But one reader at least cannot help being haunted as he reads by the notion how much better Merimee would have told it. _Le Fils du t.i.tien_--the story of the great master's lazy son, on whom even love and entire self-sacrifice--lifelong too--on the part of a great lady, cannot prevail to do more in his father's craft than one exquisite picture of herself, inscribed with a sonnet renouncing the pencil thenceforth--is the best told story in the book. But Gautier would certainly have done it even better. _Margot_, in the same fatal way and, I fear, in the same degree, suggests the country tales of Musset's own faithless love.

[Sidenote: _Emmeline._]

But the most crucial example of the "something wrong" which pursues Musset in pure prose narrative is _Emmeline_. It is quite free from those unlucky, and possibly unfair, comparisons with contemporaries which have been affixed to its companions. A maniac of parallels might indeed call it something of a modernised _Princesse de Cleves_; but this would be quite idle. The resemblance is simply in situation; that is to say, in the _publica materies_ which every artist has a right to make his own by private treatment. Emmeline Duval is a girl of great wealth and rather eccentric character, who chooses to marry (he has saved her life, or at any rate saved her from possible death and certain damage) a person of rank but no means, M. de Marsan. There is real love between the two, and it continues on his side altogether unimpaired, on hers untroubled, for years. A conventional lady-killer tries her virtue, but is sent about his business. But then there turns up one Gilbert, to whom she yields--exactly how far is not clearly indicated. M. de Marsan finds it out and takes an unusual line. He will not make any scandal, and will not even call the lover out. He will simply separate and leave her whole fortune to his wife. She throws her marriage contract into the fire (one does not presume to enquire how far this would be effective), dismisses Gilbert through the medium of her sister, and--we don't know what happened afterwards.

Now the absence of _finale_ may bribe critics of the present day; for my part, as I have ventured to say more than once before, it seems that if you accept this principle you had much better carry it through, have no middle or beginning, and even no t.i.tle, but issue, in as many copies as you please, a nice quire or ream of blank paper with your name on it.

The purchasers could cut the name out, and use it for original composition in a hundred forms, from washing bills to tragedies.

But I take what Musset has given me, and, having an intense admiration for the author of _A Saint Blaise_ and _L'Andalouse_ and the _Chanson de Fortunio_, a lively grat.i.tude to the author of _Il ne faut jurer de rien_ and _Il faut qu'une porte soit ouverte ou fermee_, call _Emmeline_ a very badly told and uninteresting story. The almost over-elaborate description of the heroine at the beginning does not fit in with her subsequent conduct; Gilbert is a nonent.i.ty; the husband, though n.o.ble in conduct, is pale in character, and the sister had much better have been left out.[236] So the rest may be silence.

[Sidenote: Gerard de Nerval--his peculiar position.]

I have been accused (quite good-naturedly) of putting Rabelais in this history because I liked him, though he was not a novelist. My conscience is easy there; and I think I have refuted the peculiar charge beforehand. But I might have a little more difficulty (though I should still lose neither heart nor hope) in the case of the ill-fated but well-beloved writer whom G.o.ds and men call Gerard de Nerval, or simply Gerard, though librarians and bibliographers sometimes insist on his legal surname, Labrunie. It certainly would be difficult, from the same point of view of strict legality, to call anything of his exactly a novel. He was a poet, a dramatist, a voyage-and-travel writer, a bibliographer (strange trade, which a.s.sociates the driest with the most "necta_w_eous" of men!) even sometimes a tale-teller by name, but even then hardly a novelist. Yet he managed to throw over the most unlikely material a novelish or at least a romantic character, which is sometimes--nay, very often--utterly wanting in professed and admitted masters of the business; and he combines with this faculty--or rather he exalts and transports it into--a strange and exquisite charm, which n.o.body else in French, except Nodier[237] (who very possibly taught Gerard something), possesses, and which, though it is rather commoner in English and in the best and now almost prehistoric German, is rare anywhere, and, in Gerard's peculiar brand of it, almost entirely unknown.

For this "Anodos"--the most unquestionably ent.i.tled to that t.i.tle of all men in letters; this wayless wanderer on the earth and above the earth; this inhabitant of mad-houses; this victim, finally, either of his own despair and sorrow or of some devilry on the part of others,[238]

unites, in the strange spell which he casts over all fit readers, what, but for him, one might have called the idiosyncrasies in strangeness of authors quite different from each other and--except at the special points of contact--from him. He is like Borrow or De Quincey (though he goes even beyond both) in the singular knack of endowing or investing known places and commonplace actions with a weird second essence and second intention. He is like Charles Lamb in his power of dropping from quaintness and almost burlesque into the most touching sentiment and emotion. Mr. Lang, in his Introduction to Poe, has noticed how Gerard resembles America's one "poet of the first order" in fashioning lines "on the further side of the border between verse and music"--a remark which applies to his prose as well.[239] He has himself admitted a kind of _sorites_ of indebtedness to Diderot, Sterne, Swift, Rabelais, Folengo, Lucian, and Petronius. But this is merely on the comic and purely intellectual side of him, while it is further confined, or nearly so, to the trick of deliberate "promiscuousness." On the emotional-romantic if not even tragic score he may write off all imputed indebtedness--save once more in some degree, to Nodier. And the consequence is that those who delight in him derive their delight from sources of the most extraordinarily various character, probably never represented by an exactly similar group in the case of any two individual lovers, but quite inexhaustible. To represent him to those who do not know him is not easy; to represent him to those who do is sure, for this very reason, to arouse mild or not mild complaints of inadequacy. And it must be clear, from what has been already said, that some critic may very likely exclaim, in reference to any selected piece, "Why, this is neither a novel nor a romance, nor even in any legitimate sense a tale!" The inestimable rejoinder already quoted,[240]--episcopal, and dignifying even that order though it was made only by a bishop _in partibus_--is the only one here.

[Sidenote: _La Boheme Galante_, _Les Filles du Feu_, and _Le Reve et la Vie_.]

The difficulty of discussing or ill.u.s.trating, in short s.p.a.ce and due proportion, the novel or _roman_ element in such a writer must be sufficiently obvious. His longer travels in Germany and the East are steeped in this element; and the shorter compositions which bear names of novel-character are often "little travels" in his native province, the Isle of France, and that larger _banlieue_ of Paris, towards Picardy and Flanders, which our Seventy Thousand saved, by dying, the other day.

But it is impossible--and might even, if possible, be superfluous--to touch the first group. Of the second there are three subdivisions, which, however, are represented with not inconsiderable variation in different issues.[241] Their t.i.tles are _La Boheme Galante_, _Les Filles du Feu_, and _Le Reve et la Vie_, the last of which contains only one section, _Aurelia_, never, if I do not mistake, revised by Gerard himself, and only published after his most tragic death. Its _supra_-t.i.tle really describes the most characteristic part or feature of all the three and of Gerard's whole work.

[Sidenote: Their general character.]

To one who always lived, as Paul de Saint-Victor put it in one of the best of those curious exercises of his mastery over words, "in the fringes[242] of the actual world," this confusion of place and no place, this inextricable blending of fact and dream, imagination and reality, was natural enough; and no one but a Philistine will find fault with the sometimes apparently mechanical and Sternian transitions which form part of its expression. There was, indeed, an inevitable _mixedness_ in that strange nature of his; and he will pa.s.s from almost "true d.i.c.kens" (he actually admits inspiration from him) in accounts of the Paris _Halles_, or of country towns, to De Quinceyish pa.s.sages, free from that slight touch of _apparatus_ which is undeniable now and then in the Opium Eater. Here are longish excursions of pure family history; there, patches of criticism in art or drama; once at least an elaborate and--for the time--very well informed as well as enthusiastic sketch of French seventeenth-century poetry. It may annoy the captious to find another kind of confusion, for which one is not sure that Gerard himself was responsible, though it is consistent enough with his peculiarities.

Pa.s.sages are redistributed among different books and pieces in a rather bewildering manner; and you occasionally rub your eyes at coming across--in a very different context, or simply shorn of its old one--something that you have met before. To others this, if not exactly an added charm, will at any rate be admitted to "grace of congruity." It would be less like Gerard if it were otherwise.

[Sidenote: Particular examples.]

In fact it is in these mixed pieces that Gerard's great attraction lies.

His regular stories, professedly of a Hoffmannesque kind, such as _La Main Enchantee_ and _Le Monstre Vert_, are good, but not extraordinarily good, and cla.s.sable with many other things of many other people. I, at least, know nothing quite like _Aurelia_ and _Sylvie_, though the dream-pieces of Landor and De Quincey have a certain likeness, and Nodier's _La Fee aux Miettes_ a closer one.

[Sidenote: _Aurelia._]

_Aurelia_ (which, whether complete in itself or not, was pretty clearly intended to be followed by other things under the general t.i.tle of _Le Reve et la Vie_) has, as might be expected, more dream than life in it.

Or rather it is like one of those actual dreams which themselves mix up life--a dream in the composition. Aurelia is the book-name of a lady, loved (actually, it seems) and in some degree responsible for her lover's aberrations of mind. He thinks he loves another, but finds he does not. The two objects of his pa.s.sion meet, and the second generously brings about a sort of reconciliation with the first. But he has to go to Paris on business, and there he becomes a mere John-a-Dreams, if not, in a mild way, a mere Tom of Bedlam. The chief drops into reality, indeed, are mentions of his actual visits to _maisons de sante_. But the thing is impossible to abstract or a.n.a.lyse, too long to translate as a whole, and too much woven in one piece to cut up. It must be read as it stands, and any person of tolerable intelligence will know in a page or two whether Gerard is the man for him or not. But when he was writing it he was already over even the fringe of ordinary sane life, and near the close of life itself. In _Sylvie_ he had not drifted so far; and it is perhaps his best diploma-piece.[243]

[Sidenote: And especially _Sylvie_.]

For _Sylvie_, with its sub-t.i.tle, "Souvenirs du Valois," surely exhibits Gerard, outside the pure travel-books, at his very best, as far as concerns that mixture of _reve_ and _realite_--the far-off goal of Gautier's[244] _Chimere_--which has been spoken of. The author comes out of a theatre where he has only seen Her, having never, though a constant worshipper, troubled himself to ask, much less to seek out, what She might be off the stage. And here we may give an actual piece of him.

We were living then in a strange kind of time,[245] one of those which are wont to come after revolutions, or the decadences of great reigns. There was no longer any gallantry of the heroic kind, as in the time of the Fronde; no vice, elegant and in full dress, as in that of the Regency; no "Directory" scepticism and foolish orgies. It was a mixture of activity, hesitation, and idleness--of brilliant utopias; of religious or philosophical aspiration; of vague enthusiasms mingled with certain instincts of a sort of Renaissance. Men were weary of past discords; of uncertain hopes, much as in the time of Petronius or Peregrinus. The materialist part of us hungered for the bouquet of roses which in the hands of Isis was to regenerate it--the G.o.ddess, eternally young and pure, appeared to us at night and made us ashamed of the hours we had lost in the day. We were not at the age of ambition, and the greedy hunt for place and honours kept us out of possible spheres of work. Only the poet's Ivory Tower remained for us, and we climbed it ever higher and higher to be clear of the mob. At the heights whither our masters guided us we breathed at last the pure air of solitude; we drank in the golden cup of legend; we were intoxicated with poetry and with love. But, alas! it was only love of vague forms; of tints roseal and azure; of metaphysical phantoms.

The real woman, seen close, revolted our ingenuousness: we would have had her a queen or a G.o.ddess, and to draw near her was fatal.

But he went from the play to his club, and there somebody asked him for what person (in such cases one regrets _laquelle_) he went so constantly to the same house; and, on the actress being named, kindly pointed out to him a third member of this club as the lady's lover-in-t.i.tle. The peculiar etiquette of the inst.i.tution demanded, it seems, that the fortunate gallant should escort the beloved home, but then go to the _cercle_ and play (they were wise enough to play whist then) for great part of the night before exercising the remainder of his rights and privileges. In the interval, apparently, other cats might be grey. And, as it happened, Gerard saw in a paper that some shares of his, long rubbish, had become of value. He would be better off; he might aspire to a portion of the lady's spare hours. But this notion, it is not surprising to hear, did not appeal to our Gerard. He sees in the same paper that a _fete_ is going to take place in his old country of the Valois; and when at last he goes home two "faces in the fire" rise for him, those of the little peasant girl Sylvie and of the chatelaine Adrienne--beautiful, triumphant, but destined to be a nun. Unable to sleep, he gets up at one in the morning, and manages to find himself at Loisy, the scene of the _fete_, in time.

One would fain go on, but duty forbids a larger allotment of s.p.a.ce; and, after all, the thing itself may be read by any one in half an hour or so, and will not, at least ought not, to be forgotten for half a lifetime--or a whole one. The finding of Sylvie, no longer a _little_ girl, but still a girl, still not married, though, as turns out, about to be so, is chequered with all sorts of things--sketches of landscape; touches of literature; black-and-white renderings of the _Voyage a Cythere_; verses to Adrienne; to the actress Aurelie (to become later the dream-Aurelia); and, lastly--in the earlier forms of the piece at any rate--s.n.a.t.c.hes of folk-song, including that really n.o.ble ballad:

Quand Jean Renaud de la guerre revint,

which falls very little, if at all, short of the greatest specimens of English, German, Danish, or Spanish.

And over and through it all, and in other pieces as well, there is the faint, quaint, music--prose, when not verse--which reminds one[246]

somehow of Browning's famous Toccata-piece. Only the "dear dead women"

are dear dead fairies; and the whole might be sung at that "Fairy's Funeral" which Christopher North imagined so well, though he did not carry it out quite impeccably.

[Sidenote: Alfred de Vigny: _Cinq-Mars_.]

The felicity of being enabled to know the causes of things, a recognised and respectable form of happiness, is also one which I have recently enjoyed in respect of Alfred de Vigny's _Cinq-Mars_. For Vigny as a poet my admiration has always been profound. He appears to me to have completed, with Agrippa d'Aubigne, Corneille, and Victor Hugo, the _quatuor_ of French poets who have the secret of magnificence;[247] and, scanty as the amount of his poetical work is, _eloa_, _Dolorida_, _Le Cor_, and the finest pa.s.sages in _Les Destinees_ have a definite variety of excellence and essence which it would not be easy to surpa.s.s in kind, though it might be in number, with the very greatest masters of poetry.

But I have never been able, frankly and fully, to enjoy his novels, especially _Cinq-Mars_. In my last reading of the chief of them I came upon an edition which contains what I had never seen before--the somewhat triumphant and strongly defiant tract, _Reflexions sur la Verite dans l'Art_, which the author prefixed to his book after its success. This tractate is indeed not quite consistent with itself, for it ends in confession that truth in art is truth in observation of human nature, not mere authenticity of fact, and that such authenticity is of merely secondary importance at best. But in the opening he had taken lines--or at any rate had said things--which, if not absolutely inconsistent with, certainly do not lead to, this sound conclusion. In writing historical novels (he tells us) he thought it better not to imitate the foreigners (it is clear that this is a polite way of indicating Scott), who in their pictures put the historical dominators of them in the background; he has himself made such persons princ.i.p.al actors. And though he admits that "a treatise on the decline and fall of feudalism in France; on the internal conditions and external relations of that country; on the question of military alliances with foreigners; on justice as administered by parliaments, and by secret commissions on charges of sorcery," might not have been read while the novel _was_; the sentence suggests, with hardly a possibility of reb.u.t.tal, that a treatise of this kind was pretty constantly in his own mind while he was writing the novel itself. And the earlier sentence about putting the more important historical characters in the foreground remains "firm,"

without any necessity for argument or suggestion.

[Sidenote: The faults in its general scheme.]

Now I have more than once in this very book, and often elsewhere, contended, rightly or wrongly, that this "practice of the foreigners,"

in _not_ making dominant historical characters their own dominant personages, is _the_ secret of success in historical novel-writing, and the very feather (and something more) in the cap of Scott himself which shows his chieftainship. And, again rightly or wrongly, I have also contended that the hand of purpose deadens and mummifies story. Vigny's own remarks, despite subsequent--if not recantation--qualification of them, show that the lie of his land, the tendency of his exertion, _was_ in these two, as I think, wrong directions. And I own that this explained to me what I had chiefly before noticed as merely a fact, without enquiring into it, that _Cinq-Mars_, admirably written as it is; possessing as it does, with a hero who might have been made interesting, a great person like Richelieu to make due and not undue use of; plenty of thrilling incident at hand, and some actually brought in; love interest _ad libitum_ and fighting hardly less so; a tragic finish from history, and opportunity for plenty of lighter contrast from Tallemant and the Memoirs--that, I say, _Cinq-Mars_, with all this and the greatness of its author in other work, has always been to me not a live book, and hardly one which I can even praise as statuesque.[248]

It is no doubt a misfortune for the book with its later readers--the earlier for nearly twenty years were free from this--that it comes into closest comparison with Dumas' best work. Its action, indeed, takes place in the very "Vingt Ans" during which we know (except from slight retrospect) nothing of what D'Artagnan and the Three were doing. But more than one or two of the same historical characters figure, and in the chapters dealing with the obscure _emeute_ which preceded the actual conspiracy, as well as in the scenes touching Anne of Austria's private apartments, the parallel is very close indeed.

[Sidenote: And in its details.]

Now of course Dumas could not write like Vigny; and though, as is pointed out elsewhere, to regard him as a vulgar fellow is the grossest of blunders as well as a great injustice, Vigny, in thought and taste and _dianoia_ generally, was as far above him as in style.[249] But that is not the question. I have said[250] that I do not quite _know_ D'Artagnan, though I think I know Athos, as a man; but as a novel-hero the Gascon seems to me to "fill all numbers." Cinq-Mars may be a succession or chain of type-personages--generous but headlong youth, spoilt favourite, conspirator and something like traitor, finally victim; but these are the "flat" characters (if one may so speak) of the treatise, not the "round" ones of the novel. And I cannot _unite_ them.

His love-affair with Marie de Gonzague leaves me cold. His friend, the younger De Thou, is hardly more than "an excellent person." The persecution of Urbain Grandier and the sufferings of the Ursuline Abbess seem to me--to use the old schoolboy word--to be hopelessly "m.u.f.fed"; and if any one will compare the accounts of the taking of the "Spanish bastion" at Perpignan with the exploit at that other bastion--Saint-Gervais at Roch.e.l.le--he will see what I mean as well as in any single instance. The second part, where we come to the actual conspiracy, is rather better than the first, if not much; and I think Vigny's presentment of Richelieu has been too much censured. Armand Duplessis was a very great man; but unless you accept the older Machiavellian and the more modern German doctrines as to what a great man may do, he must also be p.r.o.nounced a most unscrupulous one; while there is little doubt (unless you go back to Louis XI.) that Vigny was right in regarding him as the original begetter of the French Revolution. But he is not here made by any means wholly inhuman, and Vigny makes it justly clear that, if he had not killed Cinq-Mars, Cinq-Mars would have killed him. In such cases of course the person who begins may be regarded as the a.s.sa.s.sin; but it is doubtful whether this is distributive justice of the highest order. And I do not see much salvation for France in Henry d'Effiat.

This, however, is a digression from our proper subject, but one justifying itself after a fashion, inasmuch as it results from Vigny's own faulty handling of the subject itself and is appropriate to his line of argument in his _Examen_. He has written the novel not as he ought and as he ought not. The political and historical interests overshadow, confuse, and hamper the purely "fictional" (as people say now), and when he has got hold of a scene which _is_ either purely "fictional," or historical with fict.i.tious possibilities, he does not seem (to me) to know how to deal with it. There is one--of the extremest melodramatic character and opportunities--where, in a hut perched on the side of a Pyrenean gorge or canon, Richelieu's villainous tool, the magistrate Laubardemont; his mad niece, the former Ursuline Abbess, who has helped to ruin Urbain Grandier; his outcast son Jacques, who has turned Spanish officer and general bravo; and a smuggler who has also figured in the Grandier business, forgather; where the mad Abbess dies in terror, and Jacques de Laubardemont by falling through the flimsy hut-boards into the gorge, his father taking from him, by a false pretence before his death, the treaty between the Cinq-Mars conspirators and Spain. All this is sufficiently "horrid," as the girls in _Northanger Abbey_ would say, and divers French contemporaries of Vigny's from Hugo to Soulie would have made good horrors of it. In his hands it seems (to me) to miss fire. So, again, he has a well-conceived interview, in which Richelieu, for almost the last time, shows "the power of a strong mind over a weak one," and brings the King to abject submission and the surrender of Cinq-Mars, by the simple process of leaving his Majesty to settle by himself the problems that drop in from France, England, and where or whence not, during the time of the Cardinal's absence. It is less of a failure than the other, being more in Vigny's own line; but it is impossible not to remember several scenes--not one only--in _Quentin Durward_, and think how much better Scott would have done it; several in the Musketeer-trilogy, if not also in the Margot-Chicot series, and make a parallel reflection. And as a final parry by antic.i.p.ation to the objection that such comparison is "rascally," let it be said that nothing of the kind ever created any prejudice against the book in my case. I failed to get on with it long before I took the least trouble to discover critical reasons that might excuse that failure.

[Sidenote: _Stello_ less of a novel, but containing better novel-stuff.]

But if any one be of taste sufficiently like mine to find disappointment of the unpleasant kind in _Cinq-Mars_, I think I can promise him an agreeable, if somewhat chequered, surprise when, remembering _Cinq-Mars_ and basing his expectations upon it, he turns to _Stello_. It is true that the book is, as a whole, even less "precisely a novel" than Sainte-Beuve's _Volupte_. But for that very reason it escapes the display of the disabilities which _Cinq-Mars_, being, or incurring obligation to be, precisely a novel, suffers. It is true also that it exhibits that fancy for putting historical persons in the first "plan"

which he had avowed, and over which heads have been shaken. The bulk of it, indeed, consists of romanticised _histoires_ or historiettes (the narrator calls them "anecdotes") of the sad and famous fates of two French poets, Gilbert and Andre Chenier, and of our English Chatterton.

But, then, no one of these can be called "a dominant historical personage," and the known facts permit themselves to be, and are, "romanticised" effectively enough. So the flower is in each case plucked from the nettle. And there is another flower of more positive and less compensatory kind which blooms here, which is particularly welcome to some readers, and which, from _Cinq-Mars_ alone, they could hardly have expected to find in any garden of Alfred de Vigny's. For this springs from a root of ironic wit which almost approaches humour, which, though never merry, is not seldom merciful, and is very seldom actually savage, though often sad. Now irony is, to those who love it, the saving grace of everything that possesses it, almost equal in charm, and still more nearly equal in power, to the sheer beauty, which can dispense with it, but which sometimes, and not so very rarely, is found in its company.

[Sidenote: Its framework and "anecdotes."]

The substance, or rather the framework, of _Stello, ou Les Diables Bleus_, requires very little amplification of its double t.i.tle to explain it. Putting that t.i.tle in charade form, one might say that its first is a young poet who suffers from its second--like many other young persons, poetical and unpoetical, of times Romantic and un-Romantic.

Having an excessively bad fit of his complaint, he sends for a certain _docteur noir_ to treat the case. This "Black Doctor" is not a trout-fly, nor the sort of person who might be expected in a story of _diablerie_. It is even suggested that he derived the name, by which he was known to society, from the not specially individual habit of wearing black clothes. But there must have been something not quite ordinarily human about him, inasmuch as, having been resident in London at the time of Chatterton's death in 1770, he was--apparently without any signs of Old Parr-like age--a fashionable doctor at Paris in the year 1832. His visit ends, as usual, in a prescription, but a prescription of a very unusual kind. The bulk of it consists of the "anecdotes"--again perhaps not a very uncommon feature of a doctor's visit, but told at such length on the three subjects above mentioned that, with "links" and conclusion,[251] they run to nearly four hundred pages.

It is possible that some one may say "_Connu!_" both to the stories themselves and to the moral of real suffering, as opposed to mere megrim, which is so obviously deducible from them. But Stello was quite as clever as the objectors, and knew these things quite as well--perhaps, as far as the case of Gilbert is concerned, rather better than most Englishmen. It is in the manner of the Black Doctor's telling and handling that the charm lies.

[Sidenote: The death of Gilbert.]

Even for those gluttons of matter who do not care much for manner there is a good deal in the three stories. The first avails itself--as Vigny had unwisely _not_ availed himself in _Cinq-Mars_, though he was well acquainted with Shakespeare and lesser English masters--of the mixture of comic and tragic. The suffering[252] of the unfortunate youth who was partly a French Chatterton and partly a French Clare, his strange visit to the benevolent but rather ineffectual Archbishop of Paris, and the scene at his death-bed, exhibit, at nearly its best, the tragic power which Vigny possessed in a very high, though not always well exercised, degree. And the pa.s.sage of the poet's death is of such _macabre_ power that one must risk a translation:

(_The doctor has been summoned, has found the patient in his garret, bare of all furniture save a bed with tattered clothes and an old trunk._)

His face was very n.o.ble and very beautiful; he looked at me with fixed eyes, and between them and the nose, above the cheeks, he showed that nervous contraction which no ordinary convulsion can imitate, which no illness gives, but which says to the physician, "Go your ways!" and is, as it were, a standard which Death plants on his conquests. He clutched in one hand his pen, his poor last pen, inky and ragged, in the other a crust of his last piece of bread. His legs knocked together, so as to make the crazy bed crackle. I listened carefully to his hard breathing; I heard the rattle with its hollow husk; and I recognised Death in the room as a practised sailor recognises the tempest in the whistle of the wind that precedes it.