 9 to 12 miles, and is frequently not over 5 or 6 miles, per hour. There 
 would seem to be greater variation in the Southern Indian Ocean rates, 
 which, while averaging 5 to 10 miles, occasionally reach 14 to 18 miles 
 an hour, but not unfrequently in the course of their progress appear to 
 be either stationary, or perhaps not travelling at a greater speed than 
 1 to 2 miles an hour, which may well be the case if a hurricane with 
 long path is the result of successive cyclonic centres one bred from 
 the other. THOM was of opinion that the progressive velocity is 
 reduced to 100 miles in 24 hours in about lat. 20 to 22 S., and 
 that beyond 26 S. it scarcely exceeds 50 miles the causes of sup- 
 port and progression having ceased to exist, a gradual breaking up 
 is begun. 
 
 This progressive velocity is independent of the enormous intensity 
 of the wind blowing around and towards the calm centre, the steady 
 rate of which within the tropics is rarely less than from 80 to 100 miles 
 per hour, and must often, especially in the fitful gusts, and from the 
 destruction wrought, attain the velocity of 120 to 150 miles per hour. 
 
 GENERAL SUMMARY AND PRACTICAL RULES. 
 
 OLD RULE FOR FINDING THE CENTRE OF A ROTATORY GALE. Look 
 to the wind's eye, set its bearing by compass, take the eighth point to 
 the right thereof, and that will be the centre of the storm if in North 
 latitude, but the eighth point to the left of the direction of the wind 
 
GENERAL SUMMARY AND PRACTICAL RULES. 
 
 151 
 
 will be the bearing of the centre of the storm in South latitude. For 
 example : 
 
 IN NORTH LATITUDE. 
 
 IN SOUTH LATITUDE. 
 
 If the Wind be 
 
 Centre of Storm 
 will bear from 
 
 If the Wind be 
 
 Centre of Storm 
 will bear from 
 
 
 Ship 
 
 
 Ship 
 
 North 
 
 East 
 
 North 
 
 West 
 
 N.E. 
 
 S.E. 
 
 N.E. 
 
 N.W. 
 
 East 
 
 South 
 
 East 
 
 North 
 
 S.E. 
 
 S.W. 
 
 S.E. 
 
 N.E. 
 
 South 
 
 West 
 
 South 
 
 East 
 
 S.W. 
 
 N.W. 
 
 S.W. 
 
 S.E. 
 
 West 
 
 North 
 
 West 
 
 South 
 
 N.W. 
 
 N.E. 
 
 N.W. 
 
 S.W. 
 
 And similarly for the intermediate points. 
 
 BUYS BALLOT'S LAW. The law formulated by BUYS BALLOT on 
 the basis of later and more accurate observations is simply a more 
 general application of the Law of Storms which was first enunciated by 
 REDFIELD and REID, but expressed in terms of the barometer. For 
 instance : 
 
 With your face * to the wind you have 
 
 IN THE NORTHEBN HEMISPHERE, 
 
 A high barometer to your left. | A low barometer to your right. 
 
 IN THE SOUTHERN HEMISPHERE, 
 A low barometer to your left. | A high barometer to your right. 
 
 And the force or strength of the wind will depend on the barometric 
 gradients, i.e., it will be approximately proportionate to the amount by 
 which the barometers differ in a given distance. 
 
 But the lowest barometer, or centre of depression, is somewhere to 
 the right in the Northern Hemisphere, and somewhere to the left in the 
 Southern Hemisphere, not at right angles to the direction of the wind, 
 
 * The seaman having been accustomed to "face the wind" in order to determine the direction 
 of the storm's centre, the position is retained in the text, but Buys Ballot's Law is generally spoken 
 of by meteorologists in reference to the back to the wind : thus 
 Stand with your back to the wind, and you will have- 
 In the Northern Hemisphere a low barometer to your left, and 
 
 a high barometer to your right. 
 In the Southern Hemisphere a low barometer to your right, and 
 
 a high barometer to your left. 
 
 Or otherwise, in the Northern Hemisphere, stand with the left hand towards the place where the 
 barometrical reading is lowest, and with your right hand towards the place where it is highest, 
 and you will have your back to the direction of the wind; and the reverse in the Southern 
 Hemisphere. 
 
152 GENERAL SUMMARY AND PRACTICAL RULES. 
 
 as the old law of circular storms implied ; for it is an indisputable fact 
 that in cyclones or hurricanes, no less than in all storms, the wind draws 
 spirally inwards (more or less) towards the centre, instead of blowing in 
 circles returning on themselves. Hence 
 
 The newer and better RULE FOR APPROXIMATELY DETERMINING THE 
 CENTRE OF THE STORM will be 
 
 IN THE NORTHERN HEMISPHERE, 
 
 With wind North, the centre will bear more or less southward of East. 
 
 ,, N.E. ,, ,, southward of S.E. 
 
 East westward of South. 
 
 S.E. ,, westward of S.W. 
 
 ,, South ,, ,, northward of West. 
 
 S.W. northward of TS.W. 
 
 West ,, ,, eastward of North. 
 
 N.W. ,, eastward of 'S.E. 
 
 IN THE SOUTHERN HEMISPHERE, 
 
 With wind North, the centre will bear more or less southward of West. 
 
 ,, N.E. ,, westward of 'N.W. 
 
 ,, East ,, westward of North. 
 
 S.E. ,, northward of N.E. 
 
 ,, South ,, ,, northward of East. 
 
 S.W T . eastward of S.E. 
 
 ,, West ,, ,, eastward of South. 
 
 N.W. southward of S.W. 
 
 and similarly for the intermediate points in both hemispheres. 
 
 It is specially to be noted that, from the observations of able 
 meteorologists* made within the tropics during late years, the indraught 
 of the wind across the isobars of the storm's area is generally much less 
 in the front than in the rear of a cyclone. Owing to this fact, if the 
 cyclone be for a time stationary or should it have slow progressive 
 motion, as is very frequent in the Southern Indian Ocean the rear 
 winds, and those behind the trough of the cyclone, may be as dangerous 
 as the winds in the advancing front of the dangerous semicircle, unless 
 the barometer be carefully watched (see p. 131). 
 
 PKACTICAL RULES. 
 
 Heaving-to. If prudence suggests or necessity compels heaving-to, 
 then, since the ship should be put on the tack on which she will come up 
 as the wind shifts, the Rule for both hemispheres is 
 
 When in the right-hand semicircle, heave-to on the starboard tack. 
 When in the left-hand semicircle, heave-to on the port tack. 
 
 * See observations of MELDRUM, the PADRE VINES, WILLSON &c. 
 
GENERAL SUMMARY AND PRACTICAL RULES. 153 
 
 Running. In any case in which running may be deemed the most 
 judicious manoeuvre to adopt, then, if possible 
 
 Keep the wind well on the starboard quarter in the northern hemisphere. 
 Keep the wind well on the port quarter in the southern hemisphere. 
 
 The wind should be kept on the ship's quarter as much as possible 
 without causing her broaching-to. Since the winds blow towards the 
 storm's centre, running before the wind is courting disaster. 
 
 Barometer in the Tropics. As a guide to weather changes, in the 
 Tropics the diurnal variation of the barometer is ordinarily from -06 
 to '12 of an inch, and it never exceeds 15-hundredths, except for a 
 storm. The succession of changes is as follows :fatt from 9 Jh. or 
 lOh. A.M. to 3Jh. or 4h. P.M., rise from 3Jh. or 4h. P.M. to 9Jh. or 
 lOh. P.M. ; fall from 9Jh. or lOh. P.M. to 3Jh. or 4h. A.M., and rise from 
 3Jh. or 4h. A.M. to 9h. or lOh. A.M. ; any deviation from this as an 
 excessive rise or fall, or again, fall when it should rise, or rise when it 
 ought to fall is suspicious, and requires consideration, and, according 
 to the ship's position, waiting or proceeding leisurely feeling the way. 
 
 There is no such precise information derived from barometrical 
 readings in the extra-tropical regions : the diurnal variation exists in 
 the higher latitudes, but it is masked by the more frequent and 
 irregular oscillations of the instrument, due to incessant atmospheric 
 changes ; but these must be observed as assiduously as the more sure 
 indications of the tropical storm. 
 
 The most dangerous winds, and those which require the greatest 
 caution and judgment in the handling of a ship, are : 
 WHEN IN THE NORTHERN HEMISPHERE 
 Cyclone moving N.W.-ly, winds between N.N.E. and E.N.E. 
 Cyclone recurving, winds between E. N. E. and E. S. E. 
 Cyclone moving N.E.-ly, winds between E.S.E. and S.S.E. 
 
 WHEN IN THE SOUTHERN HEMISPHERE 
 Cyclone moving S.W.-ly, winds between S.S.E. and E.S.E. 
 Cyclone recurving, winds between E.S.E. and E.KE. 
 Cyclone moving S.E.-ly, winds between E.N.E. and N.N.E. 
 
 These winds, when they are encountered, are in each case the most 
 dangerous, because they are the winds characteristic of the advancing 
 front of the dangerous semicircle; then, if the barometer is falling 
 rapidly, while the wind remains steady in direction, or changes but 
 little, and at the same time is increasing in force, the ship is dangerously 
 near to, if not in, the storm's path. 
 
 Within the region of tropical hurricanes, and in the season of their 
 occurrence, it should never be forgotten that to continue to carry on 
 with a fair wind, which is increasing in strength while the barometer 
 
154 GENERAL SUMMARY AND PRACTICAL RULES. 
 
 is falling, is wholly unjustifiable. Whenever doubts arise as to a ship's 
 position in the storm, there always remains the saving mano3uvre of 
 heaving to, carefully watching the successive shifts of wind, the sea, 
 the sky, and, above all, the indications of the barometer, and then acting 
 as judgment and the rules suggest. Remember that it is of the first 
 importance that a ship should be so handled as to increase her distance 
 from the centre in the very beginning of the storm. 
 
 When in the navigable semicircle to the left of the path in the 
 N. hemisphere, but to the right of the path in the S. hemisphere 
 remember that, owing to the incurvature of the winds, the instructions 
 for heaving-to and running are as important as if the ship were in the 
 dangerous semicircle. 
 
 As cyclones sometimes travel slowly, or remain stationary for a 
 time, it is unsafe to follow closely in their wake, in the hope of profiting 
 by a favourable wind. The winds that rush in behind a depression 
 often blow more directly towards the storm's centre than do those 
 found in other parts of the storm-field. 
 
 As observed in the Barometer Manual " It is difficult to estimate 
 the distance of the centre of the cyclone from a ship. This partly 
 arises from the uncertainty as to the relation between the bearing of 
 the centre and the direction of the wind, and greatly from there being 
 no means of knowing whether the storm be of large or small dimensions. 
 If the barometer falls slowly, and the weather only gradually gets worse, 
 it is reasonable to suppose that the centre is distant ; and conversely, 
 with a rapidly falling barometer and increasing bad weather, the centre 
 may be supposed to be approaching dangerously near." 
 
 For the Mauritius hurricanes, read pp. 87 to 92, but especially 90 
 and 91 : we have no better authority than Mr. MELDRUM. 
 
 For the West Indian hurricanes, read pp. 109 to 120; the PADRE 
 BENITO VINES has thrown quite a new light on the subject. 
 
 For the Bay of Bengal, read pp. 122 to 126. 
 
 For the China Sea, read pp. 127 to 129. 
 
 When the Edinburgh Review of 1839, discussing the circular theory 
 of storms, first promulgated the notion of " infallible rules " for storm 
 sailing and this nonsense was echoed by various official authorities 
 the knell was sounded heralding the loss of innumerable lives and the 
 casting away of an immensity of property. Infallible rules were the 
 bane of shipmasters in the days when the circular theory was con- 
 sidered to be perfect, and no truer words on this subject were ever 
 written than those communicated to me years ago by the late HENRY 
 
 PlDDINGTON I 
 
QUESTIONS ON THE LAW OF STORMS. 155 
 
 " All positive rules for manoeuvring are sheer nonsense, and only 
 tend to mislead : every ship in every cyclone must have its own peculiar 
 management, dependent on the four great elements of the problem, 
 which are (1) the ship and her sea-room ; (2) the track of the cyclone ; 
 (3) its rate of travelling ; and (4) the ship's run and drift." 
 
 These remarks are especially applicable to storm-sailing, based on 
 the spiral rotation of the wind. 
 
 Questions on the Law of Storms given at the Board of Trade Examination of 
 Candidates for Master's and Mate's Certificates of Competency. 
 
 QUESTIONS ON THE LAW OP STORMS. 
 
 The Candidate is to answer in writing the whole of these Questions, numbering his 
 Answers ivith numbers corresponding with the Questions. 
 
 (1.) Q. The direction of the wind in a Cyclone being *, state the 
 
 probable bearing of the centre in the | g ^hern } Hemisphere ? 
 
 (2. ) Q. And suppose the wind during the passage of the same Cyclone were 
 
 found to change towards the *, what would be the ship's position with 
 
 reference to the line of progression of the centre of the Cyclone, and what action 
 would you take ? 
 
 (3.) Q. Under what conditions would the change in the direction of the wind 
 in the Cyclone be the reverse of the above ? 
 
 (4.) Q. What are the usual indications of a ship being on the line of pro- 
 gression of the centre of a Cyclone ? 
 
 (5.) Q. What are the usual indications that a ship is (a) approaching the centre 
 of a Cyclone, and (5) receding from it ? 
 
 (6.) Q. Describe the track usually taken by the Cyclones in the 
 
 North Atlantic, 
 Bay of Bengal, 
 China Seas, 
 Indian Ocean, 
 
 and state the seasons of the year in which they most frequently occur in that region ? 
 
 Directions as to answering the Questions. 
 (1.) When facing the wind: 
 In the N. hemisphere 8 to 10 points to the right of the wind's direction ; but 
 
 S. ,, ,, ,, left ,, ,, 
 
 will be the bearing of the centre of the storm. 
 
 Say, the given wind is N.N.E. in the N. hemisphere, then the centre will bear 
 E.S.E. toS.E. (see?. 152). 
 
 Say, the given wind is S.S.E. in the S. Hemisphere, then the centre will bear 
 E.N.E. toN.E. (seep. 152). 
 
 Note. When in front of the storm this is always safe. 
 
 (2.) This question is answered by remembering the words right, right, right, 
 which signify that if the wind changes by the right, the ship is in the right-hand 
 
 * The Examiner will fill in these blanks. 
 
156 QUESTIONS ON THE LAW OP STORMS. 
 
 semicircle, and she should be hove-to on the starboard tack, if compelled to heave-to ; 
 in both hemispheres alike (see p. 152). 
 
 Also, the words left, left, left, signify that if the wind changes by the left, the 
 ship is in the fe/l-hand semicircle, and the tack on which to heave-to is the port 
 tack ; in both hemispheres alike (see p. 152). 
 
 If, however, by the change in the wind running is advisable 
 
 Keep the wind well on the starboard quarter in the N. hemisphere, 
 
 ,, ,, port ,, S. ,, (seep. 153). 
 
 To run across the storm's path to get into the navigable semicircle is never 
 judicious. 
 
 (8.) The change in the direction of the wind might be the reverse of that given 
 in question (2) if the storm was stationary for a time, or one of slow progress, and 
 the ship was advancing in the line of the storm's progression faster than tbe body 
 of the storm. Also, if the ship were in the other semicircle. Or if another area of 
 low pressure were following on the heels of the first. 
 
 (4. ) When a ship is in the line of progression of the centre of a cyclone there is 
 little or no change in the direction of the wind, which, however, is increasing 
 rapidly in force, with a continuously falling barometer, a dangerous sea, and ugly 
 banks of clouds in the direction of the storm-field. 
 
 (5.) When a ship is approaching the centre, the barometer, already low, falls very 
 rapidly ; the wind blows furiously, perhaps shifting rapidly, accompanied with a 
 corresponding high cross sea. 
 
 When a ship is receding from the centre the barometer rises, the wind abates in 
 force, and the sea begins to go down. 
 
 (6.) North Atlantic. These cyclones commence near the equator, in lat. 8 to 
 18 N., trending W.N.W. and N-W.-ly; on passing the Northern Tropic they 
 recurve in 24 to 32 N., and then trend N.E.-ly. The season extends from July to 
 October, being most frequent in August and September (see p. 134). 
 
 (6. ) Bay of Bengal. These cyclones trend between West and N. by W. , from 
 the east side or middle of the Bay. The season is about the change of the monsoons 
 in April and May and again in October and November (see pp. 139-142). 
 
 (6.) China Seas. Typhoons trend Westerly, generally with Northing, when 
 they pass on towards Japan ; but sometimes the trend is Westerly with Southing, 
 when they pass into the China Sea. Typhoons that take a Northerly direction 
 occur between July and November, but are most frequent in August and September ; 
 those that cross the China Sea may occur at any season, but generally at the 
 beginning and end of the regular typhoon season (see pp. 128 and 144). 
 
 (6.) South Indian Ocean. These cyclones commence near the equator, in lat. 
 6 to 18 S., trending W.S. W. and S. W'.-ly ; they recurve near the Southern Tropic, 
 and then trend S.E.-ly. The season extends from November to May, being most 
 frequent in January, February, and March (see p. 138). 
 
14 DAY USE 
 
 RETURN TO DESK FROM WHICH BORROWS! 
 LOAN DEPT. 
 
 Renewed books are subject t 
 
 ; 
 
 jj^^jtf 
 
 HECtrtQ 
 
 LD 2lA-50m-8/57 
 
 (C8481slO)476B 
 
 General Library . 
 University of California 
 Berkeley 
 
UNIVERSITY OF CALIFORNIA LIBRARY