Some Experiments and Observations on the Colours used in Painting by the Ancients .
By Sir Humphry Davy , LL. D. F.R.S. Read February 23 , 1815 .
[ Phil. Trans. 1815 , p. 97 .
] Beside the use which may be made of what remains of ancient paintings as models for imitation , the author has endeavoured to reap the further advantage of making us acquainted with the nature and chemical composition of their colours ; for though the works of Dioscorides , Vitruvius , and Pliny contain descriptions of many substances used by the ancients as pigments , it is only by experiment that the subjects of which they speak can be identified .
The author 's experiments have been made upon colours found in the baths of Titus , in the ruins called the baths of Livia , and other ruins of ancient Rome , and in the ruins of Pompeii .
Some of these colours had been discovered in vases beneath the ruins of the palace of Titus , and were found to be the same as those used in various fresco paintings of the palace .
In one large vase , discovered about two years since , there were found , among other colours , three different kinds of red , one approaching to orange , another dull red , and a third purplish red .
The first was minium , the second and third proved to be both ochres of different tints .
Another red found in various fresco paintings differed from those found in the vase , and proved to be vermilion .
This substance , called by the Greeks Kicvv'aapp , was known -by the name of minium to the Romans , who called our modern minium by the name of cerussa usta , in consequence of the mode of making it ; which , on the authority of Pliny , is said to have been suggested by the accidental effects of a fire at the Pirseeus at Athens , by which ceruse was found converted into minium .
From the description which Pliny gives of an inferior sort of vermilion , formed by calcining certain stones found in veins of lead , the author is of opinion , that the mineral thus treated must have been a natural carbonate of lead , which becomes red when burned .
Among the yellows examined by Sir Humphry Davy , were ochres of various tints , from being mixed with different quantities of chalk , and the yellow oxide of lead or massicot .
But though we have the evidence of Vitruvius that orpiment was known to the ancients , and of Pliny that a substance nearly allied to orpiment , termed Sandarach , was used by the Romans , the author has not been able to detect either of these sulphurets of arsenic in any of the ancient fresco paintings .
Among some rubbish collected in one of the chambers of the baths of Titus were several large lumps of a deep blue frit , which , upon being analysed , were found to consist of soda , silica , and oxide of copper .
Upon examination of the different tints of blue observable in the paintings of the baths , as well as several blues in fragments of fresco painting from the ruins near the monument of Caius Cestius , and from excavations made at Pompeii , it appeared that they all consisted of the same blue frit , more or less diluted by admixture with carbonate of lime .
There appears to the author every reason to believe this to be the colour described by Theophrastus , as discovered by an Egyptian king , and anciently manufactured at Alexandria .
Vitruvius also speaks of the same colour under the name of caeruleum , made in his time at Puzzuoli , by heating together sand , flores nitri or natron , and filings of copper .
Though Pliny and Vitruvius speak of Indian blue , which appears to have been indigo , the author has not been able to discover any remains of it at this time ; nor indeed of any other blue , excepting the frit before mentioned among the opake blues used by painters .
But it is by no means uncommon to find among the ruins fragments of transparent blue glass , which are tinged with cobalt ; and it would appear , from a passage in Theophrastus , that the Greeks considered cobalt as a species of XaXKcOs , in consequence of its property of giving this blue colour .
Among the several shades of green observable in the baths of Livia , the baths of Titus , and elsewhere , the greater part are coloured by carbonate of copper ; but one of them , which approached the olive , proved to be the common green earth of Verona .
It seems not improbable that some of the greens which are now found in the state of carbonate of copper may have been originally laid on as acetates ; for it appears from Theophrastus that the ancients were well acquainted with verdigris .
The only trace of any thing approaching to the ancient purple prepared from shell-fish is in a broken vase in the baths of Titus , containing a substance which at the surface has become of a cream colour , but in the interior has a lustre approaching to that of carmine .
The colouring matter of this substance was found to be combustible , constituting about one thirtieth part of its weight , the remainder being a compound of siliceous , aluminous , and calcareous earths .
It may , therefore , be regarded as a lake ; but it would be very difficult , if not impossible , at this distance of time , to determine whether it be of animal or vegetable origin .
In either case its durability , even in the interior of the mass , is a very curious circumstance , although the part exposed to the air has suffered the changes to which such colours have been too often proved to be liable , and accordingly no traces of it remain in any of the ancient fresco paintings .
All the blacks observable in the baths of Titus or elsewhere accord with the descriptions given by ancient authors , who speak of them as carbonaceous substances , obtained either as common charcoal or as soots of woods or resins .
The browns are sometimes mere oxides of iron or ochres , and sometimes.mixtures of the oxides of iron and manganese ; and it appears that the Romans had some knowledge of the properties peculiar to the latter substance , as Sir Humphry Davy has analysed two specimens of ancient Roman purple glass , both of which were tinged with manganese .
Among the whites of the ancient paintings , the author was unable to discover any ceruse , although it is known to have been in common use on the authority of Theophrastus , Vitruvius , and Pliny .
The whites found are in general carbonate of lime , or fine white clays .
The ground to which the colours are applied in the ancient fresco paintings , is precisely such as is described by Vitruvius , powdered marble cemented by lime , highly polished and beautifully white .
With regard to the mode in which their colours were applied , Vitruvius and Pliny agree as to the employment of wax in encaustic painting , which was subsequently liquefied by heat so as to give a varnish to the painting .
But the author has not in any instance been able to detect the presence of wax , nor yet of any animal or vegetable gluten , in any of the fresco paintings , or even in the pot of colours found at Pompeii .
From the facts above stated , it appears that the Greek and Roman painters had the advantage over the great Italian masters , since the revival of civilization , in two of their colours , the Tyrian purple and the Egyptian azure , although the latter may easily and cheaply be imitated ; for if a mixture of about fifteen parts of soda , twenty parts of powdered flint , and three parts of copper filings , be strongly heated together for about two hours , a frit is produced extremely similar in appearance and degree of fusibility to the ancient blue frit .

