INTRODUCTION.

A SCIENTIFIC expedition may be said to have two histories. The one treats of the special object of the expedition, the other of the personal adventures of those concerned in it. It is only the former which finds permanent record in the Transactions of scientific societies: the other too often remains unwritten.

For many reasons I think this is a matter of regret. Mere details of observations are never looked at, except by a very limited number of specialists; to the general public such details are meaningless as well as inaccessible; whilst the ordinary student usually accepts the result merely as he finds it quoted in some standard work or text-book.

It is not because popular accounts of such expeditions do not interest a sufficient circle of readers that they have not been more frequently written, but rather, I think, because the faculties of original research and popular exposition are seldom united in the same individual. Besides this, I have found in my own experience, that on such expeditions there is so much actual work to be done, and the hours are so completely filled with it, that there is neither time nor inclination to write a diary. Thus, before the story can be committed to writing, it has lost its crispness—the interest has faded, and, from treacherous memory, incident is wanting to complete the narrative. On my expedition to Ascension last year, however, I had the good fortune to be accompanied by my wife, who found much pleasure and interest in making a daily record of our life and work there. This little book, compiled from her journal, she now lays before the public with much diffidence. It is an honest endeavour to tell a true story, and add somewhat to a neglected class of literature; as such, she hopes that the faults incidental to a first work will meet with lenient judgment.

The story can boast of no stirring interest, no thrilling adventures by land or sea. It must derive its interest chiefly from its truthfulness as a record of an attempt to solve a great problem, viz., the distance of the Earth from the Sun. The nature of this problem my wife explains in Chapter I. sufficiently, I think, to make it of interest to the many that would gladly learn something of the history of scientific progress, but who are often deterred from so doing by the minute details and heavy technicalities with which its every step is necessarily encumbered.

To the best of her ability she has given one side of the history of one step.

It now remains for me to preface her chronicle by a brief outline of the labours of those who have worked before in the same field of research.

The first attempt to measure the Sun's distance was made as follows.

      In the diagram suppose S the Sun, M the Moon, and E an observer on the Earth. When the angle S E M is a right angle, the Moon will be exactly half full. If it is less than a right angle the Moon will appear less than half illuminated to an observer at E, and vice versa. Hence, if the angle S E M is accurately measured at the instant when the Moon is half full, the proportions of the triangle S E M will be known; because the angle S E M being a right angle, and the angle having been determined by measurement, one side and two angles are known; the triangle S E M can therefore be drawn on paper, or its proportions may be computed mathematically.

Since E M represents the distance of the Earth from the Noon at the instant of observation, the proportion of this distance to that of the Sun (represented by the line E S) is determined.

In this way Aristarchus of Samos concluded that the Sun was nineteen times more distant than the Moon. This distance we now know to be more than twenty times too small—and the reason of his failure was twofold. 1st. Because, from the irregularity of the Moon's surface, it is almost impossible to estimate when she is exactly half full; 2nd. Because his means of measuring the angle of S E M were rude and imperfect. The result of Aristarchus was however adopted by astronomers till the time of Kepler.

About the year 1620, from the observations of Tycho Brahe on the planet Mars, Kepler concluded that the distance of the Sun must amount at least to 1800 diameters of the Earth, upon which he was upbraided by his friend Cruger "for removing the Sun to such a huge distance."

The first approximation to a true determination was the result of an expedition organized by the French astronomer, Cassini. Richer had been sent to South America by the French Academy. On the 1st. of October, 1672, the planet Mars approached very close to a bright star (y Aquarii), and, from observations made by Richer in Cayenne, and by Picard and Romer in France, Cassini concluded that the Sun's distance must be at least 86 millions of miles.

The results were, however, liable to very considerable uncertainty. This was due to the imperfect instruments of the time, by means of which it was hardly possible to measure angles with the required accuracy.

The first satisfactory step in advance is due to the Abbe de la Caille, whose celebrated expedition to the Cape of Good Hope took place in the year 1740. There he made a large number of observations of Mars, and from those compared with corresponding observations in the Northern Hemisphere, deduced 81 6/10 millions of miles for the Sun's distance. He afterwards combined this result with similar observations on the planet Venus, and arrived at 81 5/10 millions of miles as a final result.

In the meantime, however, the English astronomer Halley had gone to St. Helena, where he observed the Transit of Mercury on the 28th of October, 1677 (see Chap. III. p. 33), and this suggested to him the method of determining the Sun's distance by the Transits of Venus which would take place in the years 1761 and 1769.

In two remarkable memoirs presented to the Royal Society of London in 1691 and 1716, he pointed out the great advantages of this method, and urged upon astronomers the necessity of providing for the complete and accurate observation of these phenomena. Accordingly, in 1761, England sent Maskelyne to St. Helena, and Mason and Dixon were dispatched to Sumatra. The two latter astronomers were so delayed by the way, that, fearing they would not reach the appointed station in time, they decided to remain at the Cape of Good Hope; and the decision proved a fortunate one.

The St. Petersburg Academy of Sciences sent Chappe to Tobolsk, and Rumowski to Selinghinsk near Lake Baikil in Siberia.

The French sent Pingre to the Island of Rodriguez, and Le Gentil should have observed at Pondicherry.

Poor Le Gentil! He duly reached Mauritius on the 10th of July, 1760—nearly a year before the Transit. War having meanwhile broken out between France and England he was unable to reach Pondicherry; so he resolved to go to the Island of' Rodriguez instead, to join Pingre, who was already there. When on the point of starting for Rodriguez, he learned that a French frigate was about to sail from Mauritius for the coast of Coromandel. Le Gentil decided to avail himself of the opportunity thus offered to reach the point chosen by the Academy; but one delay after another occurred and it was not until the middle of March that he sailed again from Mauritius. There was not much time to be lost, for the Transit would occur on the 6th of June. Detained by frequent calms he did not reach the coast of Malabar till the 24th of May. Still there might have been time enough to prepare for the observation, had not the commander of the frigate learned that the English were masters of Mahe and Pondicherry. His only chance to escape capture was to make off as quickly as possible. This he did; steering a course for Mauritius again, to Le Gentil's utmost despair.

The 6th of June arrived. The sky was gloriously clear. From the deck of the vessel Le Gentil made the best observations he could; but, from so unsteady a platform, they could be of little value to science.

Other observers had better fortune; but the results when computed proved far from satisfactory. Different astronomers obtained results varying from 81 1/2 to 96 1/2 millions of miles; not because there was any error in the method, but because the observations were discordant.

The difficulties of actual observation proved to be far greater than had been anticipated. Instead of precise phenomena at contact, only a gradual merging, or a gradual separation of the limbs of Venus and the Sun could be observed. Thus different interpretations could be put upon the language of the observers; and, according to the interpretations, so was the result.

But the failure of the Transit of 1761 only urged to new effort for that of 1769. Astronomers throughout the world felt that if the opportunity, which would then occur, was lost, another so favourable for determining the Sun's Distance would not occur again for 105 years. Accordingly, the most strenuous exertions were made to provide for its proper observation. Encke has well said, "Whatever may be the future judgment as to the actual issue, posterity will never be able to reproach either the astronomers or the governments of that period with having g neglected to call sufficiently careful attention to the more important points, or with having failed to further and support scientific efforts with sufficient readiness."

It would occupy too much space to follow the adventures of all the observers, but some of them cannot be passed over without mention.

The French astronomer, Le Gentil—whose endeavours to observe the Transit of Venus in 1761 were defeated in the way already described—had no sooner returned to Mauritius than he set out again for Pondicherry, determined to wait there, for eight years, till the next Transit of Venus.

The eventful 3rd of June, 1769, at last arrived.

The morning was fine, and everything promised a happy issue. But, just as the critical moment approached, an unfortunate cloud eclipsed the Sun, a torrent of rain descended, and the fruit of eight years' waiting was lost. Le Gentil had profitably employed his time in studying the astronomy of the Brahmins, so his eight years in Pondicherry had been well occupied; but the agony of disappointment he must have felt at the defeat of his noble endeavours cannot but enlist the sympathy of all who know his story.

It was intended that the French astronomer Chappe, who in 1761 had observed the Transit of Venus at Tobolsk, should now observe it at the Solomon Islands. The Spanish Government, however, refused the necessary permission, but offered to convey him, along with two Spanish observers, to Mexico or California by a Spanish fleet then about to sail for South America.

This offer was accepted. Chappe, with his Spanish colleagues, selected Cape Lucas, in California, and there observed the Transit successfully. But he did not live to tell the tale at home. The plague visited the district, and Chappe was one of its first victims. Three days after the Transit, he was attacked by the malady, but had partly recovered, when his love of science led him to commit an imprudence which brought him to the grave. Despite his feeble condition, he passed the night of the 18th of June in observing an eclipse of the moon. A relapse ensued, and on the 1st. of August he died. A short time before his death he said to his friends: "know I have but a short time to live, but I have fulfilled my mission, and I die content."

The English Government took a bolder course, and did not wait for the permission of the Spaniards to visit the South Seas. The celebrated Captain Cook in command of the Endeavour, with Green (a pupil of Bradley) for astronomer, and Solander (a pupil of Linne) for naturalist, sailed on the 22nd of September, 1768 for an unknown destination. The result was one of the most brilliant and successful scientific expeditions ever undertaken. The Transit of Venus was successfully observed by Green and Cook at Tahiti, one of the Sandwich Islands; and much valuable work in connection with Natural History, Terrestrial Magnetism, and Hydrography, was accomplished.

Another English expedition was sent to Hudson's Bay. There Dymock and Wales, after encountering a, good many hardships, successfully observed the Transit; and their observations acquired a special value, because it unfortunately happened that those made at the other important northern station were liable to considerable suspicion.

Father Hell, with his assistant, Father Sainovicz, was invited by the King of Denmark to observe the Transit of Venus in his dominions. In the month of June, 1768, they left Copenhagen, accompanied by Borgrewing, a Danish observer. They reached Wardoebuus, in the north of Lapland, on the 11th of October 1768. Here they passed the winter, and duly observed the Transit of Venus in 1769. But numerous circumstances tended to throw suspicion on Hell's observations. In the first place, without any sufficient reason, he suppressed his observations for nine entire months, and many eminent astronomers did not hesitate to accuse him of having fabricated or changed them.

In 1834, his original papers were presented to the Vienna Observatory by Baron Mihch-Bellinogbansen, into whose hands they had come through the death of his uncle, Baron von Penkler, an intimate friend and patron of Father Hell.

Professor Littrow's investigation of these papers led to the discovery of Father Hell's original note-book for the days June 2-4, 1769. "These notes fully corroborate and justify previous suspicions. The chief figures, especially the times of entrance upon the solar disc, had been for the most part erased, and with a darker coloured ink. Two passages, the one relating to the observations of Sainovicz, the other to those of Borgrewing, had been so thoroughly obliterated, that Professor Littrow was only able to conjecture the three first letters of the one and the first and last letter of the other. From an investigation of such figures as remained legible and unaltered, he succeeded in finding one observation of the Ingress by Borgrewing, and one of Egress by Hell, upon which reliance appears warrantable."

"Although in reply to Lalande, Father Hell had publicly offered to exhibit the original note-book, free from erasures, and giving observations just as finally published by him, Littrow found both clear and undefaced documents containing the quantities as prepared for publication, and this note-book, which was as manifestly not designed for press. It contains remarks, notes, and comments in chronological order; the handwriting is unequal and frequently changing observations never made public are here noted down, together with many jottings and memoranda which could not have been intended for the public. The important observations were chiefly obliterated with great care and thoroughness as were also sundry remarks concerning them. There can be no doubt that the evidence is sufficient to establish this notebook as being the identical one used at Wardoehuus, and that this establishment of identity discredits the published observations and the truthfulness of Father Hell, but provides few new figures upon which reliance may be placed."

Encke also found another proof of Father Hell's dishonesty. An eclipse of the Sun occurred soon after the Transit of Venus of 1769, and afforded an excellent means of checking the longitudes of the stations. Father Hell observed this eclipse, and fortunately did not change the original record in his note-book. The time he published, however, differed from the time he recorded by 41 seconds; for, in his desire to publish better observations than be knew how to make, he altered his observation to agree with his computation, which proved to have been founded on erroneous elements. His original record was afterwards found to be a good observation.

Such revelations must throw the gravest doubt upon all the observations of Father Hell, and give to the observations of Dymock and Wales an exceptional importance.

Although the results obtained for the distance of the Sun from the Transit of 1769 did not differ so widely as those of 1761, still the agreement was by no means satisfactory.

Thus from the numerous results may be quoted the following

Sun's
Distance
Lalande obtained 96 2/10 millions of miles
Father Hell " 93 9/10 "
Hornsby " 93 1/10 "
Euler " 92 1/10 "
Pingre " 92 1/10 "
Laplace " 92 8/10 "

In 1835 Encke published his famous discussion of the Transits of Venus of 1761 and 1769. He found

From the Transit of 1761 95 8/10 millions of miles.
"           " 1769 95 2/10 "
From both Transits combined 95 36/100 "

Encke's discussion met with the general approval of astronomers at the time, and for many years was accepted as the standard determination of the Sun's distance.

In 1832, Henderson observed, at the Cape of Good Hope, the favourable Opposition of Mars of that year, and his observations, combined with similar ones at Greenwich, Cambridge and Altona in the Northern Hemisphere, gave

90 6/10 millions of miles

for the Sun's distance. The results were, however, not very accordant, and were not generally accepted as satisfactory.

In 1847 Professor Gerling proposed the observation of the planet Venus, at observatories in the Northern and Southern hemispheres, as a good means of determining the Sun's distance. He argued that as Venus approaches nearer to the Earth than Mars, she presents a more favourable opportunity for determining parallax. He also contended that the delicate and faint crescent form of Venus (like a very young moon when the planet is near conjunction) formed a telescopic object capable of the most accurate measurement.

Professor Gerling's idea, however, did not assume practical shape till it was taken up by a zealous observer, Lieut. Gilles of the United States Navy. He applied to his chief to ask from Congress a grant of 1000l. for the expenses of an expedition to Chili. He proposed to observe there, not only the planet Venus, when near inferior con unction (as suggested by Professor Gerling), but also the Oppositions of Mars which would occur in the years 1850 and 1852.

His proposal was accepted.

Gilles went to Chili, and there made more than 200 series of observations of Mars and Venus together. To combine with this splendid mass of work, only 28 corresponding observations were made in the Northern Hemisphere, and even these do not appear to have been made with exceptional care, nor to possess the accuracy required in so delicate a research. The want of sympathy and support from astronomers, which Lieut. Gilles met with, is a blot upon the history of astronomy, and Dr. Gould, in the work I have already quoted, has well said, "It is impossible to refrain from the expression of deep regret that, from all the observations of the well-equipped and richly-endowed observatories of the Northern Hemisphere, so few materials could be found toward rendering available, according to its original purpose, an expedition to which so much labour and enthusiasm had been consecrated, and to which an accomplished observer, already known for the precision of his measurements, had devoted his entire energies during so long a sojourn (three years) moreover, after the preparation and wide dissemination of ephemerides and charts of the comparison stars for both the planets during the whole period." Dr. Gould has reduced the whole mass of observations with a loving care, and obtains the result

96 1/10 millions of miles

a result confessedly unsatisfactory from the non-agreement of the various observations.

Meanwhile theorists had been at work upon the motions of the Moon and Planets,—weighing one against the other, in fact, by finding how much their mutual attractions disturb each other. Some of these disturbances or inequalities depend upon the distance of the Earth from the Sun, that is to say, if the Earth were nearer the Sun, these inequalities would be greater and vice versa. It would be out of place to detail here the various results which have been derived from the application of these methods. I need only state that the results obtained about this time, by these methods, give from 91 3/10 to 91 8/10 millions of miles for the Sun's distance; instead of 95 4/10 obtained by Encke from the Transits of Venus.

In 1862 there occurred a very favourable Opposition of Mars. Dr. Winnecke, an astronomer of Pulkowa (the Imperial Observatory of Russia), drew up a programme of observations which was more or less perfectly carried out at six observatories in the Northern, and at three observatories in the Southern Hemisphere. Two partial discussions of some of these observations appeared, giving 91 2/10 and 91 4/10 millions of Miles respectively for the Sun's distance.

Then Mr. Stone rediscussed the Transit of Venus observations of 1769, employing only the observations in which both ingress and egress were observed. He included the suspected observations of Father Hell, and by interpreting differently the language of various observers, and applying certain corrections for different phases observed, be obtained

91 7/10 millions of miles

for the Sun's distance. This quantity was in satisfactory agreement with the results of the theoretical methods, and also with the recent results of the observations of Mars. Not only was this the case, but also all the observations were brought into the most beautiful accord. The largest error of observing the duration did not amount in any case to six seconds of time, and the probable error of one observation of duration taken by chance, was only two seconds of time.. But the duration is made up of two contacts, so that two seconds combined the probable error of two contacts. According to the theory of probabilities, the probable error of one contact would then be two seconds divided by the square root of two—in other words, the probable error of one observation of contact was only one second and four-tenths of a second of time.

This result was accepted at the time with enthusiasm, and Mr. Stone received the gold medal of the Royal Astronomical Society for his labours.

About this time, however, Newcomb, the well-known American astronomer, published his masterly discussion of the observations of Mars made at nine observatories in the year 1862. The result arrived at was

92 2/10 millions of miles

for the Sun's distance. By his own rediscussion of some of the theoretical methods, Newcomb also showed that these could be reconciled with the result he found from the observations of Afars alone. He also pointed out that, according to a discussion by Powalky, the Transit of Venus of 1769 afforded a similar result, and by combining all the various methods, he arrived at the conclusion that the true mean distance of the Sun is

92 4/10 millions of miles.

An additional confirmation of Newcomb's result had been previously derived by Foucault from his determination of the velocity of light.

The angular velocity of the Earth's motion round the Sun is accurately known; hence, if the Earth's linear velocity can be determined, the radius of motion (i.e., the Sun's distance) will also be determined. Now, the proportion which the velocity of light bears to the velocity of the Earth's motion is pretty well determined by astronomical observation; for the fact that light takes an appreciable time to travel, has the effect of shifting the apparent places of the stars. By determining the maximum amount of this shifting (called the constant of aberration), the proportion of the velocity of light to the mean velocity of the Earth's motion becomes known. Thus, if the velocity of light is determined, the velocity of the Earth's motion will become known, and hence the Sun's distance.

The English astronomer, Bradley, was the first to trace out the cause of this shifting of the apparent places of the stars.

The story is that the true explanation occurred to him as he was sailing in a boat on the Thames. The wind blew directly down the river, and, when the boat was at rest, the flag, also was blown in a direction straight down the river; but when he tacked to the right, the free end of the flag was carried to the left; and when he tacked to the left, the free end of his flag, was carried to the right.

He had been much puzzled by changes in the apparent places of certain stars, which he could not account for. Now the truth flashed upon him suddenly.

The case is similar with my boat and its flag. The true direction of the wind represents the mean direction of a star; the boat tacking from side to side represents the Earth in its yearly revolution going from one side of the Sun to the other. The deviation of the flag from its true direction depends upon the velocity of the wind and the direction and velocity of the boat's motion. If, then, the velocity of light is not infinitely great, as compared with the velocity of the Earth's motion, the apparent place of a star must be changed according as the Earth is moving in one direction or the other; in the same way that the apparent direction of the flag is changed according to the direction of the boat's motion.

So Bradley set to work to reconcile his observations on this supposition, and succeeded perfectly, making, in so doing, the first determination of the constant of aberration.

This constant he found to be 20".2, whence he concluded that the velocity of light was 10,210 times as great as the velocity of the Earth's motion in its orbit. Succeeding astronomers have made careful determinations of the constant of aberration, with various results from 20".2 to 20".6.

The value of this constant which has most generally received the confidence of astronomers is the result obtained by Struve, 20".4451.

The Sun's distance found by Encke from the Transit of Venus, combined with Struve's value of the constant of aberration, gave 191 thousand miles per second as the velocity of light.

Later, Foucault, by most ingenious optical and mechanical arrangements actually determined the velocity of light, and found it to be between 185 and 186 thousand mile's per second, a velocity which, combined with Struve's value of the constant of aberration, gave

92 3/10 millions of miles

for the Sun's distance; a quantity in satisfactory accord with Newcomb's result. It was not until 1872, however, that Newcomb's conclusion received its most remarkable confirmation. In that year Leverrier communicated a paper to the Paris Academy of Sciences, in which he gave three values of the Sun's distance, which resulted from three independent researches in the theory of the Planets.

1. From the motions of Mars 92 2/10 millions of miles.
2. "           " Venus 92 3/10 "
3. Other motions of Venus 92 3/10 "

In such estimation did Leverrier hold the accuracy of these results, that be conceived it almost impossible that any direct means of observation could furnish a better determination.

He proposed, as preferable to the Transit of Venus, the mode of determining the Sun's distance by observations for the Velocity of Light and the Constant of Aberration, and recommended the Academy to take steps in these directions.

Of the Transit of Venus, he says, "the determination of the Solar parallax by means of the Transit of Venus still retains all its interest, but conditionally on its being made with exceptional precision, so that the astronomer may be able to answer for it with an accuracy exceeding 1/100 of a second of arc." (This accuracy corresponds with 1/10 of a million of miles in the Sun's distance).

Leverrier did not believe that the Transit of Venus could yield such an accuracy; and in this respect, as well as in regard to the accuracy of his theoretical conclusions (confirmed as they were by Newcomb's results), Leverrier's opinion was shared by many astronomers.

By others, however, this latter opinion of Leverrier's was not held, but it seemed desirable to all that the Transit of Venus of 1874 should be observed in the best possible manner. The aid of photography was called in; the use of the most exquisite of all angle measuring instruments — the Heliometer — was discussed and adopted; volumes were written, and papers were read and discussed as to the best stations for observation; and all the astronomical talent of all nations busied itself in preparation.

No scientific object ever before excited such widespread activity and interest; or received from Governments and individuals such hearty and substantial assistance.

The extent of labour and of toil expended upon the Transit of 1874 may probably be gathered from a glance at the following table, prepared from the Report of the Council of the Royal Astronomical Society for the year 1876, showing the stations where the Transit was more or less successfully observed:—

Station. Nationality. Ingress.
No. of
Observations
Egress.
No. of
Observations
No. of
Photographs
Heliometers.
Number
Employed
Australia          
  * Adelaide British   2    
  * Melbourne " 1 3 200  
  * Sydney " 2 2 180  
  * Windsor " 1 1    

Burmah

         
    Bamo ? 1 1    
    Beyrout ?   1    

Cape of Good Hope

British   3 14  

Ceylon

"        
    Columbo " 1 1    

Egypt

         
  * Cairo "   3    
    Gondokoro "   1    
  * Suez "   1    
  * Thebes Russian   1    
  * Thebes German   I   1
  * Thebes British   1 35  

India

         
    Bushire " 1 1    
    Calcutta " 1 1    
  * Kurrachee "        
  * Maddapore Italian 1 1    
  * Roorkee British 2 2 100  

Islands

         
  * Auckland German I E   1
  * Bourbon Dutch   3   1
  * Chatham American     8  
  * Sandwich Island British 5   60  
  * Kerguelen   2 2    
  * Kerguelen German I E    
  * Kerguelen American     26  
  * Mauritius German   2   1
  * Mauritius Lord Lind-
say's Exp.
  3 100 1
  * Mauritius British 1 1    
  * New Caledonia French 1   100  
  * Rodriguez British 3 3    
  * St. Paul French I E 500  
Japan          
  * Kobe " I E P  
  * Nagasaki American I E 60  
  * Nagasaki French I E P  
  * Yokohama Russian I E    
  * Yokohama Mexican 2 2 100  
New Zealand          
  * Bluff Harbour German I   P  
  * Christchurch British     9  
  * Queenstown American I   59  
Persia          
  * Ispahan German     19  
  * Teheran Russian I E P  
Russia and China          
    Habarowka " I      
    Jalta "   E    
  * Kiachta " I E 8  
  * Neytschinsk " I E   1
    Orianda "   E    
  * Port Passuet Russian I E 38  
  * Pekin American I E 90  
  * Pekin French I E 60  
  * Cheefoo German I E P 1
  * Tschita Russian I E   1
  * Wladiwostok "   E    
  * Wladiwostok American I E 13  
Tasmania          
  * Campbell Town American   E 55  
  * Hobart Town "     39  

In all cases where I am not sure how many observations were obtained, I have substituted the letters I and E to represent an unknown number of observers of Ingress or Egress respectively. Also, where the number of Photographs obtained is unknown, I have substituted the letter P.

With regard to the photographs, I have simply stated the number of pictures reported as obtained. It is only when these have been tested under the microscope that it is possible to say how many of them will really be useful for measurement-probably not more than one-tenth of the number will prove to be so.

At all the stations marked with an asterisk, some at least of the observers were men with previous training, provided wit h thoroughly good instruments; indeed it is very questionable if any observations made without these conditions should be admitted into the final discussion.

Now what are the results of all these observations? It is as yet impossible to say with anything like certainty, but, from partial discussions which have been published, we may draw some conclusions.

The photographic observations have resulted in failure, at least so far as the British stations are concerned. The pictures have not the sharpness necessary for the delicate measurement to which they must be subjected, and they appear besides to be affected by systematic errors inherent to the method employed.

In criticising this result, it must not be forgotten that the method is experimental; that its first application to the measurement of small angles (requiring an accuracy of 2/100 or 3/100 of a second of arc) was on the occasion of the Transit of Venus, and that no satisfactory means existed of putting the method to previous proof.

With the experience of last Transit, however, I think it is not impossible that photography may be successfully employed to observe the Transit of 1882.

The Heliometer observations are not as yet published in a form from which any accurate estimate of their value can be formed.

The eye observations of contact present the same difficulties as in former Transits.

The contact is not a sharply-marked phenomenon, but a gradual merging of two limbs. It is complicated by the effects of the atmosphere of Venus, by the irradiation of the Sun's limb, and by the undulations of our own atmosphere. These effects also vary with the description of telescope and accessories employed; with the depth and colour of sun-shade; with the aperture of the telescope and its magnifying power; and, lastly, with the imagination of the observer, with his previous impression as to what he ought to see, or hopes to see, and with the language in which he describes what he believes he has seen.

Picture for a moment the circumstances under which the observations are taken. The observer has made a long, it may be a perilous journey, and has perhaps encountered much difficulty in landing and erecting his instruments and Observatory. The weather for days before the 9th of December has been unpropitious, and the observer is worn out by long and anxious watching. But when the eventful morning arrives, the sky is cloudless, and the Sun shines in all his strength. The astronomer is exultant, and the revulsion of feeling sets his every nerve a-tingling.

Soon after the predicted time, he sees a small black indentation in the Sun's limb. The indentation increases. The black disc of the Planet has half entered upon the Sun-when, what is this? Around a portion of the black disc appears a band of light, that extends till it forms a ring round the Planet. Three minutes more and the critical instant will be past; and yet the mystery of the unexpected ring remains unsolved. Meanwhile the definition becomes less and less satisfactory, for the strong Sun has heated up the rocks which surround the Observatory, and tremulous currents of air ascend which render the image blurred and ill-defined. About this time, according to the observer's previous experience with "the model," fine sharp cusps should be rapidly approaching each other, and, when they meet, light should appear between the edge of the Planet and the Sun. But no! the cusps are not fine; they are blunted and rounded off; and this bright ring of light complicates the matter.

A few seconds more, and all will be over. Even yet the observer does not know what he ought to note. He feels that now is the supreme moment, that now he must reap the fruit of all his labour, or lose his harvest.

Still there is nothing precise to observe. So, in the tremulous image before him, he notes what best he can,—and then the Transit is over.

After the first feeling of disappointment, comes a certain feeling of satisfaction. The important contact has been observed. The circumstances were favourable. All has been done that could have been done. Thus the report unconsciously partakes of a more buoyant character than would have been the case could the observer precisely recall his feelings at the instant of observation, or than if he knew that to-morrow, and the next day, and the next, he could repeat his observation again and again.

Is it then a wonder that any person who attempts to select "corresponding phases" of the Transit, from the uncertain and incongruous records of the various observers, should vacillate in his opinion as to the true interpretation of their language? Is it a wonder that anyone who has previously formed a strong opinion as to what the result should be, and who has made himself familiar with the effects of different interpretations on the final figures, should unconsciously give such interpretation as will lead to a result agreeing with his preconceived ideas?

I do not think that it is; the wonder would rather be if it were otherwise.

Accordingly we find that various results have already appeared. The first of these was obtained by M. Puiseux from the French observations at Pekin and St. Paul: it gave

92 3/10 millions of miles

for the Sun's distance, and so far confirmed Newcomb's and Leverrier's conclusions. The next combination, however, gave a very different result—more nearly Encke's distance. So the French, like the Germans, very wisely resolved to publish only the observations made, without drawing any conclusions from them, leaving the definitive result to be deduced from the combined observations of astronomers of all nations, according to the recommendation of our Astronomer Royal.

This was indeed a wise recommendation, for partial discussions can tend only to hinder true progress.

The British public, however, as is their manner, upset this wise resolution, and Parliament demanded immediate value for its money. "What is the result?" was the impatient cry of honourable members, doubtless waiting with restless anxiety for information as to the Solar Parallax. And so the Astronomer Royal had to prepare a Blue Book, and give the result of the British expeditions alone. The result contained in this report was

93 4/10 millions of miles

for the Sun's distance.

Then Mr. Stone (H.M. Astronomer at the Cape) rediscussed the same observations, and deduced

91 8/10 millions of miles

or a result differing by more than one and a half millions of miles from the Astronomer Royal's result.

Subsequently, introducing a wider range of observations, the Astronomer Royal announced in his Report to the Board of Visitors, that the observations seemed now to point to about

92 7/10 millions of miles.

Finally Captain Tupman, the chief of the British Expeditions, from a still more extended discussion deduced

92 5/10 millions of miles,

a result agreeing precisely with that of Newcomb and Leverrier.

With admirable candour and fairness, however, he states

"Although the above results" (referring to the results from Ingress and Egress taken separately), "present such an unexpected agreement, it cannot be said that the mean" (equivalent to a Sun's distance of

92 4/10 millions of miles)

"is entitled to much confidence."

* * * * * *

"Any selection of times made after the investigation of the effects of parallax, will always expose the result to the suspicion of having been 'doctored."'

In these opinions I heartily concur.

Of course, no determination of the Sun's distance is likely to be correct which is irreconcileable with a reasonable interpretation of a well-observed Transit of Venus—like that of 1874; and therefore, as a confirmation of other results, the eye observations of contact may prove useful. But of this I am convinced, that no eye observations of a Transit of Venus can ever satisfactorily determine the Sun's distance; though it is not impossible that photography, properly employed, may accomplish the desired end.

One very useful lesson taught by this discussion of the Transit of 1874 is, that the apparent agreement of the observations, in Mr. Stone's discussion of the 1769 Transit, is entirely illusory. In the Transit of 1874, the probable error of an observation of contact amounted, at least, to 7 or 8 seconds of time. These observations were made with the best modern appliances, very, very far superior to any of those used in 1769. From what we know of the instruments, and the nature of the observation, the real probable error of contact in 1769 must have been at least 10 seconds (instead of one and a half), and the resulting distance of the Sun may be anything from 90 1/2 to 95 millions of miles.

But between the observation and discussion of the Transit of Venus of 1874, results of other interesting investigations were published.

In 1872 Dr. Galle of Breslau had proposed a series of observations on the minor Planet Flora, at its Opposition in the autumn of 1873, for the purpose of determining the Sun's distance. He contended, that, though this Planet would not approach so near to the Earth as Mars and Venus do in certain circumstances, yet its minute disc, exactly like a Star, would form a better object for exact measurement, and one less liable to systematic error of bisection. He secured the co-operation of Observatories in the Northern and Southern Hemispheres, and, by combining the observations so obtained, he derived

92 1/2 millions of miles

for the Sun's distance.

This result was also confirmatory of the Newcomb-Leverrier value, and the method offered great probability of freedom from systematic error. But at some of the most important Observatories the instrumental equipment was not satisfactory for so delicate an inquiry; and the result therefore hardly possesses the importance which it would otherwise have.

Meanwhile the French Academy had adopted the suggestions of Leverrier, and M. Cornu was selected to redetermine the Velocity of Light. The investigation was executed with eminent skill and care, and the result, combined with Bradley's determination of the constant of aberration, gave 92 2/10 millions of miles for the Sun's distance, a result also in agreement with that of Newcomb and Leverrier; but when combined with the far more refined and more generally accepted determination of Struve, the result is

93 millions of miles

Such was the state of our knowledge of the Sun's distance in the beginning of 1877.

Opinions were divided. Few, if any, still adhered to the old value of Encke (95 4/10 millions of miles), but some firmly maintained the accuracy of the other extreme (91 6/10 millions of miles).

The vast majority of Astronomers adopted the Newcomb-Leverrier value (92 2/10 millions of miles), but very few believed the constant to be yet definitively established. It was still possible that these two coincident results might each be subject to a small systematic error; and some investigation, to which systematic error could not possibly be attributed, was earnestly desired.

In 1857 the Astronomer Royal had proposed the mode of observation described in Chap, I. p. 9, and gave it as his opinion that it was the best of all methods to determine the Sun's distance.

This proposal had never been satisfactorily carried put, and yet it offered many advantages. It required no co-operation, and the whole of the observations might be made by the same observer with the same instrument, so that systematic errors would be entirely avoided.

Combining the suggestions of the Astronomer Royal and of Dr. Galle, Lord Lindsay and I proposed in 1874 the observation of a Minor Planet in the evening and early morning, as the best method of determining the parallax; and we showed that, by employing the Heliometer in the observations, there was a probability of realizing a higher accuracy than had ever before been attained. The practical form which the proposal took, was to observe the Planet Juno on the occasion of Lord Lindsay's Expedition to Mauritius; and this was duly done. On account of the late arrival of Lord Lindsay's yacht at Mauritius, only a very small proportion of the intended observations were secured; but these, on reduction, proved to demonstration the extreme accuracy of the method. Though the observations are not numerous enough to give a determination of the Sun's distance of the highest precision, it is still interesting to find that the tendency of the result was to confirm the Cornu-Struve result derived from the Velocity of Light and the Constant of Aberration.

From all the observations of Juno combined, the result was

93 3/10 millions of miles,

or rejecting one outstanding result

92 8/10 millions of miles.

The Opposition of Mars in 1877 offered, so far as geometrical conditions are concerned, the most favourable opportunity of the century to determine the parallax, by observations at a single station.

I thought it would be a matter of the greatest regret if such an opportunity were lost. Having mentioned the matter to Lord Lindsay, he, in the kindest manner and in fullest sympathy with the importance of the object, at once placed his Heliometer at my disposal.

I had already much experience in the use of this instrument, and had spent many months in finding out its errors and the best means of correcting them. This great labour therefore would not have to be repeated.

The combination of circumstances was altogether so fortunate, that the Royal Astronomical Society, on the earnest recommendation of the Astronomer Royal, guaranteed the Ł500 which I considered necessary for the expenses of the expedition. Through the good offices of the Astronomer Royal, I also received such letters from the Lords of the Admiralty, and such efficient assistance was in consequence given to me at Ascension, that the money voted by the Society proved more than sufficient. Numerous evening and morning observations of Mars were secured; and the reductions, now far advanced, promise a result of very great accuracy.

I may state here that the observations of one week (Sept. 4-11), which are now reduced, confirm the tendency of the Juno result and of the Cornu-Struve value of the Sun's distance; but a good many months must still elapse before the final result, from all the observations, can be deduced.

Another Astronomer, Mr. Maxwell Hall, of Jamaica, observed Mars exactly on the plan of the Astronomer Royal. The details of his observations are not yet published, but the result be arrived at also confirms the tendency of the Cornu-Struve, Juno, and Mars results.

Without entering into greater detail, I may state that, if these recent results are confirmed, the Sun's distance will prove to be nearer to 93 than to 92 millions of miles.

It may appear strange to the uninitiated that Astronomers should be in doubt about so large a quantity as a million of miles, but perhaps a familiar illustration will convey some idea of the difficulty of the problem.

The apparent size of the Earth, looked at from the Sun is about that of a globe 5 1/2 inches in diameter viewed at a mile distant.

If this 5 1/2 inch globe is shifted 57 feet nearer to the observer, it will be increased in apparent diameter just as much as the Earth would be if shifted a million of miles nearer to the Sun—or, as if the 5 1/2 inch globe had not been shifted, but had been increased by 6/100 of an inch in diameter. Hence, in measuring the parallax, an angular error, corresponding with 6/100 of an inch viewed at a mile distance, will produce the error in question.

If any one desires to form an adequate idea of the difficulties of measuring the Sun's distance to a million of miles, let him try to measure the thickness of a florin-piece, looked at, edge on, a mile off.

I have endeavoured in the preceding sketch to outline the History of the great Problem which occupied our time and thoughts at Ascension. I shall be well satisfied if I have enabled any one to realize somewhat of its nobility and interest, and the consequent intensity of our anxiety for a successful result.

DAVID GILL.

LONDON: November, 1878.


† I quote Dr. Gould's account in his admirable introduction to the U. S. Naval Astronomical Expedition to Chili.

‡ I do not think I have exaggerated the difficulty and uncertainty of the observation in the slightest degree, and I quote in corroboration the words of a most conscientious observer:—
    "Sky perfectly clear; no cloud.
    "On focussing, after changing the micrometer, to my astonishment I saw a completion of light round the planet, perfectly distinct, and such as I should have said, from previous model-practice, was immediately after contact. This is the time recorded. I remained looking at it for about two minutes, but could see no instantaneous phenomenon of contact, no black drop, nor anything resembling the model. I noticed that this light did not appear to thicken as I should have expected for a considerable time after that recorded, but as I considered, from my previous experience [with the model], that the contact had occurred, and was, unable to get, accurately, any further change until the planet was visibly on the Sun, I cannot say that the time as noted is at all satisfactory."
    Had all observers been equally hard to please as to the precision of the phenomenon, the result of the Transit would have been—no observations. But these words will undoubtedly recall to many an observer the unsatisfactory character of the phenomenon which he had to note.


Chapter I