The first of these three plans, it is true, is the most scientific, since it depends upon the mind, and a thorough knowledge of the relations and effects of colours; while the second depends wholly upon the eye, and is simply the method of sense. This distinction applies also to the two methods which have prevailed with different artists, by the most ordinary of which the painter blends his colours and forms his tints upon his palette, as was probably done by Titian and the Venetian school; and the other by which he makes as it were a palette of his picture, applying his colours unbroken, and producing his combinations therewith upon his canvass, as appears to have been the method of Rubens and Reynolds. To the first, a good eye is principally necessary; while the latter depends upon a better knowledge of colours, - it is also more favourable to the brilliancy, purity, and durability of colouring, according to the foregoing maxims. Here also the best practice is a compound of the two extreme methods, and in some measure essential to good colouring, and appears to have been acted upon in the landscapes of Gainsborough and Wilson.

The practice of producing tints and hues by grinding pigments together, instead of blending them on the palette, has fallen into disuse, whether advantageously or otherwise may be questioned; but to this practice may be attributed some peculiarities of the tints and textures of the pictures of the Flemish school, they being perhaps results of intimate combination from grinding, and consequently of a more powerful chemical action among the ingredients compounded. It conduces also undoubtedly to that union upon which tone and mellowness depend, when the same pigments which lie near together in a picture are employed to form intermediate hues and tints; but this practice conducts to foulness when the colours of such pigments are not pure and true, and do not assimilate well in mixture chemically.

The superiority of Rubens and the Flemings, and of Titian and the Venetian school, in colouring and effect, is to be attributed in a considerable measure to their sketching their designs in colours experimentally with a full palette; and this practice, as derived from Reynolds, is common with the best masters of our own school, who resort also, in executing their works, to nature, with an improved knowledge of colours and colouring. This attention to colouring and effect from the first study and ground of a picture to the finishing, contributes also a beauty and durability to the work which no superinduced colouring can accomplish.

We could not well avoid this digression on the modes of practice, upon which durability so much depends: for so concurrent is permanence with reputation, and so important is it, that had the sculpture of the Greeks been no more durable than their painting, literature could not have preserved the fame of their artists; and the reputation of their painters in our time is principally conceded to the transcendent excellence of their sculpture which still remains. Hence the modem artist, who is inattentive to the durability of his materials, may content himself with sharing some portion only of the reputation of the engraver, whose art will become to modern times in this respect, what sculpture in relief has been to the antient; for as to colouring in particular, which has been called mechanical and subordinate, it is the only department of painting which cannot be copied and transferred mechanically by a copperplate and a press, but requires a cultivated taste and judgment, a fine eye, and an able hand, united immediately in the work.

Abstractedly considered, it is probable that the durability of colour in substances is uniformly dependent upon the state in which they exist chemically or by constitution, with respect to what we have before regarded as the two concurrent and essential principles of light and colours, which we have distinguished by the terms oxygen and hydrogen; and this may account for the apparent capriciousness by which a pigment is found sometimes durable and sometimes not so, particularly the lakes, carmines, and most vegetal colours, the fugitiveness of which depends as much upon the state of their bases as upon the natural infirmity of their colouring matter. If the pigment or its base be in a state which the chemists have termed a protoxide, it will, by a gradual acquisition of oxygen from light, air, or moisture, change or fade, till, being saturated, or becoming a peroxide, it is no farther subject to change by the election of oxygen.

On the other hand, pigments and bases assume similar states with respect to the hydrogenous principle of light, etc. in which they may be termed prothydrtds and perhijdrids, in which states they are subject to changes opposite but analogous to the preceding; and to this latter influence metallic colours and their bases are principally subject: upon the whole, however, oxygen being the more active of the two principles, colours are in a greater measure subject to its influence. Such is the subtile chemistry, in the simplest view we are able to take of it, upon which the changes of colours in pigments depend individually; and upon these states of pigments depend also the various ways in which they affect each other in mixture.

It would not be difficult to explain, upon the same agency, why pigments are more subject to oxidation and fading in a water vehicle, and to hydrogenation and darkening in one of oil.

With respect also to permanence, it is worthy of remark, that recent pigments, both natural and artificial, like recent pictures, wines, etc. undergo amelioration by the influences of time, temperature, atmosphere, etc. which it is better they should suffer in the state of pigments than upon the canvass. Their drying in oil is in general also improved by age, and they approach nearer altogether to the condition of native pigments: it is worthy of attention, also, that grinding is not only a mechanical but a chemical agent, affecting considerable change in the colour and constitution of pigments, every chemical compound being more or less subject to decomposition by friction. It seldom happens, however, that pigments are injured by excess of grinding.

These effects of time, etc. are additional reasons for the esteem in which the colours as well as the works of celebrated deceased artists are commonly held; and this accords with the common remark, that time effects a mellow and harmonious change upon pictures: but sometimes it produces changes altogether unfavourable. To insure the former, and prevent these latter changes, the attention of the artist, in the course of his colouring, should be directed to the employment of such colours and pigments as are prone to adapt themselves in changing to the intended key of his colouring and the right effect of his picture. For example, if he design a cool effect, ultramarine has a tendency through time to predominate, and to aid the natural key of blue: he will, therefore, compromise the permanence of his effect, if in such case he employ a declining or changeable blue, or if he introduce such reds and yellows as have a tendency to warmth or foxiness, by which the colouring of many pictures has been destroyed. In a glowing or warm key the case is in some measure reversed, - not wholly so, for it is observable that those pictures have best preserved their colouring and harmony in which the blue has been most lasting, by its counteracting the change of colour in the vehicle, and that suffusion of dusky yellow which time usually bestows upon pictures even of the best complexion.

Newly discovered pigments, however flattering in appearance or in working, are to be employed with caution, or even suspicion, till experience has obtained them the stamp of excellence. Good pigments have ever been prized with so true an estimation of their value by all people, whether barbarian or refined, that it may be doubted if a really excellent one has ever been lost to the world; and to produce such after the ages of research which have passed, and for which all who have had eyes have been in a measure qualified, is, we may be assured, no ordinary result either of accident or design. Accordingly, most of the resplendent pigments, fruits of the fecundity of modern chemistry, have been found deficient of the intrinsic and sterling excellences which have given value and reputation to some of the antient and approved. Thus the splendid yellow chromates of lead, which withstand the action of the sunbeam, become by time, foul air, and the influence of other pigments, inferior even to the ochres. So again Indian yellow, which also powerfully resists the sun, is soon destroyed in oil, and changed by time, etc. Again, the vivid and dazzling reds of iodine are chameleon colours, subject to the most sudden and opposite changes, and yield the palm of excellence to the fine solid colours of vermilion, whose name they usurp and falsify. So again the brilliant blues of cobalt, beautiful and abundant as they are, which resist the sunbeam powerfully and have no present imperfection, are always tending to greenness and obscurity, and must yield their pretensions on the palette to the unrivalled excellences of ultramarine.

These, among the chief productions of modern chemistry, are valuable for the ordinary and temporary purposes of painting; but they captivate the eye by a meretricious beauty which misleads the judgment, and are to be introduced with great caution in the more elevated practice of the art: for we are not sufficiently acquainted with the many effects upon which colours are dependent to determine by what substances or circumstances a pigment may or may not be changed or preserved in all cases; it is, therefore, only by long experience and varied practice that the value of a pigment can be fully ascertained, and the exceplions to its use established, although it may stand the most necessary tests upon which its durability is to be presumed.

It is upon the test of time, and not according to their beauty or other qualities, that we have pronounced a preference of antient to modern, and of natural to artificial, pigments. Those known to have been employed by the antients are, for the most part, natural pigments, while the pigments of modern discovery are almost entirely productions of art. Upon these and other points of durability, etc. the tables of Chap. xxii. may be consulted.

As to the individual permanency or fugitiveness of pigments, we have noted them under their respective heads as they occur in the following chapters, and in what respects they are affected by vehicles we have remarked in a chapter on Vehicles.