INTRODUCTION 
I. THE DOUBLE ROOT OF STYLE
EDWIG RICHTER relates in his reminiscences how once, when he was in Tivoli as a young man, he and three friends set out to paint part of the 
landscape, all four firmly resolved not to deviate from nature by a hair's?breadth; and although the subject was the same, and each quite credit?ably reproduced what his eyes had seen, the result was four totally different pictures, as different from each other as the personalities of the four painters. Whence the narrator drew the conclusion that there is no such thing as ob?jective vision, and that form and colour are always apprehended differently according to temperament. 
For the art historian, there is nothing surprising in this observation. It has long been realised that every painter paints "with his blood". All the distinc?tion between individual masters and their "hand" is ultimately based on the fact that we recognise such types of individual creation. With taste set in the same direction (we should probably find the four Tivoli landscapes rather similar, of a Preraphaelite type), the line will be in one case more angular, in another rounder, its movement here rather halting and slow, there more streaming and urgent. And, just as proportions tend now to the slender, now to the broad, so the modelling of the human body appeals 
something rather full and fleshy, while the same curves and hollows will be seen by another with more reticence, with much more economy. It is the same with light and colour. The sincerest intention to observe accurately cannot pre?vent a colour looking now warmer, now cooler, a shadow now softer, now harder, a light now more languid, now more vivid and glancing. If we are no longer bound by a common subject from nature, these individual styles become, of course, much more distinct. Botticelli and Lorenzo di Credi are two painters related by epoch and race, both Florentines of the later Quattro?cento. But when Botticelli* 1 draws a female body, its stature and shape is perceived in a way peculiar to him, and as radically and un- mistakably different from any female nude of Lorenzo's* as an oak from a lime. The impetuosity of Botticelli's drawing endows every form with a peculiar verve and animation. In Lorenzo's deliberate modelling, vision is essentially fulfilled by the object in repose. Nothing is more illuminating than to compare the similar curve of the arm in the two pictures. The sharp elbow, the spirited line of the forearm, the radiant spread of the fingers on the breast, the energy which charges every line-that is Botticelli. Credi, on the other hand, produces a more flaccid effect. Though very convincingly modelled, that is, conceived in volumes, his form still does not possess the impetus of Botticelli's con?tours. That is a difference of temperament, and that difference penetrates throughout, whether we compare the whole or the details. In the drawing of a mere nostril, we have to recognise the essential character of a style. For Credi, a definite person posed. That is not the case with Botticelli, yet it is not difficult to sec that the conception of form in the two artists is bound up with a definite notion of beautiful form and beautiful movement, and if Botticelli has given full play to his ideal of form in the slender erectness of his figure, even with Credi we feel that the special case of reality has in no way prevented him from express?ing his temperament in the pose and pro?portions of his figure. 
The psychologist of style finds a parti?cularly rich booty in the stylised drapery of this epoch. With relatively few elements, an enormous variety of widely differing individual expression has here come to birth. Hundreds of artists have depicted the Virgin seated with the drapery pouched between the knees, and every time a form has been found which reveals a whole man. And yet it is not only in the great line of Italian renaissance art, but even in the painterly 1 style of the Dutch genre painters of the seventeenth century that drapery has Lorenzo di Credi
this psychological significance. 
As is well known, satin was a favourite subject of Terborch's,* and he painted it specially well. It seems as if the fine material could not look 
otherwise than it is shown here, yet it is only the artist's innate dis?
tinction which speaks to us in his forms, and even Metsu * saw the 
phenomenon of these fold-formations essentially differently. The fabric is apprehended as something rather weighty in fall and fold, the ridge of the fold is less delicate, each of its curves lacks elegance, and from the whole 
sequence of folds, the pleasing ease, the brio has vanished. It is still satin, and painted by a master, but seen beside Terborch's, Metsu's fabric looks almost dull. 
And now, in our picture, that is not merely the result of a chance off-day. 
The spectacle is repeated, and so characteristic is it that we can continue on 
1 Malerisch. This word has, in the German, two distinct meanings, one objective, a quality residing in the object, the other subjective, a mode of apprehension and creation. To avoid confusion, they have been distinguished in English as "picturesque" and "painterly" re?spectively. (Tr.) the same lines if we proceed to the analysis of figures and grouping. Consider the bare arm of the music-making lady in Terborch's picture-how finely it is felt in joint and movement, and how much heavier Metsu's figure seems?not because it is less skilfully drawn, but because it is felt differently. In Terborch, the grouping is light and the figures are bathed in air. Metsu gives something more massive and compact. An accumulation such as the bundled folds of the thick table-cloth with the writing materials could not be found in Terborch. 
And so on. And if, in our reproduction, there is little trace of the shimmer?ing lightness of Terborch's tonal gradations; the rhythm of the whole still speaks an audible language, and it requires no special persuasion to see in the equipoise of the parts an art inwardly related to the drawing of the folds. 
The problem remains identical in the trees of landscape painters. A bough, a fragment of a bough, and we can say whether Ruysdael or Hobbema is the painter, not from isolated external features of the "manner", but because all the essentials of the sense of form exist even in the smallest fragment. Hobbema's * trees, even when he paints the same species as Ruysdael, * will always seem lighter, their outlines are freer, they rise more airily in space. Ruysdael's graver style charges the line with a peculiar ponderous emphasis, he loves the slowly undulating outline, he holds the masses of foliage more compactly together, and thoroughly characteristic of his pictures is the way in which he prevents any separation of the individual forms, but gives a close?knit weft. Trees and mountain contours meet in sombre contact. While Hobbema, on the other hand, loves the graceful, bounding line, the diffused mass, the variegated terrain, and charming vignettes and vistas-every part seems like a picture within a picture. 
\"\(Tith ever-increasing subtlety, we must try in this way to reveal the connec?tion of the part with the whole, so that we may arrive at the definition of individual types of style, not only in design, but in lighting and colour. We shall realise that a certain conception of form is necessarily bound up with a certain tonality and shall gradually come to understand the whole complex of personal characteristics of style as the expression of a certain temperament. For descriptive art history there is much to be done here. 
The course of the development of art, however, cannot simply be reduced to a series of separate points. Individuals fall into larger groups. Botticelli and Lorenzo di Credi, for all their differences, have still, as Florentines, a certain resemblance when compared with any Venetian, and Hobbema and Ruysdael, however divergent they may be, are immediately homogeneous as soon as to them, as Dutchmen, a Fleming like Rubens is opposed. That is to say: to the personal style must be added the style of the school, the country, the race. 
Let us define Dutch art by contrasting it with Flemish art. The flat meadows round Antwerp present in themselves no other scene than the Dutch pastures to which native artists have given the expression of the most widespread tranquillity. But when Rubens* handles these themes, the subject looks totally different: the earth rolls in vigorous waves, tree-trunks writhe passion?ately upwards, and their foliage is handled so completely in closed masses that Ruysdael and Hobbema in comparison appear as equally delicate sil?houettists. Dutch subtlety beside Flemish massiveness. In comparison with the energy of movement in Rubens' design, Dutch design in general is restful, whether it be the rise of a hill or the curve of a petal. No Dutch tree-trunk has the dramatic force of the Flemish movement, and even Ruysdael's mighty oaks look slender beside Rubens' trees. Rubens raises the horizon high and makes the picture heavy, the Dutch relation of sky and earth is radically different: the horizon lies low, and it even happens that four-fifths of the picture is given up to air. 
These are considerations which only become valuable when they can be generalised. The subtlety of Dutch landscape must be linked up with allied phenomena and pursued into the domain of the tectonic. The courses of a brick wall or the weaving of a basket are felt in Holland as peculiarly as the foliage of the trees. It is characteristic that not only a miniaturist like Dow but even a narrator like Jan Steen has time, in the midst of the most boisterous scene, for the accurate drawing of a wicker work. And the network of whitened joints on a brick wall, the pattern of neatly set flagstones, all these small details are really enjoyed by the architectural painters. As to Dutch architec?ture proper, however, we may say that stone seems here to have achieved a quite specific lightness. A typical building such as the Rathaus of Amsterdam avoids everything which, had it been conceived by a Flemish imagination, might have invested the great mass of stone with an appearance of weight.
\Ve encounter here at all points the bases of national feeling, where the sense of form comes into immediate contact with spiritual and moral elements, and art history has grateful tasks before it as soon as it takes up systematically this question of the national psychology of form. Everything hangs together. The still poses of the Dutch figure pictures also form the bases of the objects of the architectural world. But if we bring Rembrandt into the matter, with his feeling for the living quality of light which, withdrawing from every sub?stantial form, moves mysteriously in infinite spaces, we might easily be tempted to develop the observation into an analysis of Germanic art in con?trast to Romanesque art in general. 
But here the problem already branches. Although in the seventeenth cen?tury, Dutchman and Fleming are still clearly distinguishable, we cannot forthwith base a general judgment of a national type on one single epoch. Different times give birth to different art. Epoch and race interact. We must first establish how many general traits a style contains before we can give it the name of a national style in a special sense. However profoundly Rubens may impress his personality on his landscape, and however many talents may veer to his pole, we cannot admit that he was an expression of "permanent" national character to the same extent as contemporary Dutch art. The colour of time is stronger with him. His art is powerfully affected by a particular cultural current, the mode of feeling of Roman baroque, and so it is he, rather than the "timeless" Dutch artists, who challenges us to form an idea of what we must call "period" style. 
This idea is best to be obtained in Italy, because the development there fulfilled itself independently of outside influences and the general nature of the Italian character remains fully recognisable throughout. The transition from renaissance to baroque is a classic example of how a new zeitgeist en?forces a new form. 
Here we enter upon much-trodden paths. Nothing is more natural to an history than to draw parallels between periods of culture and periods of style. The columns and arches of the High Renaissance speak as intelligibly of the spirit of the time as the figures of Raphael, and a baroque building represents the transformation of ideals no less clearly than a comparison between the sweeping gestures of Guido Reni and the noble restraint and dignity of the Sistine Madonna. 
Let us this time remain on strictly architectural ground. The central idea of the Italian Renaissance is that of perfect proportion. In the human figure as in the edifice, this epoch strove to achieve the image of perfection at rest within itself. Every form developed to self-existent being, the whole freely co-ordinated: nothing but independently living parts. The column, the panel, the volume of a single element of a space as of a whole space-nothing here but forms in which the human being may find an existence satisfied in itself, extending beyond human measure, but always accessible to the imagination. With infinite content, the mind apprehends this art as the image of a higher, free existence in which it may participate. 
The baroque uses the same system of forms, but in place of the perfect, the completed, gives the restless, the becoming, in place of the limited, the con?ceivable, gives the limitless, the colossal. The ideal of beautiful proportion vanishes, interest concentrates not on being, but on happening. The masses, heavy and thickset, come into movement. Architecture ceases to be what it was in the Renaissance, an art of articulation, and the composition of the building, which once raised the impression of freedom to its highest pitch, yields to a conglomeration of parts without true independence. 
This analysis is certainly not exhaustive, but it will serve to show in what way styles express their epoch. It is obviously a new ideal of life which speaks to us from Italian baroque, and although we have placed architecture first as being the most express embodiment of that ideal, the contemporary painters and sculptors say the same thing in their own language, and whoever tries to reduce the psychic bases of style to abstract principles will probably find the decisive word here more readily than with the architects. The relation?ship of the individual to the world has changed, a new domain of feeling has opened, the soul aspires to dissolution in the sublimity of the huge, the in?finite. "Emotion and movement at all costs." Thus does the Cicerone formu?late the nature of this art. 
We have, in thus sketching three examples of individual style, national style, and period style, illustrated the aims of an art history which conceives style primarily as expression, expression of the temper of an age and a nation as well as expression of the individual temperament. It is obvious that with all that, the quality of the work of art is not touched: temperament certainly makes no work of art, but it is what we might call the material element of style taken in the broad sense that the particular ideal of beauty ( of the in?dividual as of the community) is included in it too. Works of art history of this kind are still far from the perfection they might attain, but the task is inviting and grateful. 
Artists are certainly not readily interested in historical questions of style. They take work exclusively from the standpoint of quality-is it good, is it self-sufficing, has nature found a vigorous and clear presentment? Everything else is more or less indifferent. We have but to read Hans van Marees when he writes that he is learning to attach less and less value to schools and person?alities in order only to keep in view the solution of the artistic problem, which is ultimately the same for Michelangelo as for Bartholomew van der Helst. Art historians who, on the other hand, take the differences between the finished products as their point of departure have always been exposed to the scorn of the artists: they have taken the detail for the essence: they cling just to the non-artistic side in man in wishing to understand art as expression only. We can very \Vell analyse the temperament of an artist and still not explain how the work came into being, and the description of all the differences be?tween Raphael and Rembrandt is merely an evasion of the main problem, because the important point is not to show the difference between the two but how both, in different ways, produced the same thing-namely, great art. 
It is hardly necessary here to take up the cudgels for the art historian and defend his work before a dubious public. The artist quite naturally places the general canon of art in the foreground, but we must not carp at the historical observer with his interest in the variety of forms in which art appears, and it remains no mean problem to discover the conditions which, as material element-call it temperament, zeitgeist, or racial character-determine the style of individuals, periods, and peoples. 
Yet an analysis with quality and expression as its objects by no means exhausts the facts. There is a third factor-and here we arrive at the crux of this enquiry-the mode of representation as such. Every artist finds certain visual possibilities before him, to which he is bound. Not everything is possible at all times. Vision itself has its history, and the revelation of these visual strata must be regarded as the primary task of art history. 
Let us try to make the matter clear by examples. There are hardly two artists who, although contemporaries, are more widely divergent by tempera?ment than the baroque master Bernini and the Dutch painter Terborch Con?fronted with the turbulent figures of Bernini, who will think of the peaceful, delicate little pictures ofTerborch? And yet, if we were to lay drawings by the two masters side by side and compare the general features of the technique, we should have to admit that there is here a perfect kinship. In both, there is that manner of seeing in patches instead of lines, something which we can call painterly, which is the distinguishing feature of the seventeenth century in comparison with the sixteenth. We encounter here a kind of vision in which the most heterogeneous artists can participate because it obviously does not bind them to a special mode of expression. Certainly an artist like Bernini needed the painterly style to say what he had to say, and it is absurd to wonder how he would have expressed himself in the draughtsmanly style of the sixteenth century. But we are clearly dealing with other concepts here than when we speak, for instance, of the energy of the baroque handling of masses in contrast to the repose and reserve of the High Renaissance. Greater or less movement are expressional factors which can be measured by one and the same standard: painterly and draughtsmanly, on the other hand, are like two languages, in which everything can be said, although each has its strength in a different direction and may have proceeded to visibility from a different angle. 
Another example. We can analyse Raphael's line from the point of view of expression, describe its great noble gait in contrast to the pettier fussiness of Quattrocento outlines: we can feel in the movement of the line in Giorgione's Venus the kinship with the Sistine Madonna and, turning to sculpture, dis?cover in Sansovino's youthful Bacchus the new, long, continuous line, and nobody will deny that we feel in this great creation the breath of the new sixteenth century feeling: it is no mere superficial history-writing to connect in this way form and spirit. But the phenomenon has another side. By ex?plaining great line, we have not explained line. It is by no means a matter of course that Raphael and Giorgione and Sansovino sought expressive force and formal beauty in line. But it is again a question of international connections. The same period is for the north, too, a period of line, and two artists who, as personalities, have little in common, Michelangelo and Hans Holbein the Younger, resemble each other in that they both represent the type of quite strictly linear design. In other words, there can be discovered in the history of style a substratum of concepts referring to representation as such, and one could envisage a history of the development of occidental seeing, for which the variations in individual and national characteristics would cease to have any importance. It is certainly no easy task to reveal this inward visual de?velopment, because the representational possibilities of an epoch are never shown in abstract purity but, as is natural, are always bound to a certain ex?pressional content, and the observer is then generally inclined to seek in the expression the explanation of the whole artistic product. 
When Raphael erects his pictorial edifices and, by strict observance of rules, achieves the impression of reserve and dignity to an unprecedented degree, we can find in his special problem the impulse and the goal, and yet the tectonics of Raphael are not entirely to be attributed to an intention born of a state of mind: it is rather a question of a representational form of his epoch which he only perfected in a certain way and used for his own ends. Similar solemn ambitions were not lacking later, but it was impossible to revert to his formulas. French classicism of the seventeenth century rests on another visual basis, and hence, with a similar intention, necessarily arrives at other results. By attributing everything to expression alone, we make the false assumption that for every state of mind the same expressional methods were always available. 
And when we speak of the progress of imitation, of the new impressions of nature which an epoch produced, that is also a material element which is bound to a priori forms of representation. The observations of the seventeenth century were not merely woven into the fabric of Cinquecento art. The whole groundwork changed. It is a mistake for art history to work with the clumsy notion of the imitation of nature, as though it were merely a homogeneous process of increasing perfection. All the increase in the "surrender to nature" does not explain how a landscape by Ruysdael differs from one by Patenir, and by the "progressive conquest of reality" we have still not explained the contrast between a head by Frans Hals and one by Diirer. The imitative content, the subject matter, may be as different in itself as possible, the decisive point remains that the conception in each case is based on a different visual schema-a schema which, however, is far more deeply rooted than in mere questions of the progress of imitation. It conditions the architectural work as well as the work of representative art, and a Roman baroque fa<;ade has the same visual denominator as a landscape by Van Goyen. 
2. THE MOST GENERAL REPRESENTATIONAL FORMS
This volume is occupied with the discussion of these universal forms of representation. It does not analyse the beauty of Leonardo but the element in which that beauty became manifest. It does not analyse the representation of nature according to its imitational content, and how, for instance, the naturalism of the sixteenth century may be distinguished from that of the seventeenth, but the mode of perception which lies at the root of the repre?sentative arts in the various centuries. 
Let us try to sift out these basic forms in the domain of more modern art. We denote the series of periods with the names Early Renaissance, High Renaissance, and Baroque, names which mean little and must lead to mis?understanding in their application to south and north, but are hardly to be ousted now. Unfortunately, the symbolic analogy bud, bloom, decay, plays a secondary and misleading part. If there is in fact a qualitative difference between the fifteenth and sixteenth centuries, in the sense that the fifteenth had gradually to acquire by labour the insight into effects which was at the free disposal of the sixteenth, the (classic) art of the Cinquecento and the (baroque) art of the Seicento are equal in point of value. The word classic here denotes no judgment of value, for baroque has its classicism too. Baroque (or, let us say, modern art) is neither a rise nor a decline from classic, but a totally different art. The occidental development of modern times cannot simply be reduced to a curve with rise,height, and decline: it has two culmin?ating points. We can turn our sympathy to one or to the other, but we must realise that that is an arbitrary judgment, just as it is an arbitrary judgment to say that the rose-bush lives its supreme moment in the formation of the flower, the apple-tree in that of the fruit. 
For the sake of simplicity, we must speak of the sixteenth and seventeenth centuries as units of style, although these periods signify no homogeneous production, and, in particular, the features of the Seicento had begun to take shape long before the year 1600, just as, on the other hand, they long con?tinued to affect the appearance of the eighteenth century. Our object is to compare type with type, the finished with the finished. Of course, in the strictest sense of the word, there is nothing "finished": all historical material is subject to continual transformation; but we must make up our minds to establish the distinctions at a fruitful point, and there to let them speak as contrasts, if we are not to let the whole development slip through our fingers. The preliminary stages of the High Renaissance are not to be ignored, but they represent an archaic form of art, an art of primitives, for whom established pictorial form does not yet exist. But to expose the individual differences which lead from the style of the sixteenth century to that of the seventeenth must be left to a detailed historical survey which will, to tell the truth, only do justice to its task when it has the determining concepts at its disposal. 
If we are not mistaken, the development can be reduced, as a provisional formulation, to the following five pairs of concepts: 
(1) The development from the linear to the painterly, i.e. the development of line as the path of vision and guide of the eye, and the gradual depreciation of line: in more general terms, the perception of the object by its tangible character-in outline and surfaces-on the one hand, and on the other, a per?ception which is by way of surrendering itself to the mere visual appearance and can abandon "tangible" design. In the former case the stress is laid on the limits of things; in the other the work tends to look limitless. Seeing by volumes and outlines isolates objects: for the painterly eye, they merge. In the one case interest lies more in the perception of individual material objects as solid, tangible bodies; in the other, in the apprehension of the world as a shifting semblance. (2) The development from plane to recession. Classic I art reduces the parts of a total form to a sequence of planes, the baroque emphasises depth. Plane is the element of line, extension in one plane the form of the greatest explicit?ness: with the discounting of the contour comes the discounting of the plane, and the eye relates objects essentially in the direction of forwards and back?wards. This is no qualitative difference: with a greater power of representing spatial depths, the innovation has nothing directly to do: it signifies rather a radically different mode of representation, just as "plane style" in our sense is not the style of primitive art, but makes its appearance only at the moment at which foreshortening and spatial illusion are completely mastered. 
(3) The development from closed to open form. Every work of art must be a finite whole, and it is a defect if we do not feel that it is self-contained, but the interpretation of this demand in the sixteenth and seventeenth cen?turies is so different that, in comparison with the loose form of the baroque, classic design may be taken as the form of closed composition. The relaxation of rules, the yielding of tectonic strength, or whatever name we may give to the process, does not merely signify an enhancement of interest, but is a new mode of representation consistently carried out, and hence this factor is to be adopted among the basic forms of representation. 
(4) The development from multiplicity to unity. In the system of a classic composition, the single parts, however firmly they may be rooted in the whole, maintain a certain independence. It is not the anarchy of primitive art: the part is conditioned by the whole, and yet does not cease to have its own life. For the spectator, that presupposes an articulation, a progress from part to part, which is a very different operation from perception as a whole, such as the seventeenth century applies and demands. In both styles unity is the chief aim (in contrast to the pre-classic period which did not yet understand the idea in its true sense), but in the one case unity is achieved by a harmony of free parts, in the other, by a union of parts in a single theme, or by the sub?ordination, to one unconditioned dominant, of all other elements. 
(5) The absolute and the relative clarity of the subject. This is a contrast which at first borders on the contrast between linear and painterly. The representation of things as they are, taken singly and accessible to plastic feeling, and the representation of things as they look, seen as a whole, and rather by their non-plastic qualities. But it is a special feature of the classic age that it developed an ideal of perfect clarity which the fifteenth century only vaguely suspected, and which the seventeenth voluntarily sacrificed. Not that artistic form had become confused, for that always produces an unpleasing effect, but the explicitness of the subject is not longer the sole purpose of the presentment. Composition, light, and colour no longer merely serve to define form, but have their own life. There are cases in which absolute clarity has been partly abandoned merely to enhance effect, but "relative" clarity, as a great all-embracing mode of representation, first entered the history of art at the moment at which reality is beheld with an eye to other effects. Even here it is not a difference of quality if the baroque departed from the ideals of the age of Diirer and Raphael, but, as we have said, a different attitude to the world. 
3, IMITATION AND DECORATION 
The representational forms here described are of such general significance that even widely divergent natures such as Terborch and Bernini-to repeat an example already used-can find room within one and the same type. The community of style in these two painters rests on what, for people of the seventeenth century, was a matter of course-certain basic conditions to which the impression of living form is bound without a more special expres?sional value being attached to them. 
They can be treated as forms of representation or forms of beholding: in these forms nature is seen, and in these forms art manifests its contents. But it is dangerous to speak only of certain "states of the eye" by which conception is determined: every artistic conception is, of its very nature, organised accord?ing to certain notions of pleasure. Hence our five pairs of concepts have an imitative and a decorative significance. Every kind of reproduction of nature moves within a definite decorative schema. Linear vision is permanently bound up with a certain idea of beauty and so is painterly vision. If an ad?vanced type of art dissolves the line and replaces it by the restless mass, that happens not only in the interests of a new verisimilitude, but in the interests of a new beauty too. And in the same way we must say that representation in a plane type certainly corresponds to a certain stage of observation, but even here the schema has obviously a decorative side. The schema certainly yields nothing of itself, but it contains the possibility of developing beauties in the arrangement of planes which the recessional style no longer possesses and can no longer possess. And we can continue in the same way with the whole series. 
But then, if these more general concepts also envisage a special type of beauty, do we not come back to the beginning, where style was conceived as the direct expression of temperament, be it the temperament of a time, of a people, or of an individual? And in that case, would not the only new factor be that the section was cut lower down, the phenomena, to a certain extent, reduced to a greater common denominator? 
In speaking thus, we should fail to realise that the second terms of our pairs of concepts belong of their very nature to a different species, in so far as these concepts, in their transformations, obey an inward necessity. They represent a rational psychological process. The transition from tangible, plastic, to purely visual, painterly perception follows a natural logic, and could not be reversed. Nor could the transition from tectonic to a-tectonic, from the rigid to the free conformity to law. 
To use a parable. The stone, rolling down the mountain side, can assume quite different motions according to the gradient of the slope, the hardness or softness of the ground, etc., but all these possibilities are subject to one and the same law of gravity. So, in human psychology, there are certain developments which can be regarded as subject to natural law in the same way as physical growth. They can undergo the most manifold variations, they can be totally or partially checked, but, once the rolling has started, the operation of certain laws may be observed throughout. Nobody is going to maintain that the "eye" passes through : developments on its own ac?count. Conditioned and condi- tioning, it always impinges on other spiritual spheres. There is certainly no visual schema which, arising only from its own premisses, could be imposed on the world as a stereotyped pat?tern. But although men have at all times seen what they wanted to see, that does not exclude the possibility that a law remains operative throughout all change. To determine this law would be a central problem, the central problem of a history of art. 
We shall return to this point at the end of our enquiry. 
GENERAL OBSERVATIONS 
I. LINEAR (DRAUGHTSMANLY, PLASTIC) AND PAINTERLY
TACTILE AND VISUAL PICTURE 
IF we wish to reduce the difference between the art of Diirer and the art of Rembrandt to its most general formulation, we say that Diirer is a 
draughtsman and Rembrandt a painter. In speaking thus, we are aware of having gone beyond a personal judgment and characterised a difference of epoch. Occidental painting, which was draughtsmanly in the sixteenth century, developed especially on the painterly side in the seventeenth. Even if there is only one Rembrandt, a decisive readjustment of the eye took place every?where, and whoever has any interest in clearing up his relation to the world of visible forms must first get to grips with these radically different modes of vision. The painterly mode is the later, and cannot be conceived without the earlier, but it is not absolutely superior. The linear style developed values which the painterly style no longer possessed and no longer wanted to possess. They are two conceptions of the world, differently orientated in taste and in their interest in the world, and yet each capable of giving a perfect picture of visible things. 
Although in the phenomenon of linear style, line signifies only part of the matter, and the outline cannot be detached from the form it encloses, we can still use the popular definition and say for once as a beginning-linear style sees in lines, painterly in masses. Linear vision, therefore, means that the sense and beauty of things is first sought in the outline-interior forms have their outline too-that the eye is led along the boundaries and induced to feel along the edges, while seeing in masses takes place where the attention withdraws from the edges, where the outline has become more or less indifferent to the eye as the path of vision, and the primary element of the impression is things seen as patches. It is here indifferent whether such patches speak as colour or only as lights and darks. 
The mere presence oflight and shade, even if they play an important part, is still not the factor which decides as to the painterly character of the picture. Linear art, too, has to deal with bodies and space, and needs lights and shadows to obtain the impression of plasticity. But line as fixed boundary is assigned a superior or equal value to them. Leonardo is rightly regarded as the father of chiaroscuro, and his Last Supper in particular is the picture in which, for the first time in later art, light and shade are applied as a factor of composition on a large scale, yet what would these lights and darks be without the royally sure guidance which is exercised by the line? Everything depends on how far a preponderating significance is assigned to or withdrawn from the edges, whether they must be read as lines or not. In the one case, the line means a track moving evenly round the form, to which the spectator can confidently entrust himself; in the other, the picture is dominated by lights and shadows, not exactly indeterminate, yet without stress on the boundaries. Only here and there does a bit of palpable outline emerge: it has ceased to exist as a uniformly sure guide through the sum of the form. Therefore, what makes the difference between Diirer and Rembrandt is not a less or more in the exploita?tion of light and shade, but the fact that in the one case the masses appear with stressed, in the other with unstressed edges. 
As soon as the depreciation of line as boundary takes place, painterly possi?bilities set in. Then it is as if at all points everything was enlivened by a mysterious movement. ??'hile the strongly stressed outline fixes the present?ment, it lies in the essence of a painterly representation to give it an in?determinate character: form begins to play; lights and shadows become an independent element, they seek and hold each other from height to height, from depth to depth; the whole takes on the semblance of a movement ceaselessly emanating, never ending. Whether the movement be leaping and vehement, or only a gentle quiver and flicker, it remains for the spectator inexhaustible. 
We can thus further define the difference between the styles by saying that linear vision sharply distinguishes form from form, while the painterly eye on the other hand aims at that movement which passes over the sum of things. In the one case, uniformly clear lines which separate; in the other, unstressed boundaries which favour combination. Many elements go to produce the impression of a general movement-we shall speak of these-but the emanci?pation of the masses of light and shade till they pursue each other in inde?pendent interpiay remains the basis of a painterly impression. And that means, too, that here not the separate form but the total picture is the thing that counts, for it is only in the whole that that mysterious interflow of form and light and colour can take effect, and it is obvious that here the immaterial and incorporeal must mean as much as concrete objects .. 
When Di.irer* or Cranach places a nude as a light object on a dark ground, the elements remain radically distinct: background is background, figure is figure, and the Venus or Eve we see before us produces the effect of a white silhouette on a dark foil. Conversely, if a nude in Rembrandt* stands out on a dark ground, the light of the body seems as it were to emanate from the darkness of the picture space: it is as if everything were of the same stuff. The distinctness of the object in this case is not necessarily impaired. While the form remains perfectly clear, that peculiar union between the modelling lights and darks can have acquired a life of its own, and without the exigencies of the object being in any way prejudiced, figure and space, corporeal and incorporeal, can unite in the impression of an independent tonal movement. 
But certainly-to make a preliminary remark-it is of considerable advan?tage to "painters" to liberate lights and darks from their function of mere form-definition. A painterly impression most easily comes about when the lighting no longer subserves the distinctness of the objects, but passes over them: that is to say, when the shadows no longer adhere to the forms, but, in the conflict between the distinctness of the object and the illumination, the eye more willingly surrenders to the play of tones and forms in the picture. A painterly illumination-say in a church interior-is not the one which will make the columns and walls as distinct as possible, but, on the contrary, the one which will glide over the form and partially veil it. And in the same way, the silhouettes-if the notion can be used at all in this connection-will be apt to become inexpressive: a painterly silhouette never coincides with the form of the object. As soon as it speaks too clearly of the object, it isolates itself and checks the coalescence of the masses in the picture. 
But with all that the decisive word is not yet said. We must go back to the fundamental difference between draughtsmanly and painterly representation as even antiquity understood it-the former represents things as they are, the latter as they seem to be. This definition sounds rather rough, and to philo?sophic ears, almost intolerable. For is not everything appearance? And what kind of a sense has it to speak of things as they are? In art, however, these notions have their permanent right of existence. There is a style which, essen?tially objective in outlook, aims at perceiving things and expressing them in their solid, tangible relations, and conversely, there is a style which, more subjective in attitude, bases the representation on the picture, in which the visual appearance of things looks real to the eye, and which has often retained so little resemblance to our conception of the real form of things. 
Linear style is the style of distinctness plastically felt. The evenly firm and clear boundaries of solid objects give the spectator a feeling of security, as if he could move along them with his fingers, and all the modelling shadows follow the form so completely that the sense of touch is actually challenged. Representation and thing are, so to speak, identical. The painterly style, on the other hand, has more or less emancipated itself from things as they are. For it, there is no longer a continuous outline and the plastic surfaces are dissolved. Drawing and modelling no longer coincide in the geometric sense with the underlying plastic form, but give only the visual semblance of the thing. 
Where nature shows a curve, we perhaps find here an angle, and instead of an evenly progressive increase and decrease of light, light and shade now appear fitfully, in ungradated masses. Only the appearance of reality is seized -something quite different from what linear art created with its plastically conditioned vision, and just for that reason, the signs which the painterly style uses can have no further direct relation to the real form. The pictorial form remains indeterminate, and must not settle into those lines and curves which correspond to the tangibility of real objects.
The tracing out of a figure with an evenly clear line has still an element of physical grasping. The operation which the eye performs resembles the opera?tion of the hand which feels along the body, and the modelling which repeats reality in the gradation of light also appeals to the sense of touch. A painterly representation, on the other hand, excludes this analogy. It has its roots only in the eye and appeals only to the eye, and just as the child ceases to take hold of things in order to "grasp" them, so mankind has ceased to test the picture for its tactile values. A more developed art has learned to surrender itself to mere appearance. 
With that, the whole notion of the pictorial has shifted. The tactile picture has become the visual picture-the most decisive revolution which art history knows. 
Now we need not, of course, immediately think of the ultimate formula?tions of modern impressionist painting if we wish to form an idea of the change from the linear to the painterly type. The picture of a busy street, say, as Monet painted it, in which nothing whatsoever coincides with the form which we think we know in life, a picture with this bewildering alienation of the sign .from the thing is certainly not to be found in the age of Rembrandt, but the principle of impressionism is already there. Everybody knows the example of the turning wheel. In our impression of it, the spokes vanish, and in their place there appear indefinite concentric rings, and even the roundness of the felly has lost its pure geometric form. Now not only Velasquez, but even so discreet an artist as Nicolas Maes has painted this impression. Only when the wheel has been made indistinct does it begin to turn. A triumph of seeming over being. 
And yet that is, after all, only an extreme case. The new representation includes the stationary as well as the moving. When a sphere at rest is no longer represented by a geometrically pure circular form, but with a broken line, and where the modelling of the surface of a cube has degenerated into separate blocks of light and dark, instead of proceeding uniformly by im?perceptible gradations, we stand everywhere on impressionist ground. 
And now if it is true that the painterly style does not body forth things in themselves, but represents the world as seen, that is to say, as it actually appears to the eye, that also implies that the various parts of a picture are seen as a unity from the same distance. That seems to be a matter of course, but it is not so at all. The distance required for distinct seeing is relative: different things demand different vicinities of the eye. In one and the same form?complex, totally different problems may be presented to the eye. For instance, we see the forms of a head quite distinctly, but the pattern of the lace collar beneath it requires closer approach, or at least, a special adjustment of the eye if its forms are to become distinct. The linear style, as representation of being, had no difficulty in making this concession to the distinctness of the object. It was quite natural that things, each in its particular form, should be rendered in such a way as to be perfectly distinct. The demand for unified visual per?ception is radically non-existent for this type of art in its purest developments. Holbein, in his portraits, pursues the design of small goldsmith's work and embroidery into their smallest details. Frans Hals, on the other hand, occa?sionally painted a lace collar only as a white shimmer. He did not wish to give more than is perceived by the eye taking a general view of the whole. But, of course, the shimmer must look as if all the details were actually there, and the indistinctness were only the momentary effect of distance. 
The measure of what can be seen as a unity has been taken very variously. Although we are accustomed to describe only the higher degrees as impres?sionism, we must always bear in mind that these do not signify something essentially new. It would be difficult to fix the point at which the merely "painterly" ceases and "impressionist" begins. Everything is transition. And in the same way, it is hardly possible either to establish any ultimate expres?sion of impressionism which might be taken as its classic completion. That is