ods were put to the test of experiment with a considerable degree of success.

Some doubt is said to exist as to whether Newton was the real author of this mistake, but, as Professor Poynting remarked in a lecture at the Royal Institution a few years ago, there is something not altogether unpleasing in the belief that even Newton could make a mistake. His faulty arithmetic showed that there was, at any rate, one quality which he shared with his fallible fellow-men.

When the attractive force of a mountain is to be studied, the experiment, in its simplest form, is somewhat as follows: A weight hanging at the end of a thread-that is, a plumb-line more or less similar to the plumb-line employed by a mason, but far more sensitive and provided with more exact means of measurement-is placed first in some suitable position not too far away from the mountain, but well out of the range of its attraction, and its position noted on a scale of divisions when it hangs freely suspended, and, therefore, perpendicular to the earth's surface. The plumb-line is then brought up as close as may be to one side of the mountain. When this is done the plumb-line is found to be drawn a little to one side of its previous line of suspension-that is to say, a little out of the perpendicular and towards the mountain. The amount of this displacement is measured on the scale of divisions, and the length of the plumb-line is also measured. From these data the astronomer can calculate the ratio of the horizontal pull of the mountain to the pull of the earth.

Finally, the mountain is most carefully surveyed, and the densities of pieces of the rock of which it is composed are measured. Knowing these densities and the volume of the mountain we can estimate the mass of the mountain in pounds or kilograms, according to the system selected; and

when this is done we know the mass of the mountain, the pull of the mountain, the pull of the earth, and their distances, and from these, knowing the law of gravitation, quoted above, we can deduce the other quantity involved, the mass of the earth.

The first investigator to actually determine the mean density of the earth by this method was M. Bouguer, who was a member of one of two scientific commissions sent out by France about 1740 to measure the lengths of degrees of latitude in Peru and Lapland-that is, at points near to and remote from the equator-in order to settle finally the shape of the earth, whether it is flattened at the poles, as Newton supposed, or drawn out, as had then lately been suggested. The members of these commissions, which, by the way, settled the question in favor of Newton's views, did not confine themselves to investigating the shape of the earth; and M. Bouguer, in particular, seized the opportunity of testing the "mountain mass method" of weighing the earth thus afforded him by his visit to the great mountains of the Andes. M. Bouguer made two distinct sets of measurements. In the first he studied the swing of a pendulum at the sealevel, then at a point 10,000 feet higher, on the great plateau on which Quito stands, and, finally, on the top of Pichincha, which is about 6,000 feet above Quito. He knew that if a pendulum were lifted to a great height above a wide plain or over the open sea, say, for example, by means of a balloon, its swing would gradually grow slower as gravity decreased at the higher levels; and he calculated from the swing of his pendulum at Quito that gravity there was greater than the calculated amount for the height at which he worked, owing to the down pull of the great tableland beneath him.

Bouguer's second set of observations

was made near Chimborazo, a mountain 20,000 feet high, by the plumb-line method as described in outline above, only in a far more refined form. His difficulties were very great, for he was obliged to work above the line of perpetual snow. His labors began with a troublesome and even perilous journey of many hours over rocks and snow-fields, and when the site selected for the first set of observations was reached he had to fight against snowfalls, which threatened to bury the instruments, the tents, and even the observers themselves. At the second station, which was below the snow-line, he hoped for better conditions; but here he encountered gales of wind, and it was still so cold as to hinder the working of his instruments. Under these circumstances it is not surprising that the results obtained were, as Bouguer himself recognized, of little permanent scientific value. The cause for wonder was that he got any results at all. But his time and labors were not wasted. His observations proved that the earth, as a whole, is denser than the mountains upon it; that it is not a mere hollow shell, as some people in those days still supposed, nor yet a hollow globe filled with water, as others had insisted. Besides, he had broken new ground, and before very long his experiments were repeated under more favorable conditions and with better results.

The next experiment by the mountain-mass method was made in the neighborhood of Schiehallion, in Perthshire, thirty years later, under the auspices of the Royal Society, who, at the instance of Maskelyne, then Astronomer Royal, appointed "a committee to consider of a proper hill whereon to try the experiment, and to prepare everything necessary for carrying the design into execution."

A few years ago the inhabitants of a certain remote island were considerably

excited by the absurd proceedings of a party of visitors to their shores, who did many things which seemed stupid, not to use a stronger term, to the islanders, and at length lost the last vestiges of their respect by boiling water in tin pots on a mountain top in order to find out how high the mountain was. I have sometimes wondered what the hard-headed natives of Perthshire can have thought of the party of gentlemen who came to Schiehallion about the year 1774, and proceeded to watch plumb-lines hanging in the air, and to peep at stars through telescopes in order to discover the weight of the earth. But, be that as it may, after two months or so spent in observing, and two years more in surveying the mountain, making contour maps giving the volume and distance of every part of it from the two stations at which the observations of its attraction had been made for Maskelyne did not follow the method of Bouguer exactly, but observed the attraction of the mountain from two opposite sidesand after determining the density of various fragments of the rock of which Schiehallion is composed, Maskelyne and his colleagues came to the conclusion that the mean density of the earth must be four and a half times that of water-that is, that the earth must contain four and a half times as much matter as a globe of water of its own size, or, again, that its mass must be equal to that of a globe of water four and a half times as big as the earth. This value was presently raised to five, as the result of further determinations of the density of the rock, and we have every reason to suppose that this latter value is not very far from the truth.

I should tire my reader were I to go further into this part of our subject and describe one by one the various experiments following more or less similar lines that have been made since the completion of Maskelyne's

celebrated experiments. Moreover, interesting and ingenious as these experiments were, all were vitiated by a fatal defect. The accuracy of the conclusions reached depends in every case on two chief points. First, correct measurements of the attractive forces of the mountain masses studied are necessary, and this, doubtless, was attained in many if not in every one of the various investigations. Secondly, a fairly correct knowledge of the density of the rocks forming the mountains is required, and here the experiments in every case break down. We cannot learn with certainty the true mean densities of the rocks forming a mountain; at the best we can only make rough guesses at them. Consequently, of late years the attention of astronomers has been turned to the other methods to which I have alluded. These, though equally difficult to carry out, are not subject to this fatal objection. I may point out, however, before we proceed, that it would be quite reasonable, now the weight of the earth has been fixed by these other and sounder methods, to turn the above experiment about and apply the results obtained to the complementary problem of "weighing mountains."

"Of all experiments," exclaimed Professor Boys, a few years ago, in the course of a lecture at the Royal Institution, "the one which has most excited my admiration is the famous experiment of Cavendish." For this method of weighing the earth no costly expeditions to distant mountains, and no elaborate surveys requiring years for their performance are demanded. For the "Cavendish experiment," in fact, nothing is wanted but a few bits of wire, some strips of wood, balls of metal, and a case to protect the apparatus from "the wind," as Cavendish expressed it. If you possess these and certain other similar trifles, and if you possess, also, the genius for experi

menting of a Cavendish or of a Boys, you can weigh the earth. If, in addition, you possess one of the wonderful silica threads discovered a few years ago by Professor Boys, you can construct an apparatus hardly too big to go inside a man's hat-box, with which you may do the thing to a nicety.

That great though most eccentric man, the Honorable Henry Cavendish, was, as I have said, the first to carry out in a laboratory the operation of weighing the earth, but the actual originator of the Cavendish experiment was the Rev. John Michell, who constructed the necessary apparatus, but died before he had an opportunity of testing the value of his ideas by mak ing an experiment. After Mr. Michell's death his apparatus passed into the hands of Dr. Wollaston, and he handed it on to Mr. Cavendish, who, after making some modifications, performed the first "Cavendish experiment" in 179798. Cavendish found the mean density of the earth to be 5.45 times that of water, and we may take it that this was the first really trustworthy measurement. The experiment, in outline, was as follows:

Two equal balls of lead, each two inches in diameter, were attached to the remote ends of a light wooden rod six feet long, which was suspended horizontally at its centre, by means of a wire forty inches long, inside a narrow wooden case to protect it from draughts. Outside the case two much more massive balls, also of lead, twelve inches in diameter, were suspended by rods from a beam, which worked on a pivot. This pivot was placed above the wire by which the rod carrying the small balls was suspended, so that the large balls could be swung at will into various positions outside the case. For example, they could be placed transversely by putting the two beams at right angles to one another, or brought close up to the smaller balls,

counterpoised by weights, do you think the addition of one middle-sized boy to the population of the scale pan would seem to make much difference to a man who was weighing them? That is the sort of difference that had to be measured-a difference of one part in seventy or eighty million parts. It will give a still better idea of the degree of perfection to which the art of weighing was brought by Professor Poynting if I add that the degree of accuracy was such as would be required, in this imaginary experiment, to detect whether or no the boy had both his boots on.

But splendid as this work was, the high-water mark was reached, perhaps, by Professor Boys in a recent repetition of the Cavendish experiment. Cavendish, as I have said, suspended the beam of his "torsion balance," as such an instrument as that used by Cavendish is called, by means of a fine wire, and the accuracy of his results depended on the elasticity of this wire. Now, unfortunately, metallic wires are not perfectly elastic, and when frequently used are subject to "fatigue"; and so there was a defect in the experiment, which remained uncorrected until a few years ago, when Professor Boys discovered how to produce threads not liable to this fault. These astonishing threads, which were made of melted quartz, are finer by far than the finest wire-so fine, in fact, that a single grain of sand spun into one of them might yield a thread a thousand miles long; moreover, they surpass The Cornhill Magazine.

steel in strength, and are marvellously elastic. Armed with quartz threads Mr. Boys was able to reduce the size of the Cavendish apparatus, and at the same time greatly to increase its sensibility. This and great personal skill enabled him to make what is probably the best measurement yet obtained of the earth's mean density-viz. 5.5270. And so we find that the work of Maskelyne, the work of Cavendish, the work of Poynting, that of Boys, and, indeed, that of half a score others about whom I have said nothing, supports, almost without an exception, Newton's guess at the weight of the earth.

We are often told that we live in a material age, that the days of chivalry are gone, and that even science devotes herself to-day to the merely useful, and is too apt to neglect the search after abstract truth. Perhaps this incomplete recital of the progress of a great research during a period of nearly two centuries, including as it does some splendid contributions which have been made within quite recent years, may serve as a reminder that though science reveals herself to many of us chiefly through her more obviously useful and profitable discoveries and inventions, yet those who look for them will still find among us not a few men as ready as any of their predecessors to devote days and nights to hard labor for no other fee than the hope of discovering a new truth, overthrowing an ancient error, or extending in some other way the boundaries of knowledge.

W. A. Shenstone.