Dear Theo,

Today I received your letter with enclosure. I was very pleased with your letter, because I noticed a few things in it which I want to talk over. To begin at the beginning; what you write about a certain study of a basket with apples is very well observed, but does this observation come from yourself??? because I fancy, I should almost say I am sure, that you used not to see that kind of thing: However this may be, here we are on our way to agreeing more about the colours.

Go more deeply into those questions, for that will be useful to you, and those are the things that Burger and Mantz and Silvestre knew.

Just to explain how that study was painted - simply this: green and red are complementary colours. Now in the apples there is a red which is very vulgar in itself; further, next to it some greenish things. But there are also one or two apples of another colour, of a certain pink which makes the whole thing right.

That pink is the broken colour, got by mixing the above-mentioned red and the above-mentioned green. That's why there is harmony between the colours.

Added to this is a second contrast, the background forms a contrast to the foreground, the one is a neutral colour, got by mixing blue with orange; the other, the same neutral colour simply changed by adding some yellow.

But I am awfully glad that you notice a combination of colour, be it through direct or indirect personal perception. Further, that one of the studies seemed to you a variation on the brown-grey theme, well, that certainly is the case, but all three potato studies are like that, with this difference, that one is a study in terre de Sienne, the second in terre de Sienne brûlée, the third in yellow ochre and red ochre.

The latter - that is the largest one - is in my opinion the best - notwithstanding the dull black background which I purposely left dull because the ochres are also naturally non-transparent colours. As to that study, the largest one of the potatoes, it is made by changing, by breaking, those untransparent ochres with a transparent blue. As red ochre with yellow ochre gives orange, their combination with blue is more neutral, and against that neutralized colour, they become either more red or more yellow.

The highest light in that whole picture is simply some pure yellow ochre. The reason why this dull yellow stands out so is because it is put in a wide field of, be it neutral, violet; because... red ochre with blue gives violet tones.

Well, the birds' nests were also purposely painted against a black background, because I want it to be obvious in these studies that the objects do not appear in their natural surroundings, but against a conventional background. A living nest in nature is quite different - one hardly sees the nest itself, one sees the birds.

But when one wants to paint nests from one's collection of nests, one cannot express strongly enough the fact that the background and the surroundings in nature are quite different, therefore I simply painted the background black. But it is a fact that in a still life a coloured background can be beautiful - in Amsterdam I saw still lifes by Miss Vos that were excellent, much more beautiful than those by Blaisse Desgoffe [a French still-life painter of the mid-nineteenth century] - really like Van Beyeren. I couldn't help thinking that those simple still lifes of hers had far more artistic value than many pretentious pictures by other Amsterdam painters.

They struck me as very well done. Especially one with a golden vase, a few empty oyster shells, a broken coconut shell and a crust of bread. I will send you the book by Blanc; I hope soon to get L'art auXVIIIme Siècle; I am especially longing to hear something from de Goncourt about Chardin. Lacaze's Rembrandt is really also in the sentiment of Rembrandt's last period; it is about twelve years since I saw it, but I still remember it because it struck me, just like that head by Fabritius in Rotterdam. If I remember correctly, that nude woman in the Lacaze Collection is also very beautiful, also of a later period. The fragment, Rembrandt's “Lesson in Anatomy,” yes, I was absolutely staggered by that too. Do you remember those flesh colours - it is - de la terre - especially the feet.

You know, Frans Hals's flesh colours are also earthy, used here in the sense that you know. Often at least. Sometimes, I almost dare say always, there is also a relation of contrast between the tone of the costume and the tone of the face. Red and green are opposites; “The Singer” (Dupper Collection), who has tones of carmine in the flesh colour, has tones of green in his black sleeves, and ribbons on those sleeves of a red other than that carmine. The orange-white-blue fellow I wrote about has a relatively neutral complexion, earthy-pink, violetish, in contrast with his Frans-Hals-yellow leather suit.

The yellow fellow, citron amorti, decidedly has dull violet in his mug. Well - the darker the costume, the lighter the face is sometimes - not accidentally - at least his portrait and that of his wife in the garden contain two blackish violets (blue-violet and reddish-violet) and a plain black (yellow-black?). I repeat, reddish-violet and blue-violet, black and black, the three gloomiest things, as it were; well, the faces are very fair, extremely fair, even for Hals.

Well. Frans Hals is a colourist among colourists, a colourist like Veronese, like Rubens, like Delacroix, like Velásquez.

Of Millet, Rembrandt and, for instance, Israëls, it has truly been said that they are more harmonists than colourists. But tell me, black and white, may they be used or may they not, are they forbidden fruit?

I don't think so; Frans Hals has no less than twenty-seven blacks. White - but you know yourself what striking pictures some modern colourists make of white on white. What is the meaning of that phrase: one must not? Delacroix called them rests, used them as such. You must not have a prejudice against them, for if used only in their places, and in harmony with the rest, one may of course use all tones.

I can tell you that I often think the things by Apol, for instance, white on white, very well done. His sunset in The Hague Wood, for instance, which is in Amsterdam. That thing is damn good indeed.

No - black and white have their reason and significance, and when one tries to suppress them, it turns out wrong; to consider both neutral is certainly the most logical thing to do, white - the highest combination of the lightest red, blue, yellow; black - the highest combination of the darkest red, blue, yellow. I have nothing to say against that theory, I find it perfectly true. Well, light and brown, the tone in its value stands in direct relation to that 4th color scale from white to black. For one finds there:

That is how I understand the blacks and the whites.

When I mix red with green to a red-green or green-red, by mixing it with white, I then get pink-green or green-pink. And if you like, by adding black, I get brown-green or green-brown. Isn't that clear? When I mix yellow with violet to a violet-yellow or yellow-violet, in other terms a neutralized yellow or a neutralized violet, by adding white and black, I get greys.

Well, greys and browns, there is especially question of them when one makes colours lighter or darker, whatever their nature and their gradation of red, yellow or blue may be.

It is quite correct to speak of light and dark grays and browns, I think. But how beautiful what Silvestre says about Delacroix is - that he put a fortuitous tone on his palette, une nuance innommable violacée, that he put that one tone down somewhere, either for highest light or for deepest shadow, but that of this mud he made something which either sparkled like light or was gloomily silent like a deep shadow.

So I have heard of an experiment with a sheet of neutral coloured paper - which became greenish against a red background, reddish on a green one, bluish on orange, orange on blue, yellowish on violet, and violetish on yellow.

Just listen, suppose one wants such a muddy tone or drab colour to become light in the picture, like Delacroix said of Veronese, that he could paint a blonde nude woman with a colour like mud in such a way that she comes out fair and blonde in the picture - then the question arises - how is this possible, unless by contrast of great forces in bluish-blacks or violets, or reddish-browns?

You - who are looking for dark shadows somewhere, and think that when the shadows are dark, aye, black, that it is all wrong then, is this right? I don't think so. For then, for instance, the “Dante” by Delacroix, the “Fisherman of Zandvoort,” for instance, would be wrong. For indeed, they contain the most vigorous blue-black or violet-black values. Rembrandt and Hals, didn't they use black? and Velásquez???

Not only one, but twenty-seven blacks, I assure you. So as to “one must not use black,” are you yourself quite sure that you know what you mean by it? and do you know what you want with it? Really, think it over carefully, for you might come to the conclusion - I think this very probable - that you have learned and understood that question of tones quite wrongly, or rather have learned it vaguely and understood it vaguely. Many people do, most of them do. But in the long run Delacroix and others of his time will teach you better.

Tell me - have you noticed that those studies of mine that have black backgrounds have their highest light put in a low colour scale??? And when in this way I put my study in a lower colour scale than nature, I yet keep the harmony of tones because I become darker, not only in my shadows, but also in the same degree in my lights.

I painted my studies just as a kind of gymnastics, to rise and fall in tone, so - don't forget that I painted my white and gray moss literally with a mud colour and yet it looks light in the study.

Good-by,

Ever yours, Vincent

These things concerning complementary colours, simultaneous contrast, and the neutralizing of complementals, this question is the first and principal one; the second is the mutual influence of two kindred colours, for instance, carmine on vermilion, a pink-violet on a blue-violet. The third question is a light blue against the same dark blue, a pink against a brown-red, a citron yellow against a chamois yellow, etc. But the first question is the most important.

If you come across some good book on colour theories, mind you send it to me, for I too am far from knowing everything about it, and am searching for more every day.