the whole earth cannot be retarded exactly as though it were a rigid body. Now the tidal protuberance has not this required form, and therefore there results a slow secular distortion of the earth arising from the unequal distribution over the surface of the forces which constitute the tidal frictional couple.

The last part of the paper does not lead to results of interest to the general reader, as it is concerned with the part played by inertia in the tides of viscous, fluid, and elastic spheres.

I

INDIAN METEOROLOGY

N the article

1

to east relatively to the equatorial parts.

The exact mathematical solution for this kind of a distortion of a viscous spheroid shows that it consists in a simple cylindrical motion round the axis of rotation, each point moving from east to west with a linear velocity proportional to the cube of its distance from that axis.

The distortion of the surface of the globe consists of a motion in longitude from west to east, relatively to a point in the equator, the rate of change of longitude being proportional to the square of the sine of the

latitude.

Numerical calculation shows, however, that in the later stages of the earth's history (the development being supposed to follow the laws found in the paper on "Preces

sion") the distortion must have been very small. With a certain assumed viscosity it is found that, looking back 45,000,000 years, a point in latitude 60° would lie 14' further east than at present. From this it follows that this cause can have had little or nothing to do with the crumpling of geological strata.

As, however, the distorting force varies inversely as the sixth power of the moon's distance, it seems possible that in the very earliest stages this cause may have had sensible effects. It is therefore noteworthy that the wrinkles raised on the surface would run north and south in the equatorial regions, with a tendency towards northeast and south-west in the northern hemisphere, and north-west and south-east in the southern one. The intensity of the distorting force at the surface varies as the square of the cosine of the latitude.

An inspection of a map of the earth shows that the continents (or large wrinkles) conform more or less to this law. But Prof. Schiaparelli's map of Mars is more striking than that of the earth, when viewed by the light of this theory; but there are some objections to its ❘ application to the case of Mars. If, however, there is any truth in this, then it must be postulated, that after the wrinkles were formed the crust attained sufficient

local rigidity to resist the obliteration of the wrinkles, whilst the mean figure of the earth adjusted itself to the ellipticity appropriate to the slackening diurnal rotation: also, it must be supposed that the general direction of the existing continents has lasted through geological history.

The second problem considered in this paper is concerning the distribution of the heat, which would be generated by the internal friction of the tidal distortion.

It was shown in the preceding paper that a very large amount of heat might be thus generated, and it appeared at first sight as though this might serve to explain in part the observed increase of underground temperature; but the solution of a certain problem concerning the cooling of an infinite slab of rock 8,000 miles thick, in which heat is being generated according to a certain law of distribution, shows that the frictional heat could not possibly explain a rate of increase of underground temperature near the earth's surface of more than 1o Fahr. in 2,600

feet.

It follows, therefore, that Sir W. Thomson's investigation of the secular cooling of the earth cannot be sensibly affected by this cause.

Memorie della Società degli Spettroscopisti Italiani, 1878, vol. vii.

"Atmosphere" of the Encyclopædia Britannica it has been justly remarked that one of the most important steps that could be taken towards the development of the science of meteorology would be

India, which offers splendid contrasts of climate at all seasons, has a surface covered at one place with the richest vegetation, and at others with vast stretches of sandy deserts, and presents extensive plateaus and sharp ascending peaks, all which conditions are indispensable for collecting the data required for the solution of the problem of atmospheric physics. In working out this problem it is necessary, owing to its extreme complexity and difficulty, to give attention, not merely to questions immediately bearing on the physics of the atmosphere, but also to climatic and other practical inquiries, which may be handled with comparative ease and which afford results that contribute indirectly but very materially to the solution of the higher problem. The publications enumerated below admirably follow up this two-fold line of inquiry, and even already several important practical and theoretical conclusions seem not far from the point of being reached by the meteorologists of India.

The "Report on the Meteorology of India" is the second Annual Report issued since the administration of the Indian Meteorological Establishment was concentrated in the Central Office at Calcutta for the whole of India including British Burmah and the Islands of the Bay. In the scheme of publication of the monthly results of the observations made at the various stations over India, we note with satisfaction that the form proposed by the Permanent Committee of the Meteorological Congress at Vienna has not been adopted in some of its more important details. Thus in Mr. Blanford's tables, instead of a general monthly mean of atmospheric pressure, the mean monthly pressure for each hour of observation is given an essential requisite for the presentation of the data required in discussing various of the more important problems of international meteorology. Indeed these tables possess the very high merit of being, with perhaps one exception, entirely suited for the discussion of climatic questions of an international character-the single exception being the lumping together of the two or four daily observations of the winds into one monthly mean, instead of a monthly mean for each hour of observation as is so admirably carried out by Professor Rubenson in his annual reports of Swedish meteorology.

The most interesting part of this report is that which deals with the failure of the rains in Western and Southern India which resulted, as is only too well known, in one of the most terrible and wide-spread famines of recent years. The mode of treatment is grounded on the practice adopted by the Office, in framing forecasts of coming seasons to which we have several times drawn the attention of the readers of NATURE (vol. xiii. p. 66, &c.), and which may be described as proceeding on the assumption that there is a certain persistency in meteorological conditions; that, for instance, the longer a given state of weather has lasted, the less the probability of a speedy change; and that as regards the distributions of pressure, on which weather is so dependent, certain states of the atmosphere tend to perpetuate or reproduce themselves in the same region in such a manner as to maintain a

"Report on the Meteorology of India in 1876." By Henry F. Blanford. "The Indian Meteorologist's Vade-Mecum." By Henry F. Blanford. "Indian Meteorological Memoirs; " issued under the direction of Henry F. Blanford. Vol. i, part 2. "The Meteorology of the Bombay Presidency." By Charles Chambers, F.R.S.

constant difference between the average pressure of two neighbouring regions which, though protracted, is not permanent, but disappears after a longer or shorter time. Mr. Blanford largely inclines to trace the failure of the rains to an unusually great expanse of snow covering the southern slopes of the Himalayas, much of which fell very late in the season, and which acted as a cooling agent, bringing about an abnormal distribution of pressure, and consequently of winds, temperature, and rainfall, conditions which, once fairly established, went on reproducing themselves so that cyclonic and anti-cyclonic areas of an abnormal character gained a certain persistency over those parts of India where the rainfall was deficient and where it was in excess. Should future observations confirm this hypothesis, the result will be one of the most important yet arrived at in practical meteorology.

The least satisfactory part of the report, perhaps, is that referring to the relation of rainfall to the sun-spot period, in which too much stress appears to be laid on the results of data collected from a wide geographical superficies, and too little stress upon data referring to limited regions; the data of which regions, it may be added, require for their satisfactory discussion to be examined with reference to their seasonal as well as annual variations during the sun-spot periods.

The practical part of the "Indian Meteorologist's Vade-Mecum" being part 1 of the work, is in many respects a model-handbook for the observers for whose use it is intended. The clearness with which the difficulties attending the making of real observations of temperature are apprehended is altogether admirable; and the provisions and precautions as regards instruments, hours, and modes of observing actually taken are of such a nature as likely to secure observations of a high quality, owing to an increased intelligence, and efficiency on the part of the observers who work in accordance with the principles and instructions laid down for their guidance.

Mr. Chambers' book is an elaborate and important work on the Meteorology of the Bombay Presidency, based on all the observations made in the Presidency down to 1874. Its splendid porte-folio of eighty highly finished maps and diagrams printed in colours, as well as its excellent typography with 159 tables of results, many of them being wholly or in part laborious and elaborate analyses of the different data of observation, render the work an édition de luxe. The contributions with which this work enriches Indian meteorology are twofold, viz., the results of the hourly observations made for many years at Kurrachee, Deesa, Bombay, Poona, and Belgaum; and the monthly averages for numerous stations throughout the Presidency, from which the temperature, rainfall, and winds of this part of Asia are charted with a fulness and consequent approximation to the truth not hitherto attainable. The influence which the broad physical features of the region, such as its lofty mountain ranges, high plateaus, river valleys, and extensive sandy deserts, has on the climatology of the Presidency is worked out with great skill and ability. Still more able are the discussions of the hourly observations of pressure, temperature, humidity, and cloud, made at the five chief stations, together with many suggestive reflections on the results developed, which will well repay the reader's best attention, even though he may sometimes not see his way to agree with the opinions expressed.

A healthy feature of Indian meteorology is the vigorous manner in which the making of hourly observations is pushed at many stations which have been admirably chosen as respects the objects sought to be attained, and the comparatively full and prompt discussions of the results which are published from time to time. Of the problems handled in those discussions the most frequent as well as the most important is that of the diurnal oscillations of the barometer. To this very difficult problem Mr. J. Eliot, for example, makes a valuable contribution

in a paper on two storms in Bengal during 1876, which were accompanied with increased atmospheric pressure, and the apparent reversal of the normal diurnal oscillation of the barometer. This reversal was found to be accompanied with an instantaneous and complete change of wind direction and a great fall of temperature, which, as they occurred before the rain began to fall, proved that they were not due merely to an inrush of a strong humid current from the Bay of Bengal. The sudden chilling ot the air, accompanied as it was by an increase of pressure, also proved that the changes were not due to the internal action of a mass of air or to horizontal or surface currents from the interior, which would have been warm currents, but that they were probably produced by the downrush of a cold upper current, a conclusion which will doubtless receive further examination not only from its bearing on barometric fluctuations but also on the theory of storms.

OUR BIG GUNS

E may leave the explanation of the disaster on board

been appointed to inquire into the matter. But in the mean time it will be well to consider what are the elements of weakness, if any, in the construction of our big guns.

The system of building up large guns by shrinking coiled iron tubes over a central steel tube seems extremely well adapted to prevent a lateral explosion; for even when the steel tube has had a longitudinal crack, the gun has been frequently fired without any further evil consequence.

But our guns are manifestly deficient in longitudinal strength, for the steel tube is the only tube continuous from end to end. If, then, there should be any ring-crack in the steel tube, there is little to prevent its separation into two parts beyond the friction of the coiled tubes, and the dove-tailing by which it is attempted to join the coiled cylinders together.

Now considerable longitudinal stress on the steel tube must be caused every round by the rifling necessary to give the shot its proper rotation, and occasionally, by a jamming of the shot. Also every discharge of the gun must cause a violent vibration in every part, and should the junction of the IB coil with the C coil and breechpiece work rather loose, this would be likely to cause a ring-crack in the steel tube in that neighbourhood.

When rapidly-exploding powder was used in the service the guns were very properly rifled with an increasing twist with a view to remove every possible obstruction to the initial motion of the shot. The increasing twist is still in use notwithstanding all the efforts that have been made to manufacture a powder that will burn slowly, so as to make the propelling pressure on the shot more nearly uniform. With a view to distribute the work of giving rotation to the shot uniformly along the bore, the rifling should be calculated to give a nearly constant pressure on the studs. But this depends upon the law of explosion of the powder. And this law is very variable, and very little understood. Only we know this that the more nearly the force propelling the shot becomes constant, the more nearly the rifling approaches the uniform twist in order to obtain a constant pressure on the studs. Now the objection to the increasing twist is that it throws the chief part of the work of giving rotation towards the muzzle, where the gun is weakest. Also there is a difficulty in arranging the studs on the shot, and it now appears that the increasing twist allows the shot to slip forward when the gun is depressed. It seems, therefore, desirable to revert to the uniform twist of rifling now an improved powder is used.

But in order to give the gun additional strength in direction of its length, it seems desirable that the steel

THE recent researches of Prof. Marey on the electric discharge of the torpedo have been presented by the author in an extended memoir published last year. We propose to present to our readers the main conclusions reached by M. Marey, and the experimental demonstrations on which the principal of these are based. But before entering into details of the experiments let us indicate summarily the processes employed by M. Marey. In previous researches, made in 1871, he had at his disposal only the reactions of the muscles of the frog to analyse the electric phenomena of the torpedo; he caused to be recorded, upon an inclosed plate, the shock of a frog's muscle produced by the discharge of the electric apparatus of the torpedo. The instant of the excitation of an electric nerve or of the nervous centres of the torpedo was recognised; and it was seen that the movement of the foot of the frog presented, at the instant of excitation, a considerable retardation, equal, e.g., to fourhundreds of a second, measured on the chronographic scale. But into this total retardation entered several diverse elements, which M. Marey took into account by causing the muscle of the frog to contract by an excita tion directly acting upon it. The time lost by the muscle of the frog representing nearly the half of the total retardation, it was concluded that the time-test by the electric apparatus is equal to that of the muscle of the frog.

Since these first researches, M. Marey has been able to study more directly the electricity of the torpedo by making use of the electro-magnetic signals of M. Deprez and of Lippmann's electrometer.

M. Deprez's signal is composed of a small electromagnet provided with an extremely light armature of soft iron, which is applied to the coils when the current which traverses them is closed, and which is drawn from it, without delay in demagnetisation, at the moment of the rupture of the current, by the contraction of the tight india-rubber thread. The armature is provided with a style which traces on the inclosed cylinder the closures and ruptures of a current, the duration and frequency of these successive acts, with such perfection that it is easy thus to obtain the record of 1,000 vibrations per second. In the tracing underneath the apparatus (Fig. 1) is seen the signals which it furnishes when acted on by a noncontinuous scale of 500 simple vibrations per second.

It is this electro-magnetic signal which M. Marey placed in the circuit formed by the torpedo, whose apparatus was held between two metallic plates joined to the coils of the apparatus by two conducting-wires. We shall see, further on, what use he has been able to make of this.

The second instrument by means of which certain special points of the experiments have been made is Lippmann's capillary electrometer. This apparatus is formed essentially of a column of mercury sustained by capillarity, in a tube of extremely fine glass, the extremity of which is plunged in a bath of dilute acid. When the mercury of the apparatus and the acidulated water are placed in connection with two points of "Compte Rendu des Travaux du Laboratoire de M. Marey." T. iii. Paris: G. Masson. :877. * "Annales de l'Ecole Normale Supérieure." ze s., t. i., pp. 86-114.

an electric circuit of unequal tensions the capillary column is displaced and is carried towards the side of strongest tension. This displacement is instantaneous, and if the variations of electric tension are produced successively with great rapidity we need not fear the inertia of the capillary column. All the variations are signalled whatever be their frequency. But as the movements of the capillary column cannot be registered themselves, M. Marey has had recourse to photography in a certain number of experiments.

Let us now consider the results following the order which we have indicated at the outset.

1. A torpedo's discharge is not a continuous current; it

is formed of a series of successive waves added one upon

acts.

The fundamental experiment upon which the demonstration of this proposition rests was performed with the electro-magnetic signal (Naples, October, 1876). Having compressed one part of the apparatus of an active torpedo just drawn from the water between two metallic plates furnished with conducting wires, M. Marey placed the signal-machine of M Deprez in contact, and the magnet being stimulated he heard a shrill noise resembling that made by filing the end of a hard splinter of wood. The vibrations of the armature, therefore, had been produced by a series of successive electric In defining these vibrations one is justified in stating that the discharge of the torpedo produced by the animal as the result of a local excitation, was composed of a variable number of waves or currents succeeding each other. Fig. 2 represents two tracings so produced. The great advantage resulting from the use of the electromagnetic signal was to show definitely that the discharge is complex, an analysis which was not possible with the frog's-foot signal. The muscle used as reagent does not in fact react by means of the shocks apart from impulses which are sudden and frequent; it remains in a state of permanent contraction.

But the electro-magnetic signal, whilst showing the dissociation of the torpedo discharge, furnished no other result. It did not indicate how those successive waves follow each other, it seemed even to lead to the conclusion that one wave is quite completed when the next succeeds. At this point the induction is interrupted and the experimentalist adopts another mode of solving this question of the succession of waves in a discharge. M. Marey, in fact, being convinced that the electric action of the torpedo and muscular action should be assimilated, and wishing to see in the discharge the analogue of induced tetanus and even of voluntary contraction, could not resign himself to the admission of an absolute discontinuity between the successive acts constituting a charge. Yet the electro-magnetic signal apparatus seemed to pronounce his theory wrong. But on passing through Lippmann's electrometer a slight current from the total discharge, M. Marey observed that the column underwent a series of successive impulses, the effects of which unite together. This progression by successive jerks indicated an increase of the intensity of the discharge, an increase in which each new wave is joined to what remains from those which have preceded it. Thus we derive the proof that the electric waves are partially united to one another like the muscular shocks of a tetanised muscle.

dis

This first fact being gained, it was necessary to follow up the analysis of the torpedo-discharge, determine the nature of each of the independent electric acts which the electro-magnetic signal had revealed, measure their duration, phases, &c. These different points have been elucidated, each in its turn.

2. To measure the duration of the electric-wave in the torpedo, M. Marey has had recourse to the method devised by Guillemin for determining that of very short current, and used afterwards by Bernstein to measure the duration of the negative variation of nerves and muscles. Guillemin's method is applicable in every case where a current passes several times in succession through a metallic circuit, with duration always the same. The electric condition of the circuit is investigated during a succession of very short intervals, beginning at the moment when current is complete.

The apparatus used by Guillemin and Bernstein was the galvanometer; M. Marey preferred to use a frog's foot, which, in the successive investigation, gives a movement which can be graphically recorded, as often as there is an electric current.

The graphic method, by which each duration is transformed into a length easily measured on the paper, is easily applied in performing those experiments of which we are about to explain the principle.

Let the point o (Fig. 3) correspond to the moment of electric excitation of a torpedo-nerve, and let the successive points 1, 2, 3, &c., denote successive hundredths of a second, which correspond to very short intervals during which the torpedo apparatus is put in contact with a metallic circuit passing through a frog's foot. In the two first trials, I and 2, after the excitation of the electric nerve, there are no signals recorded; the frog's foot

remaining motionless shows that the discharge of the torpedo has not reached it, because, in fact, the phenomenon has not yet had time to take place. But at the instant 3 the frog moves, which is expressed in the diagram by a vertical stroke; at the instants 4, 5, 6, 7, 8, 9, and 10, the frog receives shocks which are indicated on the diagram by vertical lines; and finally, at the instant 11, and those succeeding, the frog shows no action, whence we conclude that the electric wave of the torpedo was finished before these last trials; and we see that, according to the tracing, the wave began three-hundredths of a second after the instant of nerve-excitation

It was exactly in the same way that M. Marey proceeded to measure the duration of the electric wave in the torpedo. An arrangement easily fixed induced electric action in e (Fig. 4) at constant intervals. A metallic contact, susceptible of being displaced at will, allowed him, during very short intervals of different lengths, to complete the circuit made by the electric wave of the torpedo to reach the frog's-foot-signal. Moreover, to avoid confusion of the curves which were registered by the successive experiments, he took care to change the position of the style each time, so that the curves appeared one under the other in order.

2

1

wave took place at instant I; that in a series of successive trials, each later than the preceding, after the excitation of the nerve, the wave was indicated at the instants 2, 3, 4, 5, and 6; and that at the 7th trial the frog gave no signal. The wave, therefore, was completed. Finally, by bringing the instant of trial nearer to that of nerve-excitation, the wave was retraced in experiments 8, 9, 10, 11, and 12; but in the 13th, occurring too soon after the instant of nerve-excitation, it was shown that the electric wave no longer existed.

The approximation of these measurements necessarily

012345 6 7 8 9 10 11 12

FIG 3.

depends on the number of successive trials, and is more delicate in proportion as they succeed each other more frequently.

3. Each electric wave presents a phase of suddenly increasing intensity, followed by a phase of gradually decreasing intensity.

On examining the tracings of electric waves obtained by the electric-magnetic signal, we observe an apparent contradiction between the indication, the wave-duration furnished by this apparatus, and that which we have just seen determined by the frog's foot. The waves traced

by the signal of Deprez seem to measure not more than one-hundredth of a second second; by Guillemin's method, on the contrary, their duration is much more considerable, being seven-hundredths of a second. This apparent contradiction results from the fact that in the torpedo the waves have not sufficient energy during the whole of their duration to act upon the signal, whereas, from beginning to end of their course they can act upon the frog's muscle, which is much more sensitive. There are, then, in every electric wave, phases of increasing intensity and decreasing intensity which remain to be determined. M. Marey has endeavoured to obtain a tracing of these phases of variable intensity by a modification of the apparatus of M. Deprez. Instead of limiting the excursion of the style between two fixed obstacles, he allowed it an excursion which varies and is proportional to the intensity of the currents acting upon it.

With this object an india-rubber thread, bent over two bridges, was stretched horizontally between the soft iron bars of the armature (Fig. 5). The bars had a groove filed on the top to receive two demi-cylinders of metal which were soldered to the lower part of the armature. In this way the nearer those parts are brought which are subjected to the magnetic attraction the greater is the resistance. Thus if we consider the armature in its different stages when gradually lowered, first it meets the elastic thread with the two demi-cylinders borne on its lower surface, and then the extensibility of the thread is very great. the thread is lowered more and more, it rests on points more and more separated, and becomes less and less extensible. Lower down the india-rubber thread stretched

But as

over the groove made in the soft-iron bars is still less extensible; and finally, when the thread has taken the curvature of the surrounding parts, it opposes any further descent of the armature with the resistance which a stretched thread of india-rubber presents against being pressed or crushed.

This apparatus, to which M. Marey has given the name electrodynamograph, has still to receive further improvements, but even as it is, it has already furnished some interesting evidence as to the decrease in volume of the electric waves from the beginning to the end of the dis

charge; also as to the shape of these waves and the occurrence in the electric tetanus produced by strychnine, &c. Of these different results we shall at present consider only one-the form of a wave is traced by the electro-dynamograph :-an investigation which brings us to the analysis of the wave-phases.

In Fig. 6 the continuous line b is the tracing of a single wave obtained with the electro-dynamograph. From it alone we already have evidence that the ascending phase is much more sudden than the descending phase as also takes place in a muscular shock. We can

account what we have learned by Guillemin's method in the preceding paragraph about the duration of a wave. All that is necessary is to produce downwards the two Produc ascending and descending lines till they intercept between them a distance equal to that which represents (on the time-line) the duration of the whole wave. Thus (Fig. 6) the pointed line a represents the actual position of the axis of abscissas and that part of the tracing which the instrument was unable to trace on account of its insufficient sensibility.

It is true that this curve is only probable, but there

GEOGRAPHICAL NOTES

AT the meeting of the Geographical Society on Monday, Sir H. Rawlinson read a paper, On the road to Merv from the Caspian. After some interesting remarks on the comparative geography of the eastern shores of the Caspian Sea, Sir Henry read some portion of the Russian letters on the earlier stages of the road to Merv, of which a summary appeared in our last issue, and afterwards gave from Russian official documents an account of two ancient cities, the probable relics of Khowrasmian timesMestorian, or Mestdovran, and Meshed. The former in past ages was one of the most important cities of Central Asia, if one may judge from the remarkable aqueducts leading into it, which were the chief arteries of an entire system of irrigation canals thoroughly watering the whole country, and from the number of its buildings, the remains of which exist to this day. The course of the aqueduct was explored by General Lomakine's orders some two or three years ago, and was traced to the Sumbar, a tributary of the Attrek, a length of some sixty-five versts. The city of Mestorian appears to have consisted of a consisted citadel and of two other inclosures with thick, high walls built of enormous bricks. The mass of the débris at the place is so extensive and in such good preservation, that it would be possible, we

points of origin and termination can determine it experimentally, as we have seen done by Guillemin's method (see 2).

We can now understand the reason of the special characteristics presented by currents induced in a secondary coil by the waves of a torpedo discharge which have been passed through an inducting coil. The phase of sudden increase of each wave is alone capable of giving birth to an inducted current.

FRANCOIS FRANCK

(To be continued.)

are told, to make use of it for building a large new town! The bricks, it may be added, are stated to be as hard as stone, and often carved and ornamented with friezes in relief, arabesques, and well-executed inscriptions; the last are sometimes in various colours, illuminated with flowers, and the letters about seven inches in height. Five versts from Mestorian is another remarkable place, known in the country as Meshed; it is, strictly speaking, an ancient necropolis. Here, according to report, is an open coffer holding the sacred books, a hanging lamp, and vases for ablutions, and although in a desert place and wholly unprotected, no one dreams of touching its contents. Sir Henry Rawlinson afterwards dealt at some length with the geography of the country further to the eastward, more especially with that on the northern slopes of the Attock, which is inhabited by three divisions of the Tekké Turcoman tribes.

We regret to record the death, on Saturday afternoon, at a comparatively early age, of Commander G. C. Musters, so well known as the explorer of Patagonia. His work, "At Home with the Patagonians," is at present the best authority we have on this inhospitable country and its people, and Mr. Musters, as readers of the work know, obtained his information by living with the Patagonians for many months as their "king," and it was only by a ruse that he managed to get away from a people who