PRINT December 1979

Mondrian’s First Diamond Compositions

THE RECENT EXHIBITION “MONDRIAN: THE Diamond Compositions” at the National Gallery of Art in Washington occasioned the publication of a major catalogue in which the curator, E.A. Carmean, Jr., discussing all known diamond-shaped paintings and related drawings, not only asserts the prominent place of the diamond compositions in the evolution of Mondrian’s style, but also clarifies his working method to an extent unparalleled in earlier discussions of specific paintings.1 Of particular interest is the discussion of the changes in some of Mondrian’s works, for example in his last painting, Victory Boogie Woogie; as to the National Gallery’s own Diamond Composition in Red, Yellow and Blue and the Philadelphia Museum’s Composition in Blue, it appears that two diamond compositions in Seuphor’s classified catalogue, which were known only from photographs, are in fact earlier states, now hidden beneath the surface of these paintings. For a further investigation of Mondrian’s procedures Carmean’s observations are most helpful.

Yet Carmean’s account of “how it all started” is not altogether satisfactory. Quite rightly, he expresses his doubts about external sources for the paintings, such as heraldic, diamond-shaped escutcheons of the deceased that can be seen in 17th-century Dutch paintings of church interiors.2 But his allegiance to traditional art-historical “evolutionary” thinking seems to me too limited.

Carmean suggests a step-by-step development toward the earliest dated diamond composition, the amazing Lozenge with Gray Lines of 1918 (The Hague, Gemeente Museum), through the works of 1914–17. Essential to the argument is the centralized, roughly diamond-shaped arrangement of elements in three paintings of 1917, the large Composition with Lines and the two smaller Compositions with Color A and B, all three at Otterlo, and in a related drawing in the Holtzman collection that he rather prematurely labels Composition based on Diamond Shape. It is conveniently ignored, however, that after these works and before the 1918 diamond composition, Mondrian made at least eight paintings of quite a different character: horizontal compositions with an allover pattern of color planes placed on a white field or divided by gray lines, the planes being cut off at the edges by the frames.3 Thus, why Mondrian took all at once to a new format, the diamond, and to a new structure, the grid, is insufficiently explained.

While these questions may never be definitely answered, certainly there is more to say if one considers the circumstances in which Mondrian painted the first diamond compositions. Carmean is mainly concerned with the painter’s development in relation to the grand tradition of French Impressionism and Cubism, tending to isolate Mondrian from his immediate artistic environment. At the risk of chauvinism, I would argue that the fundamental changes apparent in Mondrian’s first diamond compositions were largely determined by his dialogue with fellow painters in wartime Holland, notably with some of the De Stijl artists. In the Washington catalogue, Theo van Doesburg and Bart van der Leck are merely mentioned as artists whom Mondrian became friends with, “learning and sharing their style of sharp geometric forms and pure color which by 1917 led to the de Stijl movement.”4 Only later are Van Doesburg’s Counter-Compositions of 1924 and afterward discussed in connection to some diamond compositions made by Mondrian in the mid-1920s. However, both the earlier works and statements in the still largely unpublished correspondence of Mondrian testify on his response to their art in this critical period around 1916–19.5

It may be useful to note some circumstantial data. Mondrian stayed in Holland from June 1914 until June 1919; then he retuned to Paris. In Holland he lived in Amsterdam and Domburg and, from early 1915 onwards, mainly in Laren, a small village not far from Amsterdam. His first written contact with Van Doesburg was prompted by the latter’s favorable review of Mondrian’s work in a group exhibition in Amsterdam, in November of 1915. They met in January 1916 at Mondrian’s place. In April 1916 Van der Leck moved from The Hague to Laren. Within weeks he got acquainted with Mondrian and, through him, with Van Doesburg.

Mondrian communicated frequently with both. He and Van der Leck often discussed their work and their ideas. With Van Doesburg, Mondrian exchanged some 80 letters before his return to Paris, and from these it becomes clear that they actually only rarely visited each other. Van Doesburg lived in Haarlem and, from January 1917, in Leiden, where he had his own circle of friends, among them the architects J.J.P. Oud and Jan Wils, the painter Vilmos Huszar and the poet Antony Kok. Van Doesburg, in a letter to Kok dated February 7, 1916, stated that “Mondrian is not someone for having friends. He prefers to stay alone, and does not like to meet the others.”6 Thus, contrary to popular belief, the artists connected with De Stijl were not close as a group, unlike, for example, the members of Die Brücke. There were more likely two groups, a small one in Laren and a larger one centered in Leiden, Van Doesburg being their intermediary.

There was never, even at the start, complete agreement among the artists participating in the movement about formal principles now often considered to be essential to De Stijl esthetics—the exclusive use of primary colors and horizontal and vertical lines. Exceptions to these “rules” can be found in the work of every painter involved, including Mondrian. What these artists shared were certain notions of a more philosophical and ideological nature rather than an art theory in a strict sense. The first manifesto, published in De Stijl in November 1918, tells little more than that “the new art has brought forward what the new consciousness of time contains: a balance between the universal and the individual.”7 Mondrian’s proposal to put in a more precise paragraph on Neoplasticism (Nieuwe Beelding) apparently failed to please the others.8 His letters to Van Doesburg make it clear that there were minor and more serious controversies, but nothing serious enough to cause a split until late 1918, when Van der Leck left De Stijl: then others soon followed.

Why, and under what circumstances, did Mondrian start to make paintings in a diamond shape? The four that he made in 1918–19 were all based on a mathematical grid, but there are at least five rectangular paintings from the same period employing a similar structural device. I concentrate here on the diamond format, only touching upon the problem of the grid, since it would take me too far from my main point to discuss the significant role of mathematical structures in Dutch art and architecture at the beginning of the century, an idealistic tradition to which Mondrian’s paintings certainly are related.9

The use of the diamond shape should not be seen separately from Mondrian’s renewed consideration of the diagonal. As is well known, he had confined himself to straight horizontal and vertical lines and conforming color planes from 1915 on, leaving out the occasional oblique lines and arcs that had figured in his earlier Cubist paintings. Yet for him the perpendicular was by no means obligatory, at least not until 1919. In his first writings, the series “De Nieuwe Beelding in de schilderkunst” (Neoplasticism in Painting), published in De Stijl from October 1917 until October 1918, Mondrian states time and again that the new art must express the unison of opposing forces, by means of lines and color planes at right angles. The relative position of the lines and the limits of the planes are paramount, not the horizontal and vertical directions per se: “Neoplasticism visualizes the unity of the inner and the outer by right-angled position.”10

In a letter sent in February 1918 to Van Doesburg, Mondrian reports a disagreement with Van der Leck on this subject; he says he wants his color planes to be rectangular, whereas Van der Leck thinks right angles irrelevant, provided that the limits of the planes are straight lines. Indeed, in 1918 Van der Leck had reintroduced color elements with acute and obtuse angles such as trapezoids and triangles which he had abandoned the year before, no doubt temporarily influenced by Mondrian.

Mondrian’s preoccupation with the “deviations” in Van der Leck’s work was shared by Van Doesburg, whose attitude was more popish than the pope’s, as often happens with converts. When Van der Leck asked him in a letter of March 22, 1918, if he would reproduce in De Stijl new work by his friend, the painter Peter Alma, Van Doesburg replied that he objected to Alma’s “oblique destruction”; in the latter’s paintings the diagonals did not meet with any opposition within the composition, and thus, according to Van Doesburg, suggested natural space in the tradition of perspective, whereas in the new art, space should be directly determined on the picture plane (the word “beelden,” used here, is hard to translate). He concluded his criticism, which Van der Leck could take to heart as well: “Alma’s work is, in my opinion, indefensible as art. I have written to Mondrian about it and will do it more extensively in connection with your proposal to reproduce his [i.e., Alma’s] work in De Stijl. Maybe we’ll understand each other’s point of view better.”11

It was probably in response to these letters that Mondrian wrote to Van Doesburg, in April or early May 1918: “Also on the oblique line I agree with you: if combined with the perpendicular I think it is to be rejected. In Van der Leck’s case I am not sure: his works still don’t seem individual; I think that is because he works in a way totally different from ours.” Here the split between Van Doesburg and Mondrian on one side, and Van der Leck on the other, becomes already apparent.

No doubt Van Doesburg’s exposure of the disagreement, in an article on Alma’s painting The Saw and the Goldfish Bowl, published in De Stijl in June 1918, added to the trouble. There he stated that the “oblique destructive field” in the center (which can be interpreted as a distorted diamond shape) was too dominant to achieve a balance with the triangular parts in the corners of the painting. Concluding, he more or less repeated what Mondrian had written to him: “The diagonal destruction, especially in combination with the horizontal and the vertical, inside or outside the painting, is still a rudiment [of the baroque],” baroque being a dirty word in Van Doesburg’s vocabulary.12

Mondrian’s letter to Van Doesburg from April or May 1918 contains a paragraph that is still more crucial to a discussion of the diamond paintings: “Some time ago I started a painting completely in diamonds, like this: [sketch of some crossing diagonal lines]. I’ll have to see if this is possible; intellectually I would say yes. There is something to say in favor of it, because the vertical and the horizontal can be found everywhere in nature: with diagonals I could dissolve that. But I think in that case there can never be added horizontal or vertical lines, or any sort of oblique lines [i.e., other than diagonals].”

The painting Mondrian refers to in this letter must have looked surprisingly different from anything known from him: a composition with a structure of diagonal lines and/or diamond shapes only. Given the diagonal structure, it is most likely that the painting was square, not rectangular. Whether it had a diamond format is not clear, either from the description or from the sketch in the letter to Van Doesburg. That seems unlikely, however, because Mondrian would have mentioned such an important deviation from his normal practice; indeed, he did some time later, as will be shown. Now there are two possibilities, as far as I can see: the painting may be lost, or, more probably, it was the first state of one of the lozenges of 1918–19, still placed in a perpendicular position.

It is hard to say which one of the four, but I venture it is either the Lozenge with Gray Lines, dated 1918, or the much smaller Composition: Bright Color Planes with Gray Lines (Otterlo, Rijksmuseum Kröller-Müller), dated 1919. Both paintings have evidently been heavily worked on over quite a long period, and when placed in a perpendicular position they show traces of a grid with horizontal, vertical and diagonal lines, which seems to belie the opinion Mondrian had expressed to Van Doesburg. However, close observation of the intersections of the lines reveals that in both paintings there was at first a grid of diagonals only, in accordance with the sketch in the letter.

It must have been Mondrian’s intention to use the grid not just as an image by itself, but as a base for an asymmetrical composition, as he did in rectangular grid paintings of the same years. In one of his next letters to Van Doesburg, dated June 13, 1918, he relates the visit of June 7th of Huszar, who told him he had started to make paintings on a mathematical grid.13 Mondrian continues: “I was also working on a painting that I showed to Huszar and he thought it was good; also with a regular division, without my prior knowledge of his use of it. But I rework that division a lot and I do think that it will look different from his work.”

It cannot be said when or why Mondrian revised his original objections to the combination of lines in three directions and decided to lay a grid of horizontal and vertical lines over the diagonal one, but the paint of the diagonal lines had already dried when he did. In Composition: Bright Color Planes with Gray Lines the grid was subsequently covered again with paint, while making it not entirely invisible. In the Lozenge with Gray Lines at The Hague the grid is retained; only the white (or rather, white in certain areas and light gray in others) has been built up in many layers. Probably Mondrian referred to this painting and/or the smaller Philadelphia Museum Lozenge with Gray Lines when he wrote to his friend, the clergyman Van Assendelft on January 6, 1919: “Now I am working on about twelve paintings, among them some in black and white, of course different from the ones you saw. I have not made more of those, because I am changing continually—without wanting to . . . I will show some small paintings in Amsterdam again at the end of February, but they are for Bremmer and have already been paid for by him.”14

The most radical change concerning the paintings, described here as squares with a diagonal composition, was still to come—quite suddenly, as is evident from the tone of a letter to Van Doesburg which from internal evidence must be dated to February 1919: “Further I wanted to tell you that I hang several pieces now like this: ♢, in order that the composition becomes like this +; whereas in this way □ the composition is like this ×, à la Van der Leck, for example. What do you think of it? I’ll hear your opinion later, when you have seen the works. Bye, in haste, Piet.” The conclusion seems inevitable to me that (1) at least some, but probably all four paintings now known as the first diamonds were originally executed and intended to be hung in a perpendicular position, with an overt diagonal composition, and (2) Mondrian’s decision to hang them like diamonds was made within the last two or three weeks, at the most, before the exhibition, leaving little time for internal changes.

Van Doesburg’s reaction to this was positive, as is obvious from Mondrian’s next letter, dated March 3, 1919: “Dear Does, thank you for your letter . . . It pleased me that you agree in principle with hanging them like diamonds, and I believe that in practice you will think it right for some of my things, Because you really look at the things, and not just at its appearance, as, for instance, Mrs. Steenhoff does, who wrote to me that they looked like handkerchiefs placed on the wall!”

This Mrs. Steenhoff, the wife of a critic and museum director, W. Steenhoff, had seen the “handkerchiefs” at the exhibition already mentioned by Mondrian in his letter to Van Assendelft: the spring exhibition of the Hollandsche Kunsteraarskring held from February 22nd to March 23, 1919. The catalogue mentions five compositions by Mondrian, marked A, B, C, D, E, which are not for sale. According to Joosten, press reviews make it clear that among them were Composition: Bright Color Planes with Gray Lines and Composition in Diamond Shape.15 The two lozenges with gray lines were most likely not in the exhibition; in fact, they are not known ever to have been exhibited in Mondrian’s lifetime. Also, unlike the other two, they were never reproduced in De Stijl. Neither was bought by H.P. Bremmer, who advised the collector Mrs. Kröller-Müller and paid Mondrian a monthly allowance in exchange for paintings: he got the other two diamond compositions.

All this makes the position of the lozenges with gray lines highly problematic in Mondrian’s oeuvre. Did he consider them as successful, or at least as definitively finished paintings, at that time? I doubt it, especially in the case of the painting at Philadelphia, with its unusual thin layer of paint barely covering the canvas. It seems most likely to me that Mondrian abandoned them and did not show them publicly because they exposed too many of the uncertainties about such an important issue as the use of horizontal/vertical and diagonal lines which he had to resolve at that period. Yet he did not entirely reject them, otherwise he would not have hung them on the walls of his studio in Paris in the twenties, and he would not have given one away to his friend A.P. Van den Briel (after 1926) or sold the other to Walter Arensberg in 1937.

Unsolved as certain problems concerning the first diamond compositions still are, it is practically certain that Mondrian’s decision to put them in a diamond position was his final answer to the challenge of Van der Leck’s (and Alma’s) use of the diagonal. Van Doesburg’s rejection of their liberalism had obviously been a moral support for Mondrian when he tried to find his way out of the dilemma.

To make things more complicated, it could be argued that even Mondrian’s choice of the diamond format had been inspired by some works of fellow De Stijl painters. Late in 1917 Van Doesburg had designed a series of stained-glass windows for a building by the architect Jan Wils. In these windows a diamond-shaped framework was inserted in a square, the whole being divided into eight triangular parts rotated around the center. One may speculate as to how Van Doesburg would have defended his use of diagonal lines within a rectangular field while at about the same time attacking Alma for it. Presumably Van Doesburg was of the opinion that, unlike Alma, he had succeeded in realizing the right balance between the diamond shape and the perpendicular composition of the glass panels.

Another possible inspiration for Mondrian may have been Van der Leck’s design, made in 1918, for the decoration of a room in the Leeuwerik house at Blaricum, near Laren. Among other shapes, Van der Leck projected colored diamonds the size of paintings onto the walls, the floor and even the curtains of the room. Interesting also are the small drawings in the margin of the design, showing a side view of a table and a top view of a tablecloth, decorated with a typical Van der Leck image, put in a diamond position. There is no roof, but it is quite possible that Mondrian had seen these designs or had been told about them.16

Both Van Doesburg’s and Van der Leck’s designs had to do with architecture. For his part, Mondrian was aware of the problems aroused by his diamond compositions when placed in a tectonic environment. In the letter already quoted to Van Doesburg of March 3, 1919, he confided: “In relation to architecture one could possibly have some objections, but you know that I personally am still making paintings so to say (that is, separate from the building).” Indeed, while he accepted the use of the diamond shape, with its diagonal cutting edges, for autonomous painting, and expanded its possibilities considerably throughout his subsequent career, he did not apply this device directly in architecture when, after his return to Paris, he started to decorate his studio; but he hung his diamond-shaped paintings on the walls.

In retrospect, it is surprising to see that the important changes in Van Doesburg’s work of 1923–24 were already indicated by Mondrian’s work of 1918–19. Mondrian started tentatively with diagonal compositions, arguing that he could that way dissolve the natural appearance of things—the same argument later used by Van Doesburg. But in the end he preferred the equilibrium of his usual horizontal-vertical compositions within a diamond format, its dynamics remaining implicit. Van Doesburg developed from his architectural studies of 1923 the concept of “Counter-Composition” (for which he later adopted the term “Elementarism”): reintroducing diagonals within his paintings, he made the dynamics of the compositions much more explicit.

Although it is commonly believed that Mondrian condemned Van Doesburg’s use of the diagonal, in view of his own experiments of only a few years before that would seem rather strange. The controversy was not that simple. After all, Mondrian broke definitely with De Stijl as late as December 1927. Probably what offended him most was Van Doesburg’s proposition that Elementarism was a fundamental improvement on Neoplasticism, and somehow more appropriate to modern times: “Elementarism was born, in part, as a reaction to a too dogmatic and often short-sighted application of Neoplasticism, in part as its consequence, but finally and mainly it was born from a severe correction of Neoplasticist ideas.” 17 At the tenth anniversary of De Stijl, in 1927, Mondrian must have felt a stranger in his own house. That was when he wrote: “Dear Mr. Van Doesburg, after your self-righteous improvement (?) of Neoplasticism any collaboration whatsoever is impossible for me. I feel sorry that I cannot prevent the reproduction of articles and photographs in De Stijl as it now is. For the rest, sans rancune. Piet Mondriaan.”18

Carel Blotkamp is a member of the Kunsthistorisch Instituun of the Rijksuniversitein Utrecht.



1. Carmean, Jr., Mondrian, the Diamond Compositions, Washington, 1979.

2. Carmean, Jr., op. cit., pp. 18, 50–52.

3. Michel Seuphor, Piet Mondrian: Life and Work, New York, 1956, cat. nos. 285, 286, 287, 288, 289, 301. Also, two painnings known only from surviving photographs, one reproduced in De Stijl, March 2, 1919, n.p., the other published by Ankie de Jongh, De Stijl, Museumjournaal, Ivii (1972), p. 273 (fig. 2).

4. Carmean, op. cit., p. 21.

5. Relevant in this context are Mondrian’s letters no the critic and art consultant H.P. Bremmer, published by J. Joosten, “Documentatie over Mondriaan,” Museumjournaal, XIII (1968), pp. 208–15, 267–70, 321–26; to the Reverend H. Van Assendelft, Museumjournaal, XVIII (1973), pp. 172–179, 218–23; and the letters to Van Doesburg, unpublished save for some fragments. Quotations here are from transcripnions of the letters in the Van Doesburg archive, formerly at Meudon.

6. This fragment published in the Stedelijk Museum, De Stijl, Amsterdam, 1951, p. 44,

7. Manifest I of “The Style,” 1918, published in English in De Stijl, November 2, 1918, p. 4.

8. Letter to Van Doesburg, July 17, 1918: “Now about the manifesto. As you left out that part on neoplasticism than I wanted to have included . . . ”

9. Cf. Robert Welsh, “The Place of ‘Composinion 12 with Small Blue Square’ in the Art of Piet Mondrian,” Bulletin of the National Gallery of Canada (Ottawa), XXIX (1977), pp. 7f; Andrea Gasten, “Pseudo-mathematica en beeldende kunst,” Kunsternaren der Idee, Symbolistische tendenzen in Nederland, ca. 1880–1930, The Hague, 1978, pp. 59–66.

10. De Stijl, June 1, 1918, p. 90, “inner” referring to the spiritual, “outer” no the material world.

11. Letter of Van Doesburg to Van der Leck, April 3, 1918, quoted by R.W.D. Oxenaar, Bart van der Leck tot 1920, een primitief van de nieuwe tijd (Ph.d. diss., Utrecht), The Hague, 1976, p. 128f.

12. Theo van Doesburg, “Aantekeeningen bij Bijlage XII, De zaag en de goudvischkom van P. Alma,” De Stijl, June 1, 1918, pp. 91–94.

13. Two paintings of this type by Huszar are reproduced and discussed by Theo van Doesburg, “Over het zien van nieuwe schilderkunst,” De Stijl, February 2, 1919, pp. 42–44.

14. J. Joosnen, “Documentatie over Mondriaan,” Museumjournaal, XVIII (1973), p. 222.

15. Joop J. Joosnen, “Abstraction and Compositional Innovation,” Artforum, April 1973, pp. 54–59, n. 22. Steenhoff reviewed the exhibition in the weekly De Nieuwe Amsterdammer, March 8, 1919, and noted in Mondrian’s work “a return to the subdued, hazy, musical color of years ago. . . . He should, however, be cautious of the strictness as in Composition B.” This probably refers to the most colorful and strict painting, Composition in Diamond Shape, which was soon after published in De Stijl, February 3, 1920, n. p., as Kompositie B.

16. Van Doesburg’s design was illustrated in an article by J. J.P. Oud, “Glas in lood van Theo van Doesburg,” Bouwkundig Weekblad, XXXIX (1918), pp., 199–202.

17. Theo van Doesburg, “Schilderkunsn en plastiek; Elementarisme (Manifest-fragment),” De Stijl, VII (1926–27), p. 82.

18. Letter of December 4, 1927, published in Stedelijk Museum De Stijl, (note 6), p. 72.