Tricolor. The Legitimist Pretender would have no communication with the Orleanist princes except on the basis of entire submission by them to his views. The monarchical cause was thenceforth hopeless; yet the party in the Chamber determined to continue the fight and, at all events, take vengeance for what it chose to consider its betrayal by Thiers. History will not so stigmatise his policy. The socalled 'Pacte de Bordeaux' was simply a promise made by Thiers to the Royalist party in the early days of the Assembly, that the Government would be neutral between the various sections of political opinion; there was also a tacit assumption that the majority would not impede the work of the Government; and it also involved a pledge to the Republicans that the Republic ran no danger under the rule of the President. When the Royalists clearly showed their determination to wreck the existing régime, Thiers was perfectly justified in issuing the famous message of November 13, 1872, in which he said:

'The Republic exists; to desire anything else would be a new revolution and the most formidable of all. The Republic must be the Government of this nation, which having for a long time and in good faith left the direction of its destinies to an hereditary power (but having failed through faults impossible to judge to-day), at length undertakes the responsibility of herself, and herself alone, directing her own course by her elected representatives.'

This was an epoch-making announcement, and came as an unpleasant surprise to the majority who were still engaged in the hopeless task of adapting Legitimist principles to the will of the nation. It was the duty of a statesman to dissipate their illusions and save the country from a catastrophe; it would have been folly for him to have ignored the reiterated lessons of Republican victories at the elections for the Chamber, the Conseils généraux and the Municipalities. It is true that he still clung to the theory of a Republic without Republicans,' and thereby failed to win the entire support of Gambetta and his following. In this he was no doubt influenced as much by the wish to continue indispensable himself as to play his part as a moderating force.

It would be indeed ridiculous to pretend that Thiers was a Washington. Vanity swayed his actions to as large an extent during these years as it had done throughout his life; but, making all deductions for the influence of egotism on his decisions, how great was the work he accomplished

for France, and how impossible it would have been for any other man to have undertaken it! The actual bulk of the legislation completed under his inspiration was prodigious. He was, it must be remembered, not merely head of the State, but chief Minister. He carried a system of army reorganisation which before many years gave France a military force such as she had never possessed except for a brief period under the great Napoleon. This was accomplished while the Germans were actually encamped on French soil. He reformed the Conseil d'Etat and the jury system, and gave France a wholly new scheme of secondary education. He also inaugurated a protective tariff, and organised the vast financial operations necessitated by the payment of the indemnity. With all its faults, too, the Assembly was a hardworking legislature, and might read many a lesson to its more radical successors. But Thiers was the propelling and vivifying force. He found France rolling in the trough of the sea in two years he gave her a system of government and set her head straight. It was not his fault if some of his successors have steered an erratic course.

But some of the praise lavished on Thiers may justly be given to his country. Without the help of France herself his vigour and sagacity would have done little. The historian must render homage to the State which, on the morrow of such disasters, could give a steady and increasing support to a régime of common sense. France then (in the words of George Sand) learned for the first time in her history to discuss without making a revolution.' That lesson she has not altogether lost during the thirty years that have followed. This is surely no small asset to the credit of the Third Republic, and may counterbalance some strange lapses from the perfect law of liberty.

ART. V.—1. Problems in Astrophysics. By AGNES M. CLERKE. London: A. & C. Black, 1903.

2. Annals of the Astronomical Observatory of Harvard College. Vol. 37: Observations of Variable Stars of Long Period.' O. C. WENDELL. 1889-1901. Vol. 38: Variable Stars in the Cluster @ Centauri.' S. I. BAILEY. 1902. University, Cambridge, Mass., 1902.

3. Astrophysical Journal. Vols. 15, 16. Edited by G. E. HALE and others. University of Chicago Press, 1902.

THE

HE science of astrophysics, which may be defined as that branch of astronomy dealing with the physical state, chemical composition, and life-history of the heavenly bodies, may be said to date its birth from the year 1859, when Kirchhoff discovered the cause of the dark, or Fraunhofer, lines in the solar spectrum. By showing that from their examination could be inferred the presence of terrestrial elements in the sun, he opened up the possibility of extending the arm of chemical analysis to the uttermost limits of our universe, and thus compelling both sun and stars to yield up the secrets of their composition, however remote the depths of space in which they are situated, and however feeble the rays of their light which reach the observer. Astrophysics is, therefore, in a large if not an exclusive degree the science of knowledge as revealed by a single instrument, the spectroscope. It is often interesting to recall the words in which the announcement of a farreaching discovery has first been made; not, perhaps, the announcement which takes the form of a carefully written and well-weighed scientific memoir, but rather such as is to be found in an informal letter written to a friend who is also a worker at the subject, and to whom the happy discoverer of some new thing naturally turns for sympathy and interest. In the case of spectrum analysis we are fortunately in a position to do this.

On November 15, 1859, Bunsen, who was the associate of Kirchhoff in this investigation, wrote to a friend and fellowchemist in England :

'At this moment I am engaged with Kirchhoff in a research which gives us sleepless nights. Kirchhoff has made a most beautiful and most unexpected discovery; he has found out the cause of the dark lines in the solar spectrum, and he has been able both to strengthen these lines artificially in the solar spectrum and to cause their appearance in a continuous spectrum of a flame, their positions being iden

tical with those of the Fraunhofer lines. Thus the way is pointed out by which we can ascertain the composition of the sun and fixed stars with the same degree of certainty as we can ascertain with our reagents the presence of chlorine. By this method, too, the composition of terrestrial matter can be ascertained and the component parts distinguished with as great ease and delicacy as is the case with the matter contained in the sun.'

The seed planted nearly half a century ago has now grown into a vigorous tree bearing fruits that its sowers would hardly recognise.

The chemical composition of sun and stars is one only, and may almost be termed one of the less important, of the revelations of the spectroscope. As is now well known, the spectral lines, whether they be bright lines caused by the emissions from a glowing gas, or dark ones caused by the absorption of the same gas when white light is passed through it, are liable to changes both of position and of character, due not only to variations in the temperature and pressure of the gas, its state of electric or magnetic strain, its velocity of movement to or from the observer, but also to the presence of other bodies, even though they be not chemically active and exist only in the minutest quantities.

The difficulty of completely interpreting the spectral record has thus in a great measure increased with increase of knowledge. In the early days it was rashly assumed that, when the spectroscope had answered the first question put to it and had shown what elements were present in the body under examination, it had fulfilled its task, and it was therefore with feelings somewhat akin to disappointment that the physicist was forced to recognise that the problem was by no means so simple as it had appeared on first sight. Thus we find Professor Scheiner writing in 1890:- Unfortunately it can be asserted that we now know less as to the 'meaning of the Fraunhofer lines than was supposed to be 'known ten years ago."

What, however, tended to appear in the guise of a series of annoying complications, adding difficulties to the task of the spectroscopist and carrying no compensating advantages in its train, is now recognised as containing the germs of a completer knowledge than that of which the early experimenters ever dreamed. We see that because a spectrum line changes with change of pressure, magnetic strain, velocity of light-source, and other disturbing causes, the value of its record is thereby increased, inasmuch as it becomes the chronicle not only of the chemical constitution,

VOL. CXCVIII. NO. Cocov.

I

but also of the rate of movement and physical conditions of the celestial body. This record it is not easy to read. The difficulties of its interpretation lie not so much in the number of the disturbing factors as in the impossibility of imitating stellar or solar conditions in our terrestrial laboratories, and in the puzzling inconsistencies-such at least they appear to our imperfect knowledge-exhibited by the spectral lines. In many ways the very delicacy of the spectral reaction operates as a positive hindrance. Were it not that our familiarity with the phenomenon has bred something like contempt, we might well regard it as one of the most remarkable facts in nature that an infinitesimal quantity of sodium vapour in an electric arc exerts as powerful an absorption of light of definite note as the thickness of many hundred miles of similar vapour forming part of the solar envelope. It is no question here of absolute opacity to light upon which an increase of thickness would be without effect. The dark lines in a spectrum are only dark relatively to the bright background, and when viewed separate from their surroundings appear bright. In other words, the absorption is only relative. How then are we to explain why it takes no account of the quantity of the absorbing material? The fact that it does so is undoubted. If proof be required, we have only to point our spectroscope to the sun, and observe that the spectrum is identically the same whether viewed at the centre of the disc or at the edge, positions which correspond respectively to the minimum and maximum thickness of the absorbing layer. It is hence not difficult to see why all attempts at quantitative analysis with the spectroscope have proved abortive.

Another fruitful source of perplexity and uncertainty is to be found in the diverse behaviour of spectral lines, from the same substance, lying in different parts of the scale. A familiar example of this is the metal lithium, which is quite certainly a solar constituent, and yet possesses a strong and characteristic line in its spectrum which is not present as a dark, Fraunhofer line. Even in the case of sodium, whose pair of lines in the yellow was the subject of Kirchhoff's original experiment, and which was therefore the first recognised solar element, it is almost certain that the doublet of lines in the deep red is not visible in the sun. In fact, most elements present similar anomalies; even hydrogen itself fails to show its rhythmic series of lines in the upper or ultra violet end of the light scale, so that this partial deficiency of absorbing power is not confined to the low or to