The Story of a Long Inheritance

ON the 26th of July, last summer, there was a destructive tornado at Lawrence, Massachusetts. It arrived there about nine o’clock in the morning, advancing from the southwest at a rate of forty or fifty miles an hour, and rapidly passing on to the northeast. In its momentary passage, it tore down trees, wrecked buildings, killed or injured a number of persons, and left general desolation in its narrow path. It came with so little warning and passed so quickly that we have only insufficient accounts of its appearance and behavior. One man happened to be somewhat north of its path and facing it as it came, so that he had a sight of its approach. He said he saw a great quantity of rubbish of all kinds borne up into the centre of the storm, where two clouds were chasing each other around in such a way that the eastern or front cloud moved north; that is, whirling from right to left.

These local storms are, fortunately, rare in New England, but when they visit us they manifest all the characteristic features of their class. They are peculiar in their excessive violence, and in the narrow limits within which the violent winds rush around; and from this, as well as from their sudden coming and short stay, there has been more mystery attached to them than they deserve. The theory by which they are now generally explained ascribes them to the whirling ascent of a mass of inflowing air from all sides; for repeated observation demonstrates that they possess truly vortienlar motion. The evidence of their whirling was found in the displacement of buildings and overturning of trees, but these effects are complicated by the combination of the whirling with the progressive advance of the vortex. At Lawrence, for example, where the tornado was turning from right to left as it moved swiftly northeastward, the overturning of trees to the northeast on the southern side of its central track was very general; for here the two motions of progression and whirling agreed, and the wind felt by the trees was the sum of both. But on the north side of the track, while a number of trees were blown over to the west or backward, as if by the whirling component of the motion, and the little house in which the gate-tender stood at the railroad crossing was carried westward across the street and broken, killing the poor man inside of it, yet there were many other trees on the same side of the central path that were thrown down to the southeast or east, as if by the exaggerated indraught in the rear of the whirl; for on the northern or left side of the track the motion of advance would in part neutralize the whirling, particularly at a moderate distance away from the centre, where the whirl is less violent. All this has been carefully investigated at Lawrence by Mr. H. F. Mills, whose experience as an engineer gives his statements about the tornado an especial value. A careful survey was made, under his direction, of a park through which the tornado passed, and it is evident from his account, and from the positions of the overturned trees, that the destructive winds whirled about, and in the same direction as the observer on the northern side of the track saw them turning.

Nearly the whole violence of a tornado depends on its whirling. If the inflowing air moved straight inwards on radial lines, directly to the centre, there is every reason for thinking that the velocity gained would seldom be destructive, and would never be of the terrific violence seen in the whirling storms. This may be simply illustrated by watching the downward escape of water from a basin through a vent at the bottom, where exactly the same mechanical principles are at work as in the upward escape of the air in a tornado. Fill the basin to a certain depth, and let the water come to rest; carefully remove the stopper, and notice the velocity attained by the current in running out, no whirl occurring. Fill the basin to the same depth again, and set the water slowly rotating by a motion with the hand; then open the vent, and see how greatly the velocity of the current is increased. But it is significant that the high velocity now gained is not directed radially inwards, but circularly around the central vortex. The whirling may be so rapid as to produce a centrifugal force much in excess of gravity ; for, when the vortex is well formed, an open core or eddy may be seen at the centre, where the free surface of the water does not stand level, as it would under the action of gravity alone, but nearly vertical, under the combined action of gravity and the local centrifugal force of the whirl. The surface of water stands at right angles to the forces acting on it; and therefore, in the eddy, the centrifugal force, acting outwards horizontally, must be many times greater than gravity, acting downwards, in order that the resultant of the two shall be so nearly in the direction of the former. Indeed, as a result of this active whirling, the rate of radial inflow is actually diminished, as may be seen by noting that a longer time is required for the basin to be emptied when the water possesses an initial rotary motion than when it stands still. All this is undoubtedly true of tornadoes; they would be weaker and shorter-lived if they did not whirl.

We must give a few moments’ consideration to the increased velocity of the whirl as the centre is approached, for a general principle is involved here, of great importance and wide application. Tie one end of a string around a small stone or weight, and draw the other end through a small tube. Hold the tube in one hand, and the free end of the string in the other; set the stone whirling, and then draw the string through the tube, so as to shorten the radius of rotation. As the centre is approached, the stone whirls faster; and if the experiment could be performed without friction, it would be found that the linear velocity of the stone in its whirl increases as fast as the radius decreases. The two quantities, velocity and radius, vary inversely : consequently, the triangular area swept over by the string in a given brief interval of time is constant; for the area equals half the product of the distance moved over by the stone multiplied by the length of the string, and as these quantities vary inversely their product must be a constant. For this reason the mechanical principle here involved is called the “conservation of areas.”Kepler showed that the velocities of the planets in their elliptical orbits follow this law, being accelerated as they draw nearer to the sun, and retarded as they move away. Newton proved that this variation of velocity must follow from the simple laws of motion. Like the stone on the string, so the water in an eddy or the air in a tornado whirls with increasing rapidity as it is drawn in toward the centre; and there are many other examples of the same process.

If it is now clearly understood that the violent blast of the tornado is a whirling wind, and that the tornado indraught would attain only a moderate velocity if its currents were directly radial, we may seriously regret that tornadoes whirl; they would lose their terrors if they did not. But they all do. Here, for example, is the account of an eye-witness of a tornado that happened two hundred years ago in England. The Rev. Mr. A. de la Pryme records, in the twenty-third volume of the London Philosophical Transactions, that, on August 15, 1687, at two o’clock in the afternoon, there was a destructive storm at Hatfield, in Yorkshire, in which the wind “soon created a great vortex, giration and whirling among the clouds, the centre of which ever now and then dropt down in the shape of a thick long black pipe commonly called a spout; in which I could plainly and most distinctly behold a motion, like that of a screw, continually drawing upwards, or screwing up (as it were) whatever it touched.” Coming at once to modern times, we may refer to an account of tornadoes in our Southern States by H. S. Whitfield, professor of mathematics in the University of Alabama, who tells of the clear view that he had of an approaching tornado (American Journal of Science, 1871). When first seen it was about five miles away, and, judging by the angular altitude of its top, the height of its great funnel cloud must have been about forty-five hundred feet. It approached and passed south of him, about nine hundred feet away. “The gyratory motion was distinctly visible. When about a mile distant, I saw that it would go south of me, and at this time I first observed the surface drift [rubbish], which appeared like an innumerable flock of birds flying around the summit of the column;” and at about the same time the observer saw “a pine-tree, sixteen inches in diameter and sixty feet long, float out from the black vortex, at the height of a quarter of a mile, and sail round, to all appearance, as light as a feather.” Such language from a professor of mathematics is directly to the point. There are many other accounts of the same thing. The cloudy column that marks the storm has been over and over again described as a “whirling vortex, ” or as “whirling most violently upon its centre, ” or in some such phrase. A number of my students have told me of tornadoes that they have seen in the West, all agreeing as to the whirling of the vicious funnel cloud.

Tornadoes not only all whirl: they nearly all whirl in the same direction; that is, from right to left, as the tornado at Lawrence did. It appears that a few are reported to have turned the other way, but by far the greater number of them exhibit a strong family likeness in this respect, and turn “against the sun.” Their whirling, therefore, cannot be accidental: it must be controlled by some prevailing antecedent; it must be an inheritance from some preceding condition. We must look into this.

What are the conditions that give rise to tornadoes? A few years ago this question could not have been answered satisfactorily, and there is indeed still much to be learned about it; but, thanks to the weather maps of Europe and this country, it is now clear that most tornadoes are generated within the area of one of the large cyclonic storms,¹ to which we owe our spells of cloudy, rainy weather. The weather maps are so widely distributed, and publicly exhibited at so many places in our cities, that the phenomena of the larger cyclonic storms must be in a rough way familar to many persons who have perhaps little other knowledge of meteorology. They are seen to be areas of low barometric pressure, prevailingly cloudy and rainy or snowy near the centre, with their winds moving in great inward spiral circuits, and always, in our hemisphere, turning from right to left. There has never been found an exception to this rule. These cyclonic storms are so large that they may cover nearly all the States east of the Mississippi at once ; they move across country on their general eastward track at an average rate of nearly thirty miles an hour, thus determining a general succession of weather changes from west to east. They usually require from one to three days for their passage; giving us southeasterly winds with increasing dampness and cloudiness as they draw near; rain beginning as the centre approaches, and the wind increasing in strength at the same time; and as they pass on, the wind shifts, veering through the south if the centre passes north of the observer, as is generally the case, or backing through the north if the centre passes to the south, and leaving cooling and clearing westerly and northwesterly winds in their rear. Nearly all our weather changes depend upon the passage of these cyclonic storms; and the problem of weather prediction is to foresee their movements and their changes in intensity. It is an easy matter to predict on the general rule that such stormy areas move eastward at a certain rate; but such predictions often fail, for the storms have an arbitrary way of departing from average velocities. The weather prophet who discovers how to foretell the departures of single storms from the average behavior of many has a large future awaiting him.

Now, returning to tornadoes, it is found that they nearly always occur in the southeastern quadrant of cyclonic storms, and from two hundred to six hundred miles from the low-pressure centre around which the cyclonic winds turn. The reason for so well defined a habitat of the tornado in the cyclone is undoubtedly to be found in the presence there of warm and damp southerly winds, over which there is good reason to think the cooler high-level westerly winds have advanced; thus inducing a condition of instability from which an upsetting follows, and hence the indraught at the bottom by which the tornado is always characterized. But why should the indraught take on a whirling motion, and why should the whirl prevailingly turn one way, and why should that way be the same as the turning of the winds in the much larger cyclonic storm ? There is a simple and sufficient mechanical answer for this. When a subordinate whirl is set up in a larger whirl, the little one will begin to turn the same way as the larger one has been turning. They must turn the same way. In any small part of a large cyclonic storm, no one would perceive that it was possessed of a rotary motion with respect to its centre, several hundred miles away; and yet the entire whirl of a cyclonic storm is made up of such parts, every one of which has truly a slight rotation about the cyclonic centre of low pressure. If the lower air in any one such part is drawn toward its local centre, as would occur in the case of an upsetting consequent on an unstable arrangement of warm, damp lower air under cooler, drier upper air, the imperceptible cyclonic rotation will become apparent as the indraught currents move toward the centre of ascent; and when the centre is closely approached, the whirling may become very violent. The connection of one of these rotary motions with the other is so natural and so direct that we need not doubt the inheritance of the prevalent right-to-left whirling of our tornadoes from the invariably left-handed turning of our cyclonic storms. The tornadoes that turn from left to right are of rare occurrence, and for the present have no adequate explanation; their direction of turning must be called “accidental.” They hardly enter our story, for they have lost their inheritance. It is not my intention to present here an exposition of the theory of tornadoes; if any reader wishes that, he may find it as given by its master, Professor William Ferrel, in his recent Popular Treatise on the Winds. Much that is here omitted is there stated in full, and with a completeness of argument and demonstration that has placed its author at the head of American meteorologists.

It appears, then, that our tornadoes whirl because the parent cyclones, in which the tornadoes are bred, also whirl; the whirling in both being in the same direction. So far, so good; it is a clear case of inheritance, the offspring taking after the parent. But it must now be asked, Why do the cyclones turn ? They have been described as areas of low barometric pressure at the centre, toward which the winds move inward, but obliquely, so that as a whole they sidle around the centre in an incurving spiral. Why do the winds not flow in directly to the centre of low pressure on radial paths? It is a most natural expectation that the mobile air should move from where the pressure is high to where it is low; and yet every weather maj) which includes a cyclonic storm — and this will be nearly every map in winter time —shows that the winds turn in the most persistent way aside from the direct path to the centre of low pressure, and always to the right, so as to form, as a whole, an inward left-handed spiral circulation. Not an exception to this rule of riglit-to-left cyclonic turning is known; and if the objector to this wholesale statement would bring up the case of cyclonic storms in the southern hemisphere, which turn from left to right, he must remember that, while it is true that they are there said to turn in a reversed direction, this change is only because they are there looked at from the southern side instead of from the northern. It is as it two persons were looking at a transparent watch, one seeing the face and the other the back. It would hardly be worth while for them to dispute about the direction in which the hands rotate, because one sees them turning from left to right, and the other from right to left.

It appears, therefore, that the cyclonic storms are even more closely alike than the tornadoes, in this feature of rotation. Whence have they received so persistent a habit? If the tornadoes have inherited the habit, of turning from the cyclones, from what ancestor have the cyclones received it? Just as we examined conditions in which tornadoes are formed, so must we now look to see where the cyclonic storms are bred. Those that we know in this country are affairs of the temperate zone, for the most part; some of them come from the torrid zone, but only a few. about five in a hundred. The daily international synchronous weather maps, published by our Signal Service from data gathered in all parts of the world, present the striking fact that our cyclonic storms march in an irregular procession around the north pole, along with the great north polar whirl of the terrestrial winds. From latitude 30° north, the prevailing winds of our hemisphere may be described as forming a gigantic whirl from west to east around the north pole; and it is in this gigantic whirl that our cyclonic storms are generated. It is then manifest that there is the same relation between the great polar whirl and the cyclonic storms that there is between the latter and the little tornadoes. The cyclonic storms arise in a whirling atmosphere, and they must turn in the same direction that it does. It is not necessary that they should occur at the centre of the great hemi-atmospheric whirl, that is at the pole; it is enough if their centripetal motion is excited anywhere in the area of the whirl, for reasons already given in explaining the origin of the tornado whirl within the cyclone. No new special process need be called in to explain this well-marked relationship; the same general principle applies throughout, and in the southern hemisphere as well as in the northern.

If the great polar whirls should stop, the cyclonic storms also would almost disappear; for they, like the tornadoes, gain most of their distinctive features from their whirling. Our tornadoes would become rare, at the same time; for their essential antecedent conditions. namely, those found in the southeastern quadrant of our larger rotating cyclonic storms, where tornadoes for the most part occur, would also be unusual.

It is, therefore, hopeless for us to expect to get rid of the dangerous whirl of the tornadoes as long as the great parental cyclonic storms are required to turn around, because they are generated in the still greater general whirl around the north pole; and there is no reason to think that the polar whirl will stop in our day. Consider the firmness of its foundation. The sun shines strongest on our torrid zone, and warms the air there, in contrast to that which is cooled in the polar regions. The warm expanded equatorial air flows away aloft toward either pole; and for this reason we should expect, at the first glance, to find a belt of low atmospheric pressure around the equator, and caps of high pressure at the poles, where the air, being cool and somewhat compressed, would accumulate in greater amount. But, curiously, there is lowest pressure at the poles: and this because the equatorial overflow, as it runs poleward, approaching the axis around which it rotates with tire earth, is accelerated in its rotary motion, in accordance with the principle of the conservation of areas, even to such an extent as to generate a centrifugal force that holds the air somewhat away from the polar regions, and reverses the high pressure that we expected there as a result of the low polar temperatures into a low pressure, the result of high centrifugal forces. The case is closely analogous to that of the empty eddy in the basin of whirling water, where the centrifugal force held out the water from the centre. The same mechanical principle is known to he in part responsible for the moderately low pressure observed about the centre of cyclonic storms, and is supposed. on very reasonable grounds, to produce extremely low pressure in the spinning vortex of tornadoes. No direct observations can be expected of the latter; the nearest one of value is from an autographic barometer at Owensboro’, Kentucky, about a mile and a half from the track of the severe Louisville tornado of March, 1890; but even this was too far away from the vortex to show the central low pressure, just as it was too far away to be destroyed by the central whirling winds. The evidence of doors and windows burst, open in houses over which tornadoes have passed is more directly to the point. For example, we have the account of a sufferer in the Lawrence tornado, — a woman who was building a fire in her kitchen as the storm approached. Her brief story as given in the Boston Herald at the time tells us: “While caring for the fire, I heard it raining outside. It seemed to be pouring in torrents. Suddenly I heard a terrific noise and the breaking of glass behind me. Turning around, I saw that the blinds and windows had been blown out. I started toward the windows, but I guess I never got there, I heard one crash, and that was all. When I came to,

I was lying in the ruins.” So direct a narrative has every appearance of truth, and one can hardly help regarding the blowing out of the windows as evidence of the decrease of pressure outside as the vortex came over the house. Another account of the effects of the apparent explosion of a building may be found in the memoir accompanying the remarkable map, prepared under the direction of H. L. Eustis, professor of engineering in the Lawrence Scientific School of Harvard University, of the destruction caused by the tornado at West Cambridge (now Arlington), Massachusetts, on August 22, 1851. The map is doubtless the most extensive survey of the track of a tornado ever made. Professor Eustis generally refrained from theorizing, but made the following statement: “In one case, particularly, of a factory near the West Cambridge road, the whole effect produced, and to my mind well and clearly defined, was precisely what we should have if we could suddenly place in a vacuum a building filled with atmospheric air of ordinary tension. Even the foundation walls were inclined outwards, and there was every evidence of a force acting from the interior to the exterior.” This report is to be found in the Memoirs of the American Academy of Boston for 1853.

To review: The relentless violence of tornadoes is a direct result of their whirling, and the whirling is a habit which they have inherited from the rotation of the cyclonic storms in which they are bred. The cyclones have not of themselves originated the rotation that so universally characterizes them, but in turn have received the habit from the great polar whirl of the general atmospheric circulation in which they are formed: and this has come by immediate inheritance from the rotation of that persistent and inveterate spinner, old Mother Earth, The whirling that characterizes our tornadoes is therefore passed down to them in direct line of inheritance from the rotation of their great-grandparent, and you may ask any astronomer if he thinks that will soon cease. To be sure, there would be no polar whirl if there were no equatorial overflow, but there will be an overflow as long as the sun shines on the equator; and the permanence of this may also be referred to the astronomers. They will indeed tell you that the duration of sunshine cannot be expected to reach as far into the future as the endurance of the earth’s rotation; but both are enduring enough for all practical purposes.

It may be well to mention that most cyclones have no tornado offspring, for which we may be duly thankful; but others have a rather large family. Consider the extraordinarily fruitful cyclonic storm that traversed our country on the 19th of February, 1884; as its centre moved from Illinois into Canada, it gave birth to some forty or fifty vicious tornadoes in the Southern States. Most happily for us. these little whirls are short-lived: they seldom live more than half an hour, sometimes an hour, advancing in this brief time from ten to forty miles, although their parents may go on for a week or two, and cross a continent and an ocean; indeed, one cyclonic storm has been traced in apparently continuous progress all around the world. Again, just as it is not every cyclonic storm that gives birth to tornadoes, so even the tornado-breeders do not generate these violent offspring at all points on their course, but have their breeding-grounds; and alas! the favorite ground is our fruitful Mississippi Valley. As they cross over that superb stretch of country, particularly in the spring and early summer, the cyclonic indraught brings together the unlike elements from which the tornadoes arise: the warm, damp lower winds from the Gulf, and the cool, dry upper winds from the western or northwestern interior where the temperature is still low. Nowhere else in the world is there a like opportunity for the crossing of winds so strongly contrasted, and nowhere else do cyclonic storms so often give birth to tornadoes.

The same relation of short-lived offspring and long-lived parent appears between the cyclonic storms, whose life-history we measure in days or in weeks, and the great polar whirl, whose duration we may almost call immortal. The polar whirl has times of greater activity in winter, when the contrast of temperature between equator and pole is at its maximum, and at this season the most and the strongest cyclones are generated in it. In summer time, when the difference of temperature between equator and pole is least, the whirl runs slower, and its cyclones are fewer and weaker; but it is chiefly in these latter that the tornadoes are produced. The earth must therefore already have been, and continue to be for ages and ages to come, subject to cyclones and tornadoes; yet if we take a very long view of the matter, it might be allowable to say that the polar whirl is not immortal, for it presumably was not at work when the earth was glowing with its own heat; nor will it remain in operation when the heat of the sun, on which it now depends, is exhausted. The polar whirl lives all through that immensity of time in which the sun determines our climate, but the rotation of the earth, on which the whirling of the atmosphere depends, is more enduring still. In the ardent youth of the world, long past, as well as in the cold old age, in the distant future, its rotation prevails ; we must conceive of the turning being as long-lived as the earth itself. Whence did it come by this persistent habit ?

The earth turns on its axis from west to east, or, as seen from the NorthStar side, from right to left. So do the moon and the sun, and Mars, Jupiter, and Saturn, the only other members of the solar system whose rotation is certainly known. They all turn one way. Again the same strong family likeness. Not only so: the moon revolves around the earth in the same direction as both turn on their axes, and the planets all revolve around the sun in the same direction as they rotate day after day. Saturn’s rings turn in the same direction. Everywhere the same well - marked habit of turning from west to east, from right to left. The earth is by no means an isolated, lonesome old body, but one of a family of planets: the members are scattered, to be sure, but they all bear a strong resemblance to one another. Can we venture so far back in the family history as to find the ancestor from which this resemblance has come down?

A clue to guide us in this search may be found in the case of Saturn’s rings, which are believed to consist of innumerable small separate bodies crowding around the planet in closely placed orbits. Imagine that at one part of a ring its material should be collected or clotted somewhat more compactly than elsewhere. The little bodies next west of the clot would be hurried along by its attraction and drawn nearer to it, thus accelerating their revolution around Saturn; the bodies next east of it would be held back or retarded in their revolution. Those lying on the outer margin of the ring, near the clot, would be drawn in toward it, and those on the inner margin would be drawn out; and after a long time, all the material of the rings might, in this way, be gathered about the clot, — not by any sudden disruption of the ring, but by a slow process of segregation. Throughout all this process, the conditions of the formation of a tornado in a cyclone, or of a cyclone in a polar whirl, are essentially repeated; and the mass formed by the coalescence of the parts must inevitably rotate on its axis in the same direction as it revolves around Saturn.

The more ingenious and daring astronomical speculators have supposed that all the planets once existed as rings of thinly scattered matter around the sun, and that by a process of segregation like that just described the material in each ring gradually settled together and formed a planet. During the early stage of the planets, it is thought that rings may have been formed around their coalescing masses, and from these their moons have segregated. If this is admitted, we must go a step further, and say that, of all these rings, those of Saturn must have been the most regularly built, for they have not even yet broken up. They must be wonderfully well balanced.

It is, then, from the very ancient time when the planets were rings, all turning one way around the sun, that they inherit the common impulse that gives them all the same direction of axial rotation. But why did all the rings revolve the same way ? Why not some one way, and some the other? Is there, possibly, a primeval ancestor from which all the rings inherited their uniform revolution? The most venturesome theorists have dared to search even further into the past than the time of the rings, and they think that the rings were only annular segregations from a vague, irregularly scattered nebular mass, that, as a whole, turned one way in spiral courses; and that this slow turning of the primeval nebula determined the direction in which the rings revolved, and all the rest from this.

But why did the nebula turn? Why did it not stand still? It grew from chaos ; but the elemental parts of chaos possessed, presumably, some motion, unlike in its various regions; and as their mutual attractions brought them nearer together, forming the primeval nebula, we cannot suppose that they could have avoided some slow rotation. The antecedent motions of the chaotic parts would have had to be most particularly and especially adjusted to escape this result; and chaos knew nothing of particular adjustments. Hence we may infer that when the North Star looked upon our patch of chaos, and watched its segregation into the primeval nebula, he probably noticed that it took on a rotation, a slow spiral inflow of its parts, turning so as to pass from Aries to Taurus, Gemini, and the rest, from west to east, from right to left; and from that time to this, through sun, planets, and moons, winds, cyclones, and tornadoes, the habit then gained has never been lost. Literally, this is a universal habit; and what an example of the importance of forming good habits in early youth!

It is not entirely to the imagination that we must trust for pictures of these past conditions. The earth is cold, having long ago lost its surface heat; but the sun, being much larger, is still luminous, and preserves even to these late times an image of what the glowing earth was in its youth. The planetary rings are all outgrown, but from the well - balanced rings of Saturn. those extraordinary examples of retarded development, we may infer what the planetary rings once were. The primeval nebula is vastly ancient, but in some parts of the universe the nebulous phase of development is not yet passed. Look at Andromeda’s belt in the winter sky, and there a little misty object may be seen near the faintest of the three belt stars. When examined with a telescope, this is found to be a vast nebula. A magnificent photograph of the nebula has been taken by Roberts, of London. A large lens, a sensitive plate, a perfect clockwork to make the telescope follow the turning of the sky, and a fourhour exposure have brought to sight many details not visible to the eye alone; and, most wonderful to behold, there are the spiral incurvings of the nebulous streaks, such as the North Star might have seen while watching our early growth. There is the warrant for believing that our primitive nebula turned in its spiral courses, and gradually settled into revolving rings, from which the planets grew; while the sun represents the great central mass into which most of the nebulous matter was drawn.

But what a long inheritance is this! You have heard of the glacial invasion from which New England and other northern countries have lately escaped, and of which we have witnesses in the many scattered boulders on our hills: that was prehistoric, and yet it should not be regarded as ancient; it may be placed about two inches back on a line that represents the scale of time. You have seen the splendid gorge of the Hudson through the Highlands: that was begun perhaps ten feet back on the scale. You have heard of those strange reptilian tracks in the sandstones of the Connecticut Valley; those are decidedly older, possibly fifty feet back. You know the coal from the mines down in Pennsylvania: the coal plants grew in the Pennsylvania marshes long before the reptiles made tracks in the Connecticut sand flats, may be eighty or a hundred feet ago. You may have been down to Braintree, near Boston, and seen the trilobites in the slate quarry there: those are vastly more antique than the coal plants, two, or three, or four hundred feet distant on the time scale. But these examples are to be dated after the earth had taken on. practically, all its modern habits. It was then, as now, accompanied by a moon that ran around it in the same way as both bodies turned on their axes. It must then, as now, have had its lands and oceans; its tides, currents, and winds; its storms, with their clouds and rain. How much further back should we have to go to find the earth only just segregated from its ancestral ring, and how much earlier still were the nebulous rings forming from chaos? No one can say. And yet, through all this time we trace the persistent inheritance of a primeval habit that was learned in the childhood of time. If that old nebula had taken on the habit of turning the other way, the sun would rise over our western hills and set on the Atlantic; if that old nebula had turned the other way, the moon would work its way westward through the stars, and its first quarter would show us the left half illuminated, not the right; if that old nebula had turned the other way, we should here receive the tempered breezes from the ocean for our habitual winds, while western Europe would suffer under harsh easterly winds from the interior of the vast Eurasian continent; the seat of modern civilization in the Old World would be well-nigh uninhabitable, and bleak Labrador would enjoy a tempered climate. If that old nebula had turned the other way, the Lawrence tornado would have come from the east, not from the west; it would have turned from left to right, not from right to left; and the house in which the gatetender was killed would have been dashed eastward, not westward, on the north side of the track.

When we speak of inheritance, we think generally of the inheritance of property from a parent; and this means that we live in a country of established laws. Established laws give us security in the transmission of inheritance. But how local, how shortlived, how vacillating, are our human laws of inheritance compared with these eternal laws of physical inheritance, persistently in operation since the first segregation of chaos! The lesson of tornadoes would, at first sight, seem to be one of danger; but the larger lesson is one of safety, —safety under the constant operation of fixed natural laws.

William M. Davis.