CHAPTER XX

PLUTO FOUND[46]

Percival had long intended that his Observatory should be permanent, and that his work, especially on the planets, should be forever carried on there with an adequate foundation. Save for an income to his wife during her lifetime, he therefore left his whole fortune in a trust modeled on the lines of the Lowell Institute in Boston, created eighty years earlier by his kinsman John Lowell, Jr. The will provides for a single trustee who appoints his own successor; the first being his cousin Guy Lowell, the next the present trustee, Percival’s nephew, Roger Lowell Putnam. Dr. V. M. Slipher and Mr. C. O. Lampland, who have been at the Observatory from an early time, are the astronomers in charge, carrying on the founder’s principles of constantly enlarging the field of study, and using for the purpose the best instrumental equipment to be procured.

Of course the search was continued for the planet X, but without success, and for a time almost without hope, not only because its body is too small to show a disk, but also by reason of the multitude of stars of like size in that crowded part of the heavens, the Milky Way, where it is extremely difficult to detect one that has moved. It was as if out of many thousand pins thrown upon the floor one were slightly moved and someone were asked to find which it was. Mere visual observation was clearly futile, for no man could record the positions of all the points of light from one night to another. The only way to conduct a systematic search was through an enduring record, that is by taking photographs of the probable sections of the sky, and comparing two of the same section taken a few days apart to discover a point of light that had changed its place—no simple matter when more than one hundred thousand stars showed upon a single plate. This process Percival tried, but although his hopes were often raised by finding bodies that moved, they proved to be asteroids hitherto unknown,[47] and the X sought so long did not appear.[48]

Percival had felt the need of a new photographic telescope of considerable light power and a wider field, and an attempt was made to borrow such an instrument, for use while one was being manufactured, but in vain. Then came the war when optical glass for large lenses could not be obtained, and before it was over Percival had died. After his death Guy Lowell, the trustee, took up the project, but also died too soon to carry it out. At last in 1929 the lens needed was obtained, the instrument completed in the workshop of the Observatory, and the search renewed in March with much better prospects. Photographs of section after section of the region where X was expected to be were taken and examined by a Blink comparator. This is a device whereby two photographs of slightly different dates could be seen through a microscope at the same time as if superposed. But with all the improvement in apparatus months of labor revealed nothing.

After nearly a year of photographing, and comparing plates, Mr. Clyde W. Tombaugh, a young man brought up on a farm but with a natural love of astronomy, was working in this search at Flagstaff, when he suddenly found, on two plates taken January 23 and 29, 1930, a body that had moved in a way to indicate, not an asteroid, but something vastly farther off. It was followed, and appeared night after night in the path expected for X at about the distance from the sun Percival had predicted. Before giving out any information it was watched for seven weeks, until there could be no doubt from its movements that it was a planet far beyond Neptune, and was following very closely the track which his calculations had foretold. Then, on his birthday, March 13, the news was given to the world.

Recalling Percival’s own statement: “Owing to the inexactitude of our data, then, we cannot regard our results with the complacency of completeness we should like,” one inquires eagerly how nearly the actual elements in the orbit of the newly found planet agree with those he calculated. To this an answer was given by Professor Henry Norris Russell of Princeton, the leading astronomer in this country, in an article in the _Scientific American_ for December, 1930. He wrote as follows:

“The orbit, now that we know it, is found to be so similar to that which Lowell predicted from his calculations fifteen years ago that it is quite incredible that the agreement can be due to accident. Setting prediction and fact side by side we have the following table of characteristics:

_Predicted_ _Actual_ Period 282 years 249.17 Eccentricity 0.202 0.254 Longitude of perihelion 205° 202° 30′ Perihelion passage 1991.2 1989.16 Inclination about 10° 17° 9′ Longitude of node not 109° 22′ predicted

“Lowell saw in advance that the perturbations of the latitudes of Uranus and Neptune (from which alone the position of the orbit plane of the unknown planet could be calculated) were too small to give a reliable result and contented himself with the prophecy that the inclination, like the eccentricity, would be considerable. For the other four independent elements of the orbit, which are those that Lowell actually undertook to determine by his calculations, the agreement is good in all cases, the greatest discrepancy being in the period, which is notoriously difficult to determine by computations of this sort. In view of Lowell’s explicit statement that since the perturbations were small the resulting elements of the orbit could at best be rather rough approximations, the actual accordance is all that could be demanded by a severe critic.

“Even so, the table does not tell the whole story. Figure 1[49] shows the actual and the predicted orbits, the real positions of the planet at intervals from 1781 to 1989, and the positions resulting from Lowell’s calculations. It appears at once that the predicted positions of the orbit and of the planet upon it were nearest right during the 19th century and the early part of the 20th, while at earlier and later dates the error rapidly increased. Now this (speaking broadly) is just the interval covered by the observations from which the influence of the planet’s attraction could be determined and, therefore, the interval in which calculation could find the position of the planet itself with the least uncertainty.

[Illustration: Predicted and Actual Orbits of PLUTO]

“In the writer’s judgment this test is conclusive.”[50]

Later observations, and computations of the orbit of Pluto, do not vary very much from those that Professor Russell had when he wrote. Two of the most typical—giving more elements—are as follows:

_Predicted_ _Nicholson and _F. Zagar_ Mayall_ Period 282 years 249.2 248.9 Eccentricity 0.202 0.2461 0.2472 Longitude of 204.9 222° 23′ 20″ .17 222° 29′ 39″ .4 perihelion Perihelion passage 1991.2 1889.75 1888.4 Inclination about 10° 17° 6′ 58″ .4 17° 6′ 50″ .8 Semi-major axis 43. 39.60 39.58 Perihelion 34.31 29.86 29.80 distance Aphelion distance 51.69 49.35 49.36

Except for the eccentricity, and the inclination which he declared it impossible to calculate, these results have proved as near as, with the uncertainty of his data, he could have expected; and in regard to the position of the planet in its orbit it will be recalled that he found two solutions on opposite sides, both of which would account almost wholly for the residuals of Uranus. The one that came nearest to doing so he had regarded as the least probable because it placed the planet in a part of the sky that had been much searched without finding it; but it was there that Pluto appeared—a striking proof of his rigorous analytic method.

But the question of its mass has raised serious doubts whether Pluto can have caused the perturbations of Uranus from which he predicted its presence, for if it has no significant mass the whole basis of the calculation falls to the ground, and there has been found a body travelling, by a marvellous coincidence, in such an orbit that, if large enough, it would produce the perturbations but does not do so.[51] Now as there is no visible satellite to gauge its attraction, and as it will be long before Pluto in its eccentric orbit approaches Neptune or Uranus closely enough to measure accurately by that means, the mass cannot yet be determined with certainty. What is needed are measures of position of the highest possible accuracy of Neptune and Uranus, long continued and homogeneous.

The reasons for the doubt about adequate mass are two.[52] One that with the largest telescopes it shows no visible disk, and must therefore be very small in size, and hence in mass unless its density is much greater, or its albedo far less, than those of any other known planet. The other substantially that the orbits of Uranus and Neptune can be, and are more naturally, explained by assuming appropriate elements therefor, without the intervention of Pluto’s disturbing force. This is precisely what Percival stated in discussing the correctness of the residuals—that it was always possible to account for the motions of a planet, whose normal orbit about the sun is not definitely ascertained, by throwing any observed divergencies either on errors in the supposed orbit, or upon perturbations by an unknown body.

The conditions here are quite unlike those at the discovery of Neptune, for there the existence of the perturbations was clear, because fairly large, and the orbit predicted was wrong because of an error in the distance assumed; and the question was whether the presence of Neptune in the direction predicted, though in a different orbit, was an accident, or inevitable. Here the predicted orbit is substantially the actual one, adequate to account for the perturbations of Uranus if such really exist, and the question is whether they do or not. If not the discovery of Pluto is a mere unexplained coincidence which has no connection with the prediction. Whether among recognized uncertainties it is more rational to suppose a very high density, and very low albedo, with corresponding perturbations of Uranus and Neptune, whose orbits are still imperfectly known, or to conclude that a planet, which would account for these things if dense enough, revolves in fact in the appropriate path, a mere ghost of itself—a phantom but not a force—one who is not an astronomer must leave to the professionals.

In the case of both Neptune and Pluto the calculation was certainly a marvellous mathematical feat, and in accord with the usual practice whereby the discoverer of a new celestial body is entitled to propose its name the observers at Flagstaff selected from many suggestions that of “Pluto” with the symbol [Illustration: ligature, P over L]; and henceforth astronomers will be reminded of Percival Lowell, by the planet he found but never saw.

[Illustration: Decorative wreath]

APPENDIX I

_Professor Henry Norris Russell’s later views on the size of Pluto (written to the Biographer and printed with the writer’s consent)._

Later investigations have revealed a very curious situation. When once the elements of Pluto’s orbit are known, the calculation of the perturbations which it produces on another planet, such as Neptune, are greatly simplified. But the problem of finding Pluto’s mass from observations of Neptune is still none too easy, for the perturbations affect the calculated values of the elements of Neptune’s orbit, and are thus “entangled” with them in an intricate fashion.

Nicholson and Mayall, in 1930, attacked the problem, and found that the perturbations of Neptune by Pluto, throughout the interval from its discovery to the present, were almost exactly similar to the effects which would have been produced by certain small changes in the elements of Neptune’s orbit, so that, from these observations alone, it would have been quite impossible to detect Pluto’s influence. Outside this interval of time, the effects of the perturbations steadily diverge from those of the spurious changes in the orbit, but we cannot go into the future to observe them, and all we have in the past is two rather inaccurate observations made in 1795 by Lalande.[53] If the average of these two discordant observations is taken as it stands, Pluto’s mass comes out 0.9 times that of the Earth, and this determination is entitled to very little weight.

Uranus is farther from Pluto, and its perturbations are smaller; but it has been accurately observed over one and a half revolutions, as against half a revolution for Neptune, and this greatly favors the separation of the perturbations from changes in the assumed orbital elements. Professor E. W. Brown—the most distinguished living student of the subject—concludes from a careful investigation that the observations of Uranus show that Pluto’s mass cannot exceed one-half of the Earth’s and may be much less. In his latest work a great part of the complication is removed by a curiously simple device. Take the sum of the residuals of Uranus at any two dates separated by one-third of its period, and subtract from this the residual at the middle date. Brown proves—very simply—that the troublesome effect of uncertainties in the eccentricity and perihelion of the disturbed planet will be completely removed from the resulting series of numbers, leaving the perturbations much easier to detect. The curve which expresses their effects, though changed in shape, can easily be calculated. Applying this method to the longitude of Uranus, he finds, beside the casual errors of observation, certain deviations; but these change far more rapidly than perturbations due to Pluto could possibly do, and presumably arise from small errors in calculating the perturbations produced by Neptune. When these are accurately re-calculated, a minute effect of Pluto’s attraction may perhaps be revealed, but Brown concludes that “another century of accurate observations appears to be necessary for a determination which shall have a probable error less than a quarter of the Earth’s mass.”

The conclusion that Pluto’s mass is small is confirmed by its brightness. Its visual magnitude is 14.9—just equal to that which Neptune’s satellite Triton would have if brought to the same distance. (Since Pluto’s perihelion distance is less than that of Neptune, this experiment is one which Nature actually performs at times.) Now Nicholson’s observations show that the mass of Triton is between 0.06 and 0.09 times the Earth’s. It is highly probable that Pluto’s mass is about the same—in which case the perturbations which it produces, even on Neptune, will be barely perceptible, so long as observations have their present degree of accuracy.

The value of seven times the Earth’s mass, derived in Percival Lowell’s earlier calculations, must have been influenced by some error. His mathematical methods were completely sound—on Professor Brown’s excellent authority—and the orbit of Planet X which he computed resembled so closely that of the actual Pluto that no serious discordance could arise from the difference. But, in this case also, the result obtained for the mass of the perturbing planet depended essentially on the few early observations of Uranus as a star, made before its discovery as a planet, and long before the introduction of modern methods of precise observation. Errors in these are solely responsible for the inaccuracy in the results of the analytical solution.

The question arises, if Percival Lowell’s results were vitiated in this way by errors made by others more than a century before his birth, why is there an actual planet moving in an orbit which is so uncannily like the one he predicted?

There seems no escape from the conclusion that this is a matter of chance. That so close a set of chance coincidences should occur is almost incredible; but the evidence assembled by Brown permits of no other conclusion. Other equally remarkable coincidences have occurred in scientific experience. A cipher cable-gram transmitting to the Lick Observatory the place of a comet discovered in Europe was garbled in transmission, and when decoded gave an erroneous position in the heavens. Close to this position that evening another undiscovered comet was found. More recently a slight discrepancy between determinations of the atomic weight of hydrogen by the mass-spectrograph and by chemical means led to a successful search for a heavy isotype of hydrogen. Later and more precise work with the mass-spectrograph showed that the discrepancy had at first been much over-estimated. Had this error not been made, heavy hydrogen might not yet have been discovered.

Like this later error, the inaccuracy in the ancient observations, which led to an over-estimate of the mass and brightness of Pluto, was a fortunate one for science.

In any event, the initial credit for the discovery of Pluto justly belongs to Percival Lowell. His analytical methods were sound; his profound enthusiasm stimulated the search, and, even after his death, was the inspiration of the campaign which resulted in its discovery at the Observatory which he had founded.

APPENDIX II THE LOWELL OBSERVATORY _by Professor Henry Norris Russell_

The Observatory at Flagstaff is Percival Lowell’s creation. The material support which he gave it, both during his lifetime and by endowment, represents but a small part of his connection with it. He chose the site, which in its combination of excellent observing conditions and the amenities of everyday life, is still unsurpassed. He selected the permanent members of the staff and provided for the successor to the Directorship after his death. Last, but not least, he inspired a tradition of intense interest in the problems of the universe, and independent and original thought in attacking them, which survives unimpaired.

On a numerical basis—whether in number of staff, size of instruments, or annual budget—the Lowell Observatory takes a fairly modest rank in comparison with some great American foundations. But throughout its history it has produced a long and brilliant series of important discoveries and observations notable especially for originality of conception and technical skill. Percival Lowell’s own work has been fully described; it remains to summarize briefly that of the men whom he chose as his colleagues, presenting it according to its subject, rather than in chronological order.

The photography of the planets has been pursued for thirty years, mainly by the assiduous work of E. C. Slipher, and the resulting collections are unrivalled. Only a small amount of this store has been published or described in print, but among its successes may be noted the first photographs of the canals of Mars, and the demonstration by this impersonal method of the seasonal changes in the dark areas, and of the occasional appearance of clouds. It is a commonplace that any astronomer who wants photographs of the planets for any illustrative purpose instinctively applies to his friends in Flagstaff, and is not likely to be disappointed.

The discovery of Pluto, and incidentally of many hundreds of asteroids, has already been described.

An important series of measurements of the radiation from the planets was made at Flagstaff in 1921 and 1922 by Dr. W. W. Coblentz of the Bureau of Standards and Dr. C. O. Lampland. Using the 40-inch reflector, and the vacuum thermocouples which the former had developed, and employed in measurements of stellar radiation at the Lick Observatory, and working with and without a water-cell (which transmits most of the heat carried by the sunlight reflected from a planet, but stops practically all of that radiated from its own surface), they found that the true “planetary heat” from Jupiter was so small that its surface must be very cold, probably below -100° Centigrade, while that from Mars was considerable, indicating a relatively high temperature. Both conclusions have been fully confirmed by later work.

Spectroscopic observation has been equally successful. In 1912 Lowell and Slipher (V. M.) successfully attacked the difficult problem of the rotation of Uranus. One side of a rotating planet is approaching us, the other receding. If its image is thrown on a spectroscope, so that its equatorial regions fall upon the slit, the lines of the spectrum will be shifted toward the violet on one edge, and the red on the other, and will cross it at a slant instead of at right angles. This method had long before been applied to Jupiter and to Saturn and its rings, but Uranus is so faint as to discourage previous observation. Nevertheless, with the 24-inch reflector, and a single-prism spectrograph, seven satisfactory plates were obtained, with an average exposure of 2½ hours, every one of which showed a definite rotation effect. The mean result indicated that Uranus rotates in 10¾ hours, with motion retrograde, as in the case of his satellites. This result was confirmed five years latter by Leon Campbell at Harvard, who observed regular variations in the planet’s brightness with substantially the same period.

It has been known since the early days of the spectroscope that the major planets exhibit in their spectra bands produced by absorption by the gases of their atmospheres, and that these bands are strongest in the outer planets. Photographs showing this were first made by V. M. Slipher at the Lowell Observatory in 1902. To get adequate spectrograms of Neptune required exposures of 14 and 21 hours—occupying the available parts of the clear nights of a week. The results well repaid the effort. The bands which appear faintly in Jupiter are very strong in Uranus, and enormous in Neptune’s spectrum, cutting out great portions of the red and yellow, and accounting for the well-known greenish color of the planet. Only one band in the red was present in Jupiter alone.

For a quarter of a century after this discovery those bands remained one of the most perplexing riddles of astrophysics. The conviction gradually grew that they must be due to some familiar gases, but the first hint of their origin was obtained by Wildt in 1932, who showed that one band in Jupiter was produced by ammonia gas, and another probably by methane. These conclusions were confirmed by Dunham in the following year, but the general solution of the problem was reserved for Slipher and Adel, who, in 1934, announced that the whole series of unidentified bands were due to methane. The reason why they had not been identified sooner is that it requires an enormous thickness of gas to produce them. A tube 45 meters long, containing methane at 40 atmospheres pressure, produces bands comparable to those in the spectra of Saturn. The far heavier bands in Neptune indicate an atmosphere equivalent to a layer 25 miles thick at standard atmospheric pressure. The fainter bands though not yet observed in the laboratory, have been conclusively identified by the theory of band-spectra. Ammonia shows only in Jupiter and faintly in Saturn; the gas is doubtless liquefied or solidified at the very low temperatures of the outer planets.

The earth’s own atmosphere has also been the subject of discovery at Flagstaff. The light of a clear moonless sky does not come entirely from the stars and planets; about one-third of it originates in the upper air, and shows a spectrum of bright lines and bands. The familiar auroral line is the most conspicuous of these, but V. M. Slipher, making long exposures with instruments of remarkably great light-gathering power, has recently detected a large number of other bands, in the deep red and even the infra-red. Were our eyes strongly sensitive to these wave-lengths, the midnight skies would appear ruddy.

Just as the first rays of the rising sun strike the upper layers of the atmosphere many miles above the surface, new emission bands appear in the spectrum—to be drowned out soon afterwards by the twilight reflected from the lower and denser layers; and the reverse process is observable after sunset.

The origin of these remarkable and wholly unexpected radiations is not yet determined.

The spectrograph of the Observatory was also employed in observations of stars, and again led to unexpected discoveries. In 1908, while observing the spectroscopic binary Beta Scorpii, V. M. Slipher found that the K line of calcium was sharp on his plates, while all the others were broad and diffuse. Moreover, while the broad lines shifted in position as the bright star moved in its orbit, the narrow line remained stationery. Hartmann, in 1904, had observed a similar line in the spectra of Delta Orionis, and suggested that it was absorbed in a cloud of gas somewhere between the sun and the star. Slipher, extending his observations to other parts of the heavens, found that such stationery calcium lines were very generally present (in spectra of such types that they were not masked by heavier lines arising in the stars themselves), and made the bold suggestion that the absorbing medium was a “general veil” of gas occupying large volumes of interstellar space.

This hypothesis, which appeared hardly credible at that time, has been abundantly confirmed—both by the discovery of similar stationery lines of sodium, and by the theoretical researches of Eddington,—and no one now doubts that interstellar space is thinly populated by isolated metallic atoms presumably ejected from some star in the remote past, but now wandering in the outer darkness, with practically no chance of returning to the stars.

To secure satisfactory spectroscopic observations of nebulae is often very difficult. Though some of these objects are of considerable brightness, they appear as extended luminous surfaces in the heavens, and in the focal plane of the telescope. The slit of a spectroscope, which must necessarily be narrow to permit good resolution of the lines, admits but a beggarly fraction of the nebula’s light. To increase the size of the telescope helps very little, for, though more light is collected in the nebular image, this image is proportionately increased in area, and no more light enters the slit than before.

For the gaseous nebulae, whose spectra consist of separate bright lines, there is no serious difficulty; but the majority of nebulae have continuous spectra, and when the small amount of light that traverses the slit is spread out into a continuous band, it becomes so faint that prohibitively long exposures would be required to photograph it. It was at the Lowell Observatory that Dr. V. M. Slipher first devised a way of meeting this difficulty.

By employing in the camera of the spectrograph (which forms the image of the spectrum on the plate) a lens of short focus, this image became both shorter and narrower, thereby increasing the intensity of the light falling on a given point of the plate in a duplicate ratio. Moreover, since with this device the image of the slit upon the plate is much narrower than the slit itself, it became possible to open the slit more widely and admit much more of the light of the nebula, without spoiling the definition of the spectral lines.

This simple but ingenious artifice opened up a wholly new field of observation, and led to discoveries of great importance.

Within the cluster of the Pleiades, and surrounding it, are faint streaky wisps of nebulosity, which have long been known. One might have guessed that the spectrum, like that of some other filamentous nebulae, would be gaseous. But when Slipher photographed it in December 1912 (with an exposure of 21 hours, on three successive nights) he found a definite continuous spectrum, crossed by strong dark lines of hydrogen and fainter lines of helium—quite unlike the spectrum of any previously observed nebula, but “a true copy of that of the brighter stars in the Pleiades.” Careful auxiliary studies showed that the light which produced this spectrum came actually from the nebula. This suggested at once that this nebula is not self-luminous, but shines by the reflected light of the stars close to it. This conclusion has been fully verified by later observations, at Flagstaff and elsewhere. It is only under favorable conditions that one of these vast clouds (probably of thinly scattered dust) lies near enough to any star to be visibly illuminated. The rest reveal themselves as dark markings against the background of the Milky Way.

Similar observations of the Great Nebula of Orion showed that the conspicuous “nebular” lines found in its brighter portions faded out in its outer portions, leaving the hydrogen lines bright, while, at the extreme edge, only a faint continuous spectrum appeared. This again has been fully explained by Bowen’s discovery of the mechanism of excitation of nebular radiation by the ultra-violet light from exceedingly hot stars, and affords a further confirmation of it.

But the most important contribution of the new technique was in the observation of the spiral nebulae. Their spectra are continuous and so faint that previous instruments brought out only tantalizing suggestions of dark lines. With the new spectrograph, beautiful spectra were obtained, showing numerous dark lines, of just the character that might have been expected from vast clouds of stars of all spectral types. This provided the first definite indication of one of the greatest of modern astronomical discoveries—that the white nebulae are external galaxies, of enormous dimensions, and at distances beyond the dreams of an earlier generation.

By employing higher dispersion, spectra were secured which permitted the measurement of radial velocity. The first plates, of the Andromeda Nebula, revealed the almost unprecedented speed of 300 kilometers per second toward the Sun. Later measures of many other nebulae showed that this motion was, for a nebula, unusually slow, but remarkable in its direction, for practically all the others were receding.

Similar measures upon globular star-clusters showed systematic differences in various parts of the heavens, which indicated that, compared with the vast system of these clusters, the Sun is moving at the rate of nearly 300 kilometers per second—a motion which is now attributed to its revolution, in a vast orbit, about the center of the Galaxy, as a part of the general rotation of the latter.

The velocities of the nebulae reveal substantially the same solar motion, but, over and above this, an enormous velocity of recession, increasing with the faintness and probable distance of the nebulae.

This, again, was a discovery of primary importance. It has been confirmed at other observatories and observations with the largest existing telescope have revealed still greater velocities of recession in nebulae too faint to observe at Flagstaff. How this has led to the belief that the material universe is steadily expanding and that its ascertainable past history covers only some two thousand millions of years, can only be mentioned here.

This is a most remarkable record for thirty years’ work of a single observatory with a regular staff never exceeding four astronomers. But its distinction lies less in the amount of the work than in its originality and its fertile character in provoking extensive and successful researches at other observatories as well.

All this is quite in the spirit of its Founder, and, to his colleagues in the science, makes the Observatory itself seem his true monument. His body lies at rest upon the hill, but, in an unquenched spirit of eager investigation, his soul goes marching on.

FOOTNOTES

[1]It is dated Boston, August 24th, but the year does not appear. She was abroad and he at home in the summers of 1882 and 1887.

[2]Before leaving Korea he spent two delightful weeks at the Footes’.

[3]This came about a month later than ours.

[4](_Atlantic Monthly_, Nov. 1886, “A Korean Coup d’Etat”).

[5]“The Life and Letters of Lafcadio Hearn by Elizabeth Bisland,” Vol. I, p. 459.

[6]_Ib._, Vol. II, p. 28.

[7]_Ib._, Vol. II, p. 30.

[8]_Ib._, Vol. II, p. 487. See also pp. 479, 505. Percival’s “Occult Japan” a study of Shinto trances, published in 1894, he did not like at all. It struck him only “as a mood of the man, an ugly supercilious one, verging on the wickedness of a wish to hurt—there was in ‘The Soul of the Far East’ an exquisite approach to playful tenderness—utterly banished from ‘Occult Japan.’” _Id._, pp. 204, 208. By this time Hearn seems to have come to resent criticism of the Japanese.

[9]The exact elevation proved to be 12,611.

[10]These discoveries have since been doubted.

[11]The theory of the gradual loss of water is very doubtful, but Percival’s main conclusions depend on the present aridity of the planet, not on its assumed history.

[12]In a lecture shortly before his death he said: “Where Schiaparelli discovered 140, between 700 and 800 have been detected at Flagstaff.”

[13]Thereafter the equipment of the Observatory was steadily enlarged—notably by a 42-inch reflector in 1909—until now there are five domes, and much auxiliary apparatus.

[14]Vol. 19, No. 218.

[15]Percival’s statement of this may be found also in “Mars as the Abode of Life,”