does not average more than about 10°. The period from the 26th of October to the end of December 1863, was remarkable for the number of storms which occurred-about fourteen well-marked storms having swept over Great Britain in that time. During these months frost scarcely occurred. The uniform mildness of the weather was shown by the circumstance that in the end of December 245 plants were in flower in the open air in the Edinburgh Botanic Gardens, 35 of these being spring flowers; and the strange spectacle was seen of Sweet Peas and Hepaticas blooming together. Now in America storms are almost invariably succeeded, or rather terminated, by intense cold. It occasionally happens, however, that intensely cold winds terminate British storms, and in these cases it has been invariably observed that barometric pressure is very high in Iceland. The storm of the 1st and 2d December 1867 was of this sort. In this storm the wind suddenly veered to N.W., N., or N.E., suddenly rose from a light air to a storm, the barometer rapidly rose, and the temperature fell everywhere to the extent of 20° or 30°. The northerly gale was felt first at North Unst, in Shetland, from which it proceeded towards the south-east, advancing with an extended front many hundreds of miles in length. At Brussels, which was south of the area of lowest pressure accompanying this storm, no sudden change of the wind took place, but it veered successively through S.S.E., S., S. W., and W. to N.W.

STORMS OR HURRICANES OF THE TROPICS.

540. THE BAHAMA HURRICANE OF OCTORER 1866.+-On

*See a paper by Professor Balfour on "The State of Vegetation in the Royal Botanic Gardens, Edinburgh, in the end of December 1863," in 'Proceedings of the Royal Society of Edinburgh,' Session 1863-64.

The data of this hurricane have been obtained from Commander A. J. Chatfield's Record Chart of the storm; Governor Rawson's Report of the Hurricane; and observations received from Mr Andrew Lang, St Croix; the Royal Mail Steam Packet Company; the Director-General of Hospitals, &c.

Plate VIII. are exhibited the barometric pressure and the direction of the winds at different places in the West Indies at 8 P.M. on the 1st October 1866, when the centre of the storm was at Nassau. At this instant the barometer was 27.7 inches at Nassau, while at a distance of 286 miles it was as high as 29.7 inches; at Bermuda, St Thomas, and Barbadoes, it was at the average height; but to the east, north-east, and north of Barbadoes, it was above the average height of that part of the Atlantic. These figures show such extraordinary fluctuations of the barometer as never occur beyond the tropics. In Great Britain the fall of one-tenth of an inch of barometric pressure in an hour is reckoned a large fall, such as only accompanies great storms; but during this storm, when the barometer was falling most rapidly, it fell at Nassau 0.700 inch from 4 to 5 P.M. The fall of the barometer became more rapid as the centre of the storm approached; but at the centre the pressure remained nearly stationary for a time during the lull which there prevailed, and before the barometer began to rise. Since at Nassau the lull at the centre lasted from 7.20 P.M. to 8.50, we learn, from the rate at which the whole body of the hurricane was carried forward, that the calm at the centre covered an area of at least 23 miles across. This calm may be considered as bounded nearly by the isobarometric 27.800 inches, in the centre of which the pressure was only 27.700 inches. Since ordinarily, within the tropics, the barometer rises and falls so little that it may be practically regarded as stationary, such enormous changes of the barometer are the more striking; they are so sudden that the eye, in looking at the mercurial column, can plainly see it falling, and they may be regarded as registering rather than foretelling the different stages in the progress of the storm. Barometric differences much greater than these have been observed. Thus, during the hurricane which devastated Guadaloupe on the 6th September 1865, it is stated in the 'Bulletin International' that the barometer at Marie Galante, a neighbouring island, was 29.929 inches at 4 A.M., 29.646 at 6.30 A.M., 29.174 at 6.47 A.M., and

27.953 at 7.40 A.M., having thus fallen 1.693 inches in one hour and ten minutes!

541. Storm-Wave of the Sea accompanying the Hurricane. -Owing to the diminished pressure of the air at the centre as compared with what prevailed at the outskirts of the storm, the difference being fully two inches of mercury, the level of the sea at the centre would be raised about three feet, being sustained at that height by the greater pressure all round. This increase of level, when occurring at high tide, and being increased still further by violent winds blowing in upon the centre, quite accounts for the advances made by the sea over the land, especially over the low-lying islands, and the heartrending scenes of desolation which it caused. In certain parts of Turk Islands and some other flat islands, those only were saved from drowning who succeeded in climbing up the trees. The great storm which swept over Calcutta on the 5th October 1864, furnishes a notable illustration of the destructive power of the storm-wave occasioned by such low atmospheric pressures. On that occasion the sea, rising ten

feet above the highest spring tides, broke over its usual bounds, and laid the whole level country at the mouth of the Ganges under water, by which 45,000 human beings. perished.

542. Direction of the Wind.-From the chart, Plate VIII., in which the arrows represent the winds at 8 P.M. of the 1st October, it will be observed that in no instance did the winds blow directly to the centre of least pressure at Nassau, or blow round the area of low pressure in circles returning on themselves; but that in every instance they blew in a direction intermediate between these two directions. Hence the storm was rotatory, and, as the direction of the arrows show, it revolved in a direction contrary to the hands of a watch; and within the area of rotation the winds approached the centre by spirally in-moving currents of air. It will be observed that the winds nearest the centre blew more directly towards it than those at a greater distance from it. In this respect this hurricane resembled every European and American storm

which I have hitherto examined from actual observations laid down on synchronous charts.

543. Connection between the Direction of the Wind and the Centre of the Storm.-Standing back to the wind, the centre lies to the left hand of the direction in which the wind is blowing. This holds in all places north of the equator, and it furnishes the rule which must be observed by ships in steering out of the course of the storm. From this relation of the winds to the pressure is also deduced the rule for predicting the direction of the wind at particular seaports during storms. Thus, suppose at 9 A.M. it be required to know the direction in which the wind will blow in London at 9 P.M., a storm being observed advancing from Ireland towards the east. Information being had through the telegraph of the course the storm is taking, and an inference being drawn from that observed course that at 9 P.M. its centre will be near Liverpool, then at that hour the gale may be expected at London from S.S.W.

544. The only exceptions to the rule, shown by the winds on Plate VIII., are the winds at St Thomas and St Croix. But it must be considered that the storm had already passed these places; the winds were therefore light, and were now beginning to yield to the influence of the high barometer which prevailed at that time to the east and north-east.

545. Probable Track of the Centre of the Storm.-From observations received, the position of the centre was determined at seven points, which were then connected together by the hacked line on the chart. It will be observed that it began to the north-east of St Thomas, and thence pursued a W.N.W. course, coming round to Nassau; thence its course was to the N.E., passing 150 miles to the north of Bermuda. Its course thus assumed the form of a parabolic curve, with Nassau near the apex. This is the track usually followed by West Indian hurricanes.

546. Rate of the Onward Course of the Hurricane, and Velocity of the Wind.-The rate at which the whole body of the storm travelled was slow as compared with European

storms, being only fifteen miles, until it had passed the Bahamas and was nearing Bermuda, when it proceeded with the accelerated speed of thirty miles an hour. But this rate gives no idea of the violence of the hurricane, which depends not on the velocity of translation of the whole body of the storm, but on the speed with which the winds whirl round and in upon the centre. This, doubtless, rose to a steady velocity of from 80 to 100 miles an hour, and for short intervals to 120 or even 130 miles an hour, a velocity which was registered during the Guadaloupe hurricane of September 1865. The long black list of wrecks recorded, bears testimony only too emphatic to the devouring energy of the hurricance.

547. The barometric differences determining the force of the wind must have been very great. Since the pressure at Nassau at 8 P.M. was 27.70 inches, and 286 miles distant it was 29.70 inches, the mean barometric gradient for the distance was 1 inch to 143 miles. But since the barometer did not fall uniformly over the intervening space, much steeper gradients must have taken place. The following barometric observations were made by Captain Chatfield, R.N., at Nassau on the 1st :

:

Now since the onward course of the storm as it neared Nassau was at the rate of fifteen miles an hour, after which it increased gradually to thirty miles,-suppose that between 4 and 5 P.M. it advanced twenty miles, then there probably occurred at 4 P.M. a difference of 0.700 inch of pressure between Nassau and places twenty miles to the south-east, which would give a gradient of 1 inch to about twenty-nine miles. Colonel Sykes has recorded a still more remarkable case of great difference of pressure. During a cyclone off the