Go to National Library of New Zealand Te Puna Mātauranga o Aotearoa
Volume 30, 1897
This text is also available in PDF
(1 MB) Opens in new window
– 465 –

Art. LVI.—Notes on the Vertical Component of the Motions of the Earth's Atmosphere.

[Read before the Wellington Philosophical Society, 13th October, 1897.]

Plate XLV.

The very exhaustive observations which have been made in a great number of places on the earth's surface on the varying conditions of the atmosphere have resulted in a very great advance in our knowledge of the laws which govern its motions; but hitherto meteorologists have mainly investigated the horizontal motions in their velocity and direction, and in their relation to the atmospheric pressures as evidenced by the barometer. The isobaric charts compiled from the recorded facts at the same hour at a number of stations scattered over a wide area are most valuable in forecasting the probable weather conditions for a day or two in advance, and in giving us some insight into the horizontal circulations of the atmosphere. Complicated as these motions are, they are probably more simple and regular in the Australasian district than in any district of similar area (about 70° of longitude by 40° of latitude) where regular observations are established in any other part of the world.

Situated on the borderland between the two great belts of anticyclonic and cyclonic circulations in the Southern Hemisphere, and without any land-surfaces to interfere with the full development of the latter until the south-eastern part of the Australian Continent and Tasmania are reached, and with a fresh set of observations on the line of advance of the cyclones and anticyclones along the extensive barrier of New Zealand, and finally at Chatham Island, the conditions are

– 466 –

exceptionally favourable both for the normal circulation and for its observation.

In studying this horizontal circulation with the aid of the most valuable weather-charts prepared by Mr. Wragge at Brisbane, corrected by the completer information recorded in New Zealand and the Chatham Islands for that sub-district, I have been increasingly persuaded to believe that the phenomena exhibited in these charts of horizontal motion and atmospheric pressure need for their elucidation a knowledge also of the vertical circulation, and of the connection between barometric readings and upward or downward motions of the atmosphere. We know experimentally that atmospheric pressure diminishes as we attain higher levels above the earth's surface, and so by the indications of the barometer, or the temperature at which water boils, we can obtain approximately the heights of mountains; but we know also that the heights so obtained vary considerably. May not these variations be due in some measure to upward or downward motions of the atmosphere, which diminish or increase the pressure locally? Is it not also probable that low barometric readings observed over a large area of the earth's surface may be caused in some measure by an upward circulation of the atmosphere there, and high readings by a downward circulation? If such a view be permissible, it enables one to understand a little better what is going on in a cyclonic disturbance. It is evidently impossible that the air is merely gyrating with great velocity inwards and downwards towards the centre from all sides; it must get out somewhere, and the only outlet is upwards.

In a small whirlwind we can observe this upward motion in the centre, where light bodies are whirled upwards; but in a great cyclone, with an external diameter of perhaps a thousand miles or more, and where there is always a comparatively small central area, it is not easy to conceive how the great horizontal circular motion of the wind is maintained, unless there be also a vertical circulation, probably descending on the outside of the storm area and ascending on the inside. In our antarctic cyclones the southerly winds are cold and the northerly winds are warm, which would seem to indicate that the horizontal circulation has not extended round a considerable part of the circle.

Some meteorologists conceive that these antarctic cyclonic disturbances are not true cyclones, but are portions of cyclones open to the south, where we know there is a belt of low pressure, and the observed fact that, even at the southern extremity of New Zealand, easterly winds are rare gives support to this view. On the other hand, we have the not infrequent complete development of the closed cyclonic circuit in storms

– 467 –

which were traced along the south coast of Australia, the centre passing over New Zealand, and over or near Chatham Island, as on the 18th to the 22nd August; on the 3rd to the 6th, and probably on the 14th to the 16th, September; also on the 6th, 7th, and 8th October; and again on the 12th October, in 1897.

It would seem, therefore, that to extend our knowledge we require to extend our observations further to the south, and also to adopt some method by which we may note the vertical inclinations of the motions of the wind as well as its horizontal directions and its velocity. The difficulty which meets us at the outset of such an inquiry is the very limited height of the stratum or shell of air which we can easily observe. Ascending currents must be fed from below, and descending currents must be deflected horizontally or upwards when they reach the ground. At first sight it would seem best to use mountain observatories, but on consideration I think it will be admitted that the observations made in such localities, although most instructive, will be found somewhat less reliable than those made on level plains, or low islands or promontories, because horizontal winds striking the sides of mountains must always be deflected upwards. Stationary balloons or kites offer the best positions for such observations, if suitable recording instruments can be devised. I believe, however, that much valuable information can be obtained from observing the motions of a wind-vane so constructed as to show the true direction of the wind both vertically and horizontally, and fixed on a pole (at least 30 ft. high) on an open level site. I have observed the indications of such a wind-vane for the last few months, and already I have learned much from it. By the courtesy of Mr. Ferguson, the secretary and engineer of the Wellington Harbour Board, it has been placed on the lightning-rod above the time-ball on a wooden tower by the wharf, and at a height of about 70 ft. above the water. The principle of the vane is that the directing tail has a horizontal blade in addition to the usual vertical blade, and that it is pivoted at its balancing-point, so that it is free to move in the vertical and also in the horizontal plane, and to take up the true direction of the wind in both planes. (See Plate XLV.)

The first information derived from the motions of the vane is that the air moves in waves generally. These waves vary notably in inclination and period, the inclinations varying from 1° to at least 30° above or below the horizontal, the period being sometimes reckoned by minutes, sometimes by seconds, or even fractions of a second. At times the upward or downward inclination is steady (or with waves) for hours at a time, or, again, there may be for hours no regular deviation from the horizontal. These facts are of great importance, of course, in

– 468 –

architectural and engineering works, where the force and inclination of the wind have to be provided for. Hitherto, I think such great inclinations above or below the horizontal have not been recognised.

The wave - motions, too, must immensely increase the difficulties connected with flying-machines. Birds, I have no doubt, with their exquisitely adapted organism, are able instantaneously to detect the wave-motions of the air, and to utilise the waves in their soaring flight by varying correspondingly the inclination of the surface of the wing, and a young bird soon learns to balance itself and move on its wings as a land animal does on its feet. But evidently the problem how to make a machine that will balance itself in the moving waves of the air as a ship does on the moving waves of the sea is one of extreme difficulty.

The shore of Wellington Harhour, situated as it is in an amphitheatre of high hills, and close to Cook Straits, with their special local winds, is not favourable for obtaining reliable general information on the vertical component of wind motion, or on its relation to barometric indications; my observations, also, have as yet been extended over only a very short period; but so far they seem to indicate that in a marked V depression, particularly if, as often happens at Wellington, the depression assumes the form of a wide head of a valley with a narrow outlet, the inclination of the wind is upward; and in the contrary case, when the isobars are convex towards the “low,” the inclination is downward.

On the only occasion when I have as yet been able to observe the vane while the calm centre of a cyclone was passing over Wellington—on the 20th August last—there was a decided but slow upward current of air, although at the level of the vane there was no perceptible horizontal current. On the other occasions before mentioned, when the centre of a cyclone passed over Wellington lately, either the vane had not yet been set up or the centre passed at night, and the indications were not observed, and this leads me to the question of how it may be possible to make the indications of such a vane self-recording. I have devised two methods—one electrical, the other mechanical—by which it might probably be accomplished, and Mr. Ferguson has suggested a third, by the agency of light reflected from a mirror fixed to the vane; but the motions are often so quick that the ribbon moved by clockwork, on which the record would be made, must move at great speed, and the record would be very voluminous, and the greater part would be valueless. Perhaps it might be so arranged as to be set in action by electricity by a distant observer for a short time when considered advisable. In any case a system of record would be somewhat expensive, and

– 469 –

would require skilled workmen to make the necessary apparatus. A simple vane, observed even once a day at the ordinary hour of observation at various suitable localities, would give, I believe, much important information if the records were collated in a central office. More frequent observations would be often very useful under special weather conditions.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

In fine calm weather here the mechanism of land- and sea-breezes is well illustrated by the wind-vane, which shows the sea-breeze as an ascending current taking the place of the warm air heated by contact with the land, and rising. The descending land-wind follows late in the day. The regular fall and rise of the barometer which I have noticed at coast stations in the tropics, and which in settled weather is so marked and so regular that it indicates the hours like a clock, is doubtless due to the strong ascending and descending currents of air caused by the great heating of the land by day and its rapid radiation of heat at night. Here I have observed the barometer to fall 1/10 in. in a short time during a sea-breeze, shown by the vane to be ascending. The barometer rose again when the breeze ceased, and the vane became horizontal, or showed a descending current. On the other hand, I have observed the barometer to fall 0.13 in. in about seven hours during a northerly gale, while the windvane showed no regular inclination in the motion of the lower stratum of air, but only wind waves. This diminution of pressure must therefore have been caused either by a decrease in the total height of the atmosphere or by upward movement in the vertical circulation of higher strata of the atmosphere. The latter appears to me the more probable cause. Occasional abnormal differences between the barometric readings at stations very near each other—as, for instance, at Auckland and Manukau, Christchurch and Lyttelton, Dunedin and Port Chalmers—are, I think, caused by up and down movements of the lower atmosphere produced by the deflection of the wind striking on steep hillsides, rising up on one side and falling on the other. These are, it is true, minor irregularities, which do not materially affect the greater circulations; but they are instructive if they establish the principle that barometric readings do not depend entirely upon the total height of the atmosphere at the place of observation, but also in part on the vertical components of the motions of the atmosphere above the station.

The following experimental observation on the effect of a strong upward current of air on the barometer was made here lately: The signal-station on the summit of Mount Victoria, on the south side of Wellington Harbour, is 640 ft. above high-water mark by survey. The hill rises very steeply from

– 470 –

the water's edge, having an inclination of 1 in 3.5 approximately. A northerly gale striking the hill is forced upwards at this inclination, whatever the previous direction of the wind may have been with reference to the horizontal plane.

On the 8th November, 1897, it was blowing a moderately strong gale from the north, and the following observations were made by Mr. F. W. Rutherfurd with a good small aneroid barometer at about noon:—

Place of Observation. Reading of Aneroid. Inches. Difference of Height by Scale on Aneroid. Addition and Resulting Height above Sea. True Height, and Errors.
At foot of Mount Victoria, 10 ft. above high-water mark, both before and after summit reading 29.675 True height, 640 ft.
At summit of Mount Victoria, varying according to wind force—
From 28.850 725 ft. + 10 = 735 Errors, + 95.
To 28.800 770 ft. + 10 = 780 " + 140.
Mean 28.825 750 ft. + 10 = 760 " + 120.
[or] 757 " + 117.
At the back of the crest, under shalter of a wall 28.900 680 ft. + 10 = 690 " + 50.

The velocity of the wind was, of course, much greater at the summit of the hill than it was at the bottom, where no variation was observed between the readings before and after the ascent, nor was there any fluctuation in the pressure at the sea-shore producing the variations in the readings (or “jumpings” of the barometer) at the summit, which were there so marked.

This experimental observation seems to indicate that a strong upward current in the air reduces considerably the pressure of the atmosphere, and gives an abnormally low barometric reading. It shows that the heights of mountains taken by means of the barometer in windy weather are liable to considerable error, and that this will probably be in excess, and may be as great as nearly one-fourth of the height in low altitudes, perhaps not so large a proportionate error in high altitudes.

The observation also seems to indicate on a small scale the effect of upward or downward movements in the upper atmosphere, in its vertical circulation, on barometric indica-

– 471 –

tions generally. These experimental results were confirmed by a second experiment when the wind-force was about half of that in the first experiment. The details were as follows:—

Second experiment made at noon on the 22nd November, 1897, at the same place. The weather was fine, with a north breeze, fresh to strong. The wind-vane showed that the inclination was about 20° upward, with a slowly falling barometer as the sea-breeze increased in strength.

Place of Observation. Reading of Aneroid. Inches. Difference of Height by Scale on Aneroid. Addition and Resulting Height above Sea. True Height, and Errors.
At foot of Mount Victoria, 10 ft. above sea-level, both before and after summit reading 29.775 True height, 640 ft.
At summit of Mount Victoria 28.975 710 ft. +10=720 Error, + 80 ft.
On southern slope, about 30 ft. below, and 100 ft. behind crest 29.035 650 ft. +10+30=690 " + 50 ft.

The errors in this case are less than those in the first experiment, corresponding with the lesser force of the wind, but they are in the same direction. The reduced pressure caused by the upward deflection of the air was similarly less evidenced at 100 ft. in rear of the crest (a greater distance than was tried in the first experiment). The “jumpings” or fluctuations in the readings of the aneroid, which were so marked in the first experiment, were not observed under the more favourable conditions prevailing during this experiment.

A subsequent experiment was made on the 3rd December. The wind was a light north-west breeze, shown by the windvane to be generally ascending, while the barometer was falling. A large aneroid, by Casella, was compared with the small instrument before used. The results were as follows:—

Place of Observation. Large Aneroid. Small Āneroid.
(1.) 4 ft. above high-water mark 30.092 in. 30.18 in.
Summit of Mount Victoria 29.405 in. 29.38 in.
(2.) 4 ft. above high-water mark 30.058 in. 30.12 in.
Corrected height by (1) 636 ft. 704 ft.
Corrected height by (2) 600 ft. 644 ft.
Mean height, and error 618 ft. (- 22 ft.) 674 ft. (+ 34 ft.)
– 472 –

No corrections have been made for temperature, which varied between 50° and 60° in all the observations.

From these last it would appear that the heights given by the small aneroid are always somewhat in excess; but the principle is confirmed that the vertical component of the motion of the air strongly affects barometric readings.

It is only by establishing facts by observations on a small scale that we can obtain a firm basis for reasoning out the atmospheric motions in their great circulations, and I am in hope that some experienced meteorologists will take up this line of investigation, and follow it out to the discovery of the vertical systems of circulation of the atmosphere as fully as those in a horizontal direction have been traced. The combination of the two motions, or their resultants, will no doubt be found to be very complicated and intricate curves, and probably some new method, or some addition to the present system of isobars, may be found necessary to show graphically the approximately true circumstances of atmospheric circulation from time to time; but a new field of investigation appears to me to be open to meteorologists, the results of which no one can predict. This line of investigation has, indeed, been attacked already in some measure by observations of cloud motion, but the results have as yet been small. The weakness of this method is due to the constant changes in the forms and masses of the clouds as the vapour in the atmosphere becomes visible or invisible by varying temperature, and also to the difficulty of making simultaneous observations of any identical part of a cloud from the two ends of a measured base. The motions and the forms of clouds at different elevations do, however, give valuable information. A peculiar form of cirro-stratus cloud which is almost always seen here previous to the arrival of an antarctic cyclonic storm seems to indicate a vertical circulation on the advancing edge of the storm. These clouds have a rounded upper surface, peculiarly smooth and somewhat similar to a fish, or sometimes to an eel when several are joined lengthways. I imagine that these “fish” clouds are formed where ascending currents from the storm area curve over and fall again on the outside as they are cooled down in the higher atmosphere.

I have not read in any publication on meteorology of this peculiar form of cloud as associated with the advancing edge of a cyclonic storm, and it is possible that some local peculiarity at Wellington may give rise to them. It would be interesting and instructive to know if they are noticed on the southern coast of Australia and the west coast of Tasmania.

It is a matter of common observation that the upper clouds are sometimes seen to be moving in the opposite direction to that in which the lower clouds are moving—sometimes even a

– 473 –

third direction of movement may be observed in a third and higher cloud stratum. It seems quite probable that when two strata of clouds are seen to be moving in opposite directions the observer is favourably placed for observing the upper and under surfaces of a vertical circulation, the upper being visible through the spaces between clouds in the lower, and the curved ascending and descending parts in the circulation being distant, and probably obscured in clouds caused by the contact between strata at different temperatures.

Such a circulation may be observed on a small scale and at a low level when a sea-fog rolls inland in opposition to the direction of the wind and clouds above. The rapidly rising cumulus clouds, often seen before a thunderstorm, seem to show the upward movement in part of a local vertical circulation.

It appears to me improbable that variations in the atmospheric pressure at an observing-station, as shown by the barometer, are caused chiefly by variations in the total height of the atmosphere over the station. The evidence brought forward in this paper, as far as it goes, seems to show that upward or downward currents in the lower stratum of the atmosphere affect very decidedly the pressure, as shown by the barometer locally. If similar observations made in other places confirm these results we may infer that what holds good locally on a comparatively small scale is also true as regards the circulation of the atmosphere in cyclones and anticyclones, and also in the still more extensive circulation between the poles and the equator, and that the isobars in our charts do not represent absolutely greater or less heights of the atmosphere, like contour lines of a map showing hills and valleys; but that they indicate greater or less pressure, caused partly, if not mainly, by the downward or upward movements of the air in its vertical circulation.

The idea of using a balanced wind-vane with a horizontal tail-plate in addition to the vertical tail suggested itself to me in connection with these investigations after reading an article in a late number of the “Journal of the Royal Geographical Society” describing the exploration of the Gobi Desert by a Russian expedition, which was driven back by intense heat and drought. On their return to the country bordering this hot desert they noticed that a cool wind blew from the desert. They extemporised some sort of wind-vane, which was not described, but which showed them that this cool wind was falling at a steep inclination. It was evidently an overflow of the heated air, which had been cooled down at a high level, and what they observed was a portion of a vertical circulation, which they do not seem to have investigated further, having other objects in view.

– 474 –

The following are the principal points which occur to me as worthy of observation by means of balanced wind-vanes established in favourable position in connection with meteorological observatories:—


The periods of wind-waves in winds of different velocities.


The inclinations of wind-planes either in rapid or in long-period waves, or in more steady air-motions.


The connection between barometric indications of pressure and ascending or descending currents in the lower stratum of the atmosphere.


The special wind inclinations peculiar to the different parts of the horizontal circulation of cyclones and anticyclones.


Extension of these observations to higher levels by mountain observatories, or kites, or balloons.


The effects of slopes of mountains in deflecting the wind upwards (or, it may be, downwards on reverse slopes) with reference to measurements of heights by barometers.

Other points will probably suggest themselves to other minds, and probably some better form of instrument may be devised than that which I have used, and also some means of automatic record. For this latter the mechanical arrangement which I have thought out seems to me the more likely to be successful, and of this I will give a brief description, which may possibly be useful to any meteorologist who has the means to have the apparatus constructed and installed. The principle is that the standard on which the vane is placed should be tubular, and that a light stiff rod (probably of aluminium, about No. 10 B.W. gauge) should pass up through its axis, the weight of the rod being supported by a spiral spring at its lower end, adjusted so as to keep the upper end always in light but firm contact with the lower edge of the metal disc on which the vane is pivoted. The lower edge of this disc would be formed into a cam, or eccentric curve, so that when the tail end of the vane is depressed the rod will be depressed, and when the tail rises the rod will rise by the force of the spiral spring. A projecting ring on the lower end of the rod would influence the shorter arm of a balanced lever, the longer arm of which would carry a pen or pencil recording the upward and downward motions on a paper ribbon moved by clockwork, and ruled with parallel lines. These lines might probably be nine in number, corresponding with elevations or depressions of 10°, 20°, 30°, 40° above or below the horizontal.

The mechanical defects of the arrangement are obvious: (1.) Friction between the cam and the head of the rod. (2.) The tangent to the cam surface is not perpendicular to

– 475 –

the axis of the rod. (3.) The inertia of the rod retarding its movements.

Nevertheless, the arrangement is comparatively simple, and it would, I think, record the principal vertical movements of the lower atmosphere on a ribbon moving at a moderate speed—say, a minimum of 1 in. an hour, or 2 ft. in twentyfour hours. The wave-motions of the air are so rapid at times that to record them a speed of from 3 in. to 6 in. a minute would seem necessary. Possibly the kinematograph, in combination with the velocity of the wind given by the anemometer, during the short period occupied in taking the series of photographs, would best afford data for constructing graphically the wind-waves at any particular time as shown by the wind-vane.

The record of the more enduring inclinations of the windmotions would be very much confused by the wave-motions on a slow-moving ribbon, and I only suggest the slow motion to avoid the inconvenience of the great length of ribbon necessary to record fully and clearly the whole of the motions. Experiment only can decide the best speed to adopt.

During sixty-eight days in September, October, November, and December, on which observations have been made at Wellington, there were thirty-four days on which the windcurrent was upward, five days on which it was downward, and twenty-nine days on which it was horizontal. At Lincoln, Canterbury, of the twenty-four days on which observations were taken up to the present date, the current was upward on nine days, downward on two days, and horizontal on thirteen days. The preponderance of upward over downward movement is very marked in both places, and this preponderance is also reported to exist at Odessa by Professor Klossovsky.

In Wellington the phenomenon of sea-breezes, as before mentioned, may partially account for it, and the fact that both Christchurch and Wellington are frequently in the V depressions which run up the east coast of the South Island when antarctic storms pass farther south may eventually prove to be a connected fact, but the connection between horizontal and vertical movements in the air is still very obscure.

I observe in Nature of the 7th October, 1897, p. 551, that Professor A. Klossovsky, of the Odessa Observatory, has made some interesting experiments upon “the ascending and descending currents of the atmosphere by means of an anemometer turning in a vertical direction.” This method occupied my attention for some time, but I was unable to devise any plan by which the motion caused by ascending currents could be distinguished from that resulting from descending currents, as both would be in the same direction. Apparently, the

– 476 –

professor has solved the difficulty, but no information is given on the subject in the paragraph in Nature. Probably the details will be found in the “Annales of the Odessa Observatory” for 1896, to which, unfortunately, I am unable to refer, as the work is not sent to New Zealand.

Circulation of the Atmosphere.

Since writing my paper for the Sydney Scientific Congress, my observations of the balanced wind-vane and of cloudmotions have led me to the following conclusions:—


That in addition to and in connection with the great primary vertical circulation of the earth's atmosphere, by which an interchange is constantly being effected between the equatorial and polar zones, and the secondary vertical circulations, apparently in three divisions in each hemisphere, by which the primary circulation is effected, there are numerous minor circulations frequently taking place both in a vertical and a horizontal manner in the different strata of the atmosphere. During anticyclonic weather at Wellington, New Zealand, with the prevailing northerly winds, cloud movements often show minor local circulations at a level of from 5,000 ft. to 10,000 ft. above the sea. The movement is usually upwards on the north side and downwards on the south side, while horizontally the movement is cyclonic; but sometimes the circulation both horizontally and vertically seems to be reversed. The mass of air involved in these minor circulations is variable, but seems to average not more than half a mile in diameter, often probably much less. These small circulations are within and subordinate to the third order of circulations, which are concerned in our cyclones and anticyclones.


At present, then, we have indications of at least four orders of magnitude in atmospheric circulations. It is very probable that there may be many more.


We have indications also from experiment and observation that barometric variations in pressure depend on the resultant of the masses and velocities of the air in upward or downward motion in the different strata of the atmosphere over the place and at the time of the observation.

From the above considerations it would seem that balanced wind-vanes at low levels can only give information regarding the fourth order of circulations, in which I include the circulations involved in land- and sea-breezes, and that the cyclonic and anticyclonic circulations of a higher order may be so much modified locally by these minor circulations that the balanced wind-vane may often not give reliable indications concerning them. Even at high levels the minor circulations will, I think, often affect the indications of the wind-vane, which, however valuable, must always be limited in their range.